Walden University

ScholarWorks

Walden Dissertations and Doctoral Studies

Walden Dissertations and Doctoral Studies Collection

2021

Factors Influencing Family Planning Uptake Among Adolescents

and Postpartum Women in Kenya
Shiphrah Kuria-Ndiritu
Walden University

Follow this and additional works at: https://scholarworks.waldenu.edu/dissertations

Part of the Medicine and Health Sciences Commons

This Dissertation is brought to you for free and open access by the Walden Dissertations and Doctoral Studies
Collection at ScholarWorks. It has been accepted for inclusion in Walden Dissertations and Doctoral Studies by an
authorized administrator of ScholarWorks. For more information, please contact [email protected].
 Walden University

College of Health Professions

This is to certify that the doctoral study by

Shiphrah Kuria-Ndiritu

has been found to be complete and satisfactory in all respects,

and that any and all revisions required by
the review committee have been made.

Review Committee
Dr. Stacy-Ann Christian, Committee Chairperson, Public Health Faculty
Dr. Claire Robb, Committee Member, Public Health Faculty
Dr. W. Sumner Davis, University Reviewer, Public Health Faculty

Chief Academic Officer and Provost

Sue Subocz, Ph.D.

Walden University
2020
 Abstract

Factors Influencing Family Planning Uptake Among Adolescents and Postpartum

Women in Kenya

by

Shiphrah Kuria-Ndiritu

Doctoral Study Submitted in Partial Fulfillment

of the Requirements for the Degree of

Doctor of Public Health

Walden University

March 2021
 Abstract

Family planning (FP) is a cost-effective public health strategy, but the uptake is low with

marked disparity among adolescents and postpartum women. However, data on these

marginalized groups are limited. This quantitative, cross-sectional study sought to

provide information on the factors that contribute to the uptake of FP among adolescents

and postpartum women in Kenya. The 2014 Kenya Demographic Health Survey FP data

were analyzed regarding the factors associated with FP uptake among adolescents and

postpartum women as well as the differences by region. The factors were organized

according to the socioecological model (SEM) and included intrapersonal, interpersonal,

community, and organizational levels of influence. The logistic regression model was

used to determine the contribution of different factors to the uptake of FP. The results

indicated that compared to adolescents, older women have better socioeconomic status,

and a higher proportion are using modern FP methods. On bivariate analysis, factors at all

levels of the SEM were associated with uptake of FP in both groups. On regression

analysis, factors that contributed most significantly to the uptake of FP were at the

intrapersonal and organizational levels among adolescents and at the intrapersonal and

interpersonal levels among postpartum women. Recommendations include further

research on empowering adolescents to make informed choices in FP, FP interventions

focusing on more than one level of SEM, and regional disparities being addressed. The

information provided by this study can contribute to improved FP uptake and positive

social change for adolescents and postpartum women, which means better health,

economic benefits, and improved quality of life.

Factors Influencing Family Planning Uptake Among Adolescents and Postpartum

Women in Kenya

by

Shiphrah Kuria-Ndiritu

Doctoral Study Submitted in Partial Fulfillment

of the Requirements for the Degree of

Doctor of Public Health

Walden University

March 2021
 Dedication

This work is dedicated to the marginalized women who struggle to get services,

particularly skilled deliveries, and family planning, that their needs may be realized

sooner than later.

 Acknowledgments

First and foremost, I would like to thank the almighty God who has given me the

grace and good health to be resilient throughout the period when I worked on this

dissertation and throughout the very intense course work. To him, be the glory.

Secondly, special thanks to my committee members; The chair, Stacy-Ann Nicola

Christian, the second member, Dr. Claire Robb, and the university research reviewer, Dr.

W Sumner Davis, for the support and guidance without which this work would not have

been possible.

To my course and residencies’ instructors, who imparted knowledge and skills

throughout my course work, I say a big thank you all. Special thanks to Dr. Vasileios

Margaritis for the useful initial guidance on the prospectus and the proposal writing. All

faculty, your effort, and contribution are highly appreciated; it shall be used to touch and

improve lives, particularly those of women and children.

To my family, my husband, Dr. Simon Ndiritu, and my children, Neema, Abby,

and Joe, I am most grateful for standing with me and allowing me time to work on this,

sometimes even at the expense of being with you. Your support will be cherished forever.

To my extended family, particularly my siblings, whose encouragement and prayers

motivated me.

To my friends and colleagues who encouraged me and some who took the time to

give input to my work; my sincere gratitude to you. I want to mention Dr. Alice Lakati,

who guided me and encouraged me through the analysis.

 Table of Contents

List of Tables................................................................................................................... v

List of Figures ...............................................................................................................vii

Section 1: Foundation of the Study and Literature Review ............................................... 1

Introduction ............................................................................................................... 1

Description of the Research..................................................................................2

Family Planning and Adolescents.........................................................................2

Family Planning Among Postpartum Women .......................................................4

Importance of Family Planning ............................................................................4

Importance of this Study ......................................................................................6

Problem Statement ..................................................................................................... 7

Purpose of the Study ................................................................................................ 10

Research Questions and Hypotheses ........................................................................ 11

Theoretical Foundation for the Study ....................................................................... 13

Nature of the Study .................................................................................................. 18

Literature Search Strategy ........................................................................................ 20

Literature Review .................................................................................................... 21

Family Planning Methods .................................................................................. 21

Unmet Need for Family Planning ....................................................................... 26

Family Planning ................................................................................................. 48

Modern Methods of Family Planning ................................................................. 49

Contraceptive Prevalence Rate ........................................................................... 50

i
 Unmet Need for Family Planning ....................................................................... 51

Method Mix ....................................................................................................... 51

Adolescents........................................................................................................ 51

Postpartum Women and Postpartum Family Planning ........................................ 52

Assumptions ............................................................................................................ 52

Scope and Delimitations .......................................................................................... 53

Significance, Summary, and Conclusions ................................................................. 53

Section 2: Research Design and Data Collection ............................................................ 56

Introduction ............................................................................................................. 56

Research Design and Rationale ................................................................................ 56

Study Variables .................................................................................................. 56

Research Design ................................................................................................ 57

Methodology ........................................................................................................... 59

Population .......................................................................................................... 59

Sampling Strategy and Procedure ....................................................................... 59

Sampling Frame ................................................................................................. 60

Sample Size ....................................................................................................... 61

Procedure ........................................................................................................... 63

Instrumentation and Operationalization of Constructs ........................................ 64

Threats to Validity ................................................................................................... 72

Ethical Procedures ............................................................................................. 72

Summary ................................................................................................................. 73

ii
Section 3: Presentation of the Results and Findings ....................................................... 74

Introduction ............................................................................................................. 74

Data Collection of Secondary Data Set .................................................................... 74

Sampling and Sample Characteristics ................................................................. 75

Analysis ............................................................................................................. 76

Results 79

Adolescents........................................................................................................ 79

Association of Various Factors with the Uptake of Family Planning

Among Adolescents in Kenya ................................................................. 87

Extent of Association of Family Planning Use with Various

Characteristics Among Adolescents ........................................................ 97

Postpartum Women ............................................................................................ 98

Association of Various Factors with the Uptake of Family Planning

Among Postpartum Women in Kenya ................................................... 103

Association of Different Characteristics with Use of Family Planning

Among Postpartum Women .................................................................. 105

Extent of Association of Family Planning Use with Various

Characteristics Among the Postpartum Women .................................... 111

Distribution of Factors Associated with Family Planning Uptake Among

Adolescents Across the Counties in Kenya ........................................... 113

Distribution of Factors Associated with Family Planning Uptake Among

Postpartum Women Across the Regions in Kenya ................................ 118

iii
 Summary ............................................................................................................... 121

Section 4: Application to Professional Practice and Implications for Social

Change ............................................................................................................. 124

Introduction ........................................................................................................... 124

Key Findings.................................................................................................... 124

Interpretation of the Findings ................................................................................. 125

Adolescents...................................................................................................... 126

Factors Associated with Use of Family Planning Among Adolescents.............. 130

Postpartum Women .......................................................................................... 135

Different Characteristics with Family Planning Among Postpartum

Women ................................................................................................. 138

Distribution of Factors Associated with Family Planning Uptake Across

the Regions in Kenya ............................................................................ 144

Limitations of the Study ......................................................................................... 145

Recommendations.................................................................................................. 146

Implications for Professional Practice and Social Change ...................................... 148

Recommendations for Professional Practice ..................................................... 148

Kenya Demographic and Health Survey Methodological Implications ............. 149

Theoretical and Empirical Implications ............................................................ 149

Positive Social Change ..................................................................................... 149

Conclusion............................................................................................................. 150

References ................................................................................................................... 151

iv
 List of Tables

Table 1 Variables for Measuring Socioecological Model Constructs ............................ 17

Table 2 Sociodemographic Characteristics of the Respondents .................................... 81

Table 3 The Pattern of Family Planning by the Adolescents and Older Women ............ 82

Table 4 Current Use of Family Planning by Adolescents and Older Women by

Specific Method.................................................................................................. 83

Table 5 Family Planning Methods Discontinuation by Adolescents and Older

Women by Specific Method ................................................................................ 84

Table 6 Reason for Discontinuation of Family Planning Methods

Discontinuation by Adolescents and Older Women by Specific Method .............. 84

Table 7 Reasons for not Using Family Planning Methods by Adolescents .................... 86

Table 8 Information on Family Planning given to Adolescents who Obtained

Family Planning Methods .................................................................................. 87

Table 9 Summary of Chi-Square Results for Sociodemographic Characteristics

and Use of Modern Methods of Family Planning Among Adolescents ................ 90

Table 10 Extent of Association of Family Planning Use Among the Adolescents

and Various Characteristics ............................................................................... 98

Table 11 Sociodemographic Characteristics of Postpartum and Other Women .......... 101

Table 12 Pattern of use Among Postpartum and Other Women .................................. 102

Table 13 Use of Family Planning Modern Methods Among Postpartum and

Other Women ................................................................................................... 102

v
Table 14 Type of Method discontinued Among Postpartum Women and Other

Women ............................................................................................................. 103

Table 15 Chi-square Results of the Association of Sociodemographic

Characteristics and Use of Family Planning .................................................... 105

Table 16 Distribution of Various Characteristics Among the Adolescent Across

Regions ............................................................................................................ 115

Table 17 Distribution of Adolescents’ Socio-demographic Characteristics across

regions Chi-square Test Results ....................................................................... 116

Table 18 Distribution of Adolescents’ Various Characteristics Across Regions

Chi-Square Test Results ................................................................................... 117

Table 19 Distribution of Selected Characteristics of Postpartum Women Across

the Regions in Kenya........................................................................................ 119

Table 20 Distribution of Postpartum Women’s Sociodemographic

Characteristics Across Regions Chi-Square Test Results .................................. 120

Table 21 Distribution of Postpartum Women’s Various Characteristics Across

Regions Chi-Square Test Results ...................................................................... 121

vi
 List of Figures

Figure 1. Conceptual Framework for Factors Associated with Family Planning Services

Uptake ................................................................................................................... 16

vii
 1
Section 1: Foundation of the Study and Literature Review

Introduction

Family planning (FP) refers to the preparation, knowledge, and methods that

assist individuals and couples to plan and attain their desired family size and determine

the spacing of pregnancy (World Health Organization [WHO], 2018). FP was prioritized

internationally during the 1970s and 1980s with significant support, which led to an

increase in contraceptive prevalence rate (CPR) with reduced fertility globally (Cleland et

al., 2006; Mwaikambo et al., 2011). Kenya was among the first African countries to

recognize the challenges associated with their high fertility and to embrace FP (Stiegler

& Susuman, 2016). The family planning program was established when the total fertility

rate was high at around eight children per woman and a low CPR of 7% in the 1970s

(Cleland et al., 2006; Sibanda, 2010; Sindiga, 1985). With the establishment of the FP

program, a sustained decrease in the total fertility rate with a rapid increase in

contraceptive use ensued until a stagnation in the 2000s (Crichton, 2008) due to

dwindling international support that led to funding cuts for FP programs (Cleland et al.,

2006; Mwaikambo et al., 2011). The stall was more severe among specific subgroups

such as younger women and those with low formal education.

Renewed focus on FP internationally and nationally has led to some success;

however, the targets have not been met, and inequity has persisted. The discussion has

broadened to include the impact of FP not only on the health and rights of the women and

girls but also on socioeconomic development and demographic dividend (Cleland et al.,
 2
2006; Hardee et al., 2014). Hence, there is a need to provide more focused evidence to

support the ongoing efforts.

Description of the Research

Uptake of FP is a cost-effective public health strategy that faces many challenges

(Ganatra & Faundes, 2016). Adolescents and postpartum women are priority groups

because they have a higher unmet need for FP than the general population (Moore et al.,

2015; Vogel et al., 2015). Unmet need refers to the proportion of sexually active, women

of reproductive age who are capable of becoming pregnant but want to limit (do not want

more children) or to space (postpone pregnancy) their children; however, they are not

using FP methods (Kennedy et al., 2011). Unmet need is usually regardless of the reason

why they are not using FP. Hence, there is a need for evidence that can support FP

programs to enhance FP uptake in these priority groups. The current study was conducted

on factors influencing FP uptake among adolescents and postpartum women in Kenya

based on the 2014 Kenya Demographic and Health Survey (KDHS). The independent

variables and covariates are based on the socioecological model (SEM) and were grouped

as intrapersonal, interpersonal, community, and organizational.

Family Planning and Adolescents

The adolescence stage in life is critical for the realization of individuals’ potential

(Patton et al., 2016). But there has been limited investment in adolescents’ health due to

their generally good health despite facing risk and inequities when it comes to maternal

health. Despite gaps in the data for adolescents, evidence points to disparity with

adolescents’ use of contraceptive services compared to older women, which varies across
 3
and within countries (Vogel et al., 2015). Maternal health indicators such as maternal

mortality and contraceptive prevalence have made improvements globally over the last

several decades but not uniformly across all ages. The inequity in contraception uptake is

particularly critical for adolescents because contraception contributes toward securing

their future (Patton et al., 2016). Adolescents are also at significantly higher risk of

maternal mortality and other adverse pregnancy outcomes, which is related to various

factors such as the characteristics of the population, particularly socioeconomic status

(Ganatra & Faundes, 2016).

Contraception is an effective intervention in preventing unintended pregnancies,

which are associated with increased risks of poor pregnancy outcomes such as death and

unsafe abortions (Chandra-Mouli et al., 2014; Glasier et al., 2006; Nove et al., 2014). But

many births by women below 20 years of age in developing countries are unplanned,

which may be an indicator of the widespread burden of unmet need for contraception

(Bishwajit et al., 2017). Kenya is among the 10 countries with the highest teenage

pregnancy globally (Loaiza & Liang, 2013). According to the KDHS, in 2014, 18% of

girls between 15 and 19 years had begun childbearing, meaning they had already given

birth or they were pregnant (Kenya National Bureau of Statistics, 2015). Early

childbearing in Kenya is higher in some regions being highest in Nyanza followed by Rift

Valley and Coast and was lowest in Central and North Eastern region. These differences

may point to inequity in contraception access, but there may be other factors contributing

to this state. Notably, the proportion of teenagers who had begun childbearing had not

changed since the previous KDHS carried out in 2008, thus indicating no progress in the
 4
utilization of contraception in the country among teenagers. Reports from some regions

in Kenya and those from other countries suggest that adolescents face various barriers to

using contraception, including lack of access, health concerns, and fear of side effects

(Ochako et al., 2015; Woog et al., 2015). This study seeks to provide information on

adolescents in Kenya that is nationally representative.

Family Planning Among Postpartum Women

Evidence suggests that optimal birth spacing, 2 to 3 years, enhances maternal and

infant health and contributes to the reduction of maternal mortality (Ganatra & Faundes,

2016). Contraception reduces the high-risk births associated with short interval births and

reduces fertility (Brown et al., 2015). However, though after the delivery of a child, many

women desire to delay pregnancy for at least 2 years (Pasha et al., 2015), many do not

start contraception within the first year, thus risking a closely spaced pregnancy (Rossier

et al., 2015). There is a high unmet need for FP among postpartum women in low- and

middle-income countries, including Kenya, with more than half of repeat births being

within an interval that is too short (Moore et al., 2015). Unmet need for spacing is high at

29% and that for limiting at 28% in Kenya for postpartum women, with 50% of

postpartum women having short birth intervals (Moore et al., 2015).

Importance of Family Planning

FP has health benefits as well as economic benefits.

Health Benefits

Kenya has high maternal and child mortality and morbidity. The 2014 national

DHS estimated maternal mortality at 362 deaths per 100,000 live births, neonatal
 5
mortality at 22 deaths per 1,000 live births, infant mortality 39 deaths per 1000 live

births, and under-5 mortality at five deaths per 1,000 live births (Kenya National Bureau

of Statistics, 2015). A maternal death often means the loss of healthy young productive

women at the prime of life, resulting in economic and social losses. But FP is a cost-

effective public health strategy that improves maternal and child survival (Ganatra &

Faundes, 2016). Contraception has been shown to save hundreds of thousands of

women’s lives within a year by averting maternal deaths (Ahmed et al., 2012).

Economic Benefits

Kenya has continued to develop economically, but the progress is not uniform

with some areas, particularly the hard to reach arid counties and specific segments of the

population such as the youth (Kenya National Bureau of Statistics, 2018). These

marginalized areas are also under-served in regard to services and have the worst

maternal and child health indicators, including low FP uptake (Kenya National Bureau of

Statistics [KNBS], 2018). However, FP is a public health strategy that promotes

economic growth and improves maternal and child health. Fertility decline improves

women’s quality of life and affords them more time to participate in income-generating

activities, thus improving the well-being of the family (Canning & Schultz, 2012). FP has

been estimated to save billions of shillings, which can be availed for development

purposes, thus stirring economic growth and contributing to the improvement of the

general quality of life (Frost et al., 2014). FP saves money by preventing unintended

pregnancies and their adverse outcomes such as abortions and low birth weights and by

enabling women to be more productive. The government can then channel the saved
 6
funds into other development initiatives. Thus, improving the FP uptake in the

marginalized areas has the potential for positive social change. Health and economic

benefits can result in better educated and healthier children, leading to less dependency in

the society with overall enormous macroeconomic, demographic benefits (Canning &

Schultz, 2012).

Importance of this Study

The Kenyan constitution has recognized health, including reproductive health, as

a right (National Council for Law Reporting (2010). Given that FP is a cost-effective

public health strategy with many benefits, it is imperative to have all the eligible people

of Kenya access the service equitably. To achieve the sustainable development goal

benchmark of a CPR of at least 75% by 2030 in all countries, the majority of states,

including Kenya, need to accelerate the uptake of FP (Choi et al., 2015). Though the CPR

in Kenya has been increasing with the 2014 Kenya DHS reporting a CPR of 58% among

married women and 65% among sexually active unmarried women, it fell short of the

national target of 70% (Kenya National Bureau of Statistics, 2015).

Further, though there is information focusing on the general population in Kenya,

none focuses on adolescents and postpartum women. Analyzing the available data

concentrate on these groups is a cost-effective and efficient way of getting critical and

relevant information. Cost-effectiveness and efficiency are crucial given the limited

funding for health and particularly for public health globally (Shi, & Johnson, 2014). For

healthier lives, developing countries like Kenya need to develop homegrown solutions

that are in line with their realities (Agyepong et al., 2017). Local research is necessary to
 7
inform priorities and national strategies and provide evidence to support program

implementation. This study provides relevant information for the FP program that will

support increased FP uptake among the priority groups and thus harness the benefits

associated with women’s and girls’ use of contraception.

Potential Positive Social Change Implications

FP is a cost-effective public health intervention with many positive social change

implications. FP improves the health of women and their children and empowers

adolescent girls to avoid unintended pregnancy, thus enhancing their opportunity to

pursue education. Education promotes economic independency, which enhances general

quality of life (Canning & Schultz, 2012). Additionally, fertility decline reduces youth

dependency, further strengthening economic growth. Further, increased uptake of FP by

preventing unintended pregnancies and the consequences saves millions of dollars that

are made available for other social services and economic development, thus improving

the general quality of life for all and not just women (Canning & Schultz, 2012).

Problem Statement

Despite the knowledge that FP is one of the most cost-effective public health

strategies that can contribute to the improvement of maternal health, empower women

and girls and spur economic growth, there are many challenges associated with

developing successful FP programs (Stiegler & Susuman, 2016). Access to FP for all

women of reproductive age has been recognized internationally for several decades, as

evidenced by inclusion in various international commitments and initiatives such as the

International Conference on Population and Development Programme of Action, the

 8
millennium development goals, and most recently in the sustainable development goals

and the FP 2020 (Hardee, et al., 2014a; Hardee, Kumar, 2014b; Woog et al., 2015).

Despite progress, millions of women are still without access to voluntary FP (Kissoon et

al., 2015).

Many studies have been undertaken to study the various factors influencing FP

uptake in different geographical areas. For instance, a cross-sectional multi-country

analysis of DHS data for low- and middle-income countries revealed that different factors

influenced FP uptake differently. Community-level education attainment influenced FP

uptake positively, whereas gender and fertility-related norms influenced contraception

uptake negatively. Exposure to media did not have a positive influence on the FP uptake

(David & Allan, 2018; Mutumba et al., 2018). However, these results are not uniform

across all these countries, and there is a need to analyze specific country data for context-

specific results. For example, in Zambia, a qualitative study revealed that health system

factors such as long distances, stock-outs, and unfriendly policies negatively influenced

the uptake of FP (Silumbwe et al., 2018). Community-level barriers to uptake of FP

included myths and misconceptions about FP, side effects experienced by some

community members, social stigma, and harmful cultural and religious beliefs (Silumbwe

et al., 2018).

In Kenya, studies have been undertaken to determine the factors influencing FP

uptake in different areas. Limited geographical coverage and aggregate outcomes on FP

use have shown progress in the use of FP, but they tend to mask the diversity across the

country (Cahill et al., 2018). Amo-Adjei et al. (2017) did a multipronged intervention to
 9
increase FP uptake in two high fertility counties in Kenya. The interventions included

improving the health service delivery through training of health service providers and

supporting commodities availability. The initiative reported an increase in FP uptake in

the two counties with the total fertility rate dropping from 5.4 to 4.2 in Siaya and from

5.6 to 4.7 in Busia, while the unmet need for FP also dropped from 32 to 23% and 26 to

21% in Siaya and Busia counties respectively. Despite these improvements, the total

fertility rate and the unmet need are still high and need to improve further (Amo-Adjei et

al., 2017). Another study focused on the factors associated with FP uptake among

postpartum women in a county hospital. Sociodemographic factors such as age, marital

status, and being employed, as well as the quality of the services, were found to be

associated with uptake of FP among the postpartum women (Jalang’o et al., 2017). A

third of women who did not want more children were not on any contraceptive,

indicating a high unmet need for FP (Jalang’o et al., 2017).

As highlighted, various factors have been found to influence the uptake of FP.

However, these factors vary regionally, nationally, and even sub-nationally and affect the

different segments of the population differently, hence the importance of studying

specific population targets (Li et al., 2019; Stiegler & Susuman, 2016). The need for

studies focusing on particular contexts (priority populations, subnational, and community

level) has been recommended to inform FP policy and programming (Dennis et al., 2017;

Do & Hotchkiss, 2013; Patton et al., 2016). There is limited information on adolescents

and postpartum women and particularly for the marginalized areas in Kenya, where FP

uptake is the poorest. Thus, factors influencing FP uptake among adolescents and
 10
postpartum women in Kenya are not well understood, and it is not known how they differ

by region. The KDHS collects a lot of essential data on FP that is nationally

representative and is in line with national priorities and strategies; however, the analysis

is not detailed, particularly on the adolescents and postpartum women. Though these raw

data are available, the analysis is limited. This study utilized the available KDHS data to

provide nationally representative and specific information on the factors associated with

FP uptake among adolescents and postpartum women, to support the implementation of

the FP program for these priority groups.

Purpose of the Study

Given the need for focused FP data for effective program implementation to

enhance uptake of contraception, this study sought to examine the factors that contribute

to FP uptake among priority groups (adolescents and postpartum women) in Kenya. The

data were from the Kenya DHS (KNBS, 2015). Another purpose of this study was to

provide feedback that can inform subsequent KDHS questions in the future. It will also

act as a baseline for comparison with future surveys on the status of factors influencing

FP uptake across the counties for postpartum women and adolescents and youths.

The primary dependent variable was the uptake of FP, but other critical measures

of FP utilization such as unmet need and discontinuation were examined. The

independent variables included the personal, interpersonal, community, and

organizational factors. These factors have been described in relation to FP uptake and the

associations determined among adolescents and postpartum women. The differences by

regions were also a focus. The sociodemographic characteristics included education,

 11
wealth status, and residence (rural vs. urban), and intrapersonal characteristics included

exposure to media, and knowledge of FP (Do & Hotchkiss, 2013; Jalu et al., 2019;

Mutumba et al., 2018). Interpersonal factors included involvement of partners and

making decisions for the woman (Shahabuddin et al., 2019; Wegs et al., 2016).

Community factors included having heard FP messages from community leaders (Jalu et

al., 2019; Silumbwe et al., 2018; Wegs et al., 2016). Organizational factors included

mainly interaction with health workers (Kumar et al., 2020). Characteristics related to

utilization of maternal and child health services such as facility delivery, seeking

antenatal care (ANC) and postnatal care (PNC) were treated as covariates (David &

Allan, 2018; Do & Hotchkiss, 2013; Vogel et al., 2015).

Research Questions and Hypotheses

Secondary data analysis was done to answer the following research questions and

the associated hypotheses:

Research Question 1: To what extent are intrapersonal, interpersonal, community,

and organizational factors (independent variables) associated with the uptake of FP

(dependent variables) among adolescents in Kenya?

H01: Intrapersonal, interpersonal, community, and organizational factors are not

associated with the uptake of FP among adolescents in Kenya.

Ha1: Intrapersonal, interpersonal, community, and organizational factors are

associated with the uptake of FP among adolescents in Kenya.

 12
Research Question 2: To what extent are intrapersonal, interpersonal, community,

and organizational factors (independent variables) associated with the uptake of FP

(dependent variables) among postpartum women in Kenya?

H02: Intrapersonal, interpersonal, community, and organizational factors are not

associated with the uptake of family planning among postpartum women in Kenya.

Ha2: Intrapersonal, interpersonal, community, and organizational factors are

associated with the uptake of family planning among postpartum women in Kenya.

Research Question 3: To what extent are the intrapersonal, interpersonal,

community, and organizational factors (independent variables) associated with family

planning uptake (dependent variables) among postpartum women and adolescents differ

by counties in Kenya?

H03: Factors associated with family planning uptake among postpartum women

and adolescents in Kenya do not differ by counties.

Ha3: Factors associated with family planning uptake among postpartum women

and adolescents in Kenya differ by region.

Intrapersonal factors included information and knowledge on FP, exposure to

media, ability to make decisions, and perceptions on FP. Interpersonal factors included

partner and other family involvement in decision. Community factors included mean age

at marriage, mean age at first birth and mean age of first sexual intercourse, household

decision-making norms, and the ideal number of children perceptions. Organizational

factors included access and availability to the services (distance and cost) and counseling

on FP. Co-variables were also analyzed: age, education level, wealth status, area of
 13
residence, obstetric history (parity, seeking ANC and delivery services), and utilization of

other maternal and child health services such as PNC and immunizations. FP uptake was

evaluated in terms of those who use FP and those who do not use, in relation to their need

for birth spacing or limiting births and sexual activity. Discontinuation of methods was

described and related to independent variables.

Theoretical Foundation for the Study

Theory refers to the systematic explanations concerning a particular issue and

attempts to answer the why question, particularly as regards a behavior (Babbie, 2017).

Thus, theoretical frameworks or models direct research to focus on the relevant areas.

Theories are built through analysis of observations that focus on making sense of

occurrences, coming up with various propositions or hypotheses. These prepositions are

then used as bases of research questions, and from the research, the theory is further

tested (de Vaus, 2001).

This study used the SEM, which is based on the ecological model that was

proposed by Urie Bronfenbrenner (Bronfenbrenner, 1989). It was introduced in the 1970s

as a conceptual model for understanding human development and was further developed

into a theory (Bronfenbrenner, 1977). Bronfenbrenner (1977) advocated for a broader

approach in studying human beings that would consider their interaction with the

changing environment, both the physical and the social context. The model posits that

behavior is influenced by the interaction with the environment and implies reciprocal

causation. The environment was described as micro, meso, exo, and macro, and different

variations based on the model have been developed (McLeroy et al., 1988).
 14
The SEM used in this study has been developed to describe intrapersonal

(individual), interpersonal (family, friends), community, organizational, and policy level

influences on behavior (McLeroy et al., 1988; National Institute of Health, 2005). This

model is in line with the current thinking in health promotion that behavior change should

not only focus on the individual but also on the environment created by family, friends,

and the community as well as the organizational and policy interactions with the

individual (National Institute of Health, 2005). This ecological approach considers a

broad perspective in dealing with complex public health challenges. It has been adapted

by the Centers for Diseases and Prevention for various health promotion initiatives (Sallis

et al., 2008). The SEM has been used widely to study individuals in their ecological

contexts, including adolescents (Neal & Neal, 2013). The use of maternal and child

health services, including FP are influenced by the interaction of the individual with

others (family, friends, and community), health systems, and policies (Kissoon et al.,

2015). FP uptake is a complex health behavior that must, therefore, be approached from a

broad perspective rather than a simplistic view.

Other studies have successfully used the SEM to study the uptake of maternal

health services, including FP. For example, Shahabuddin et al. (2019) studied the

maternal healthcare-seeking behavior of married adolescent girls in Nepal. They found

that at the individual level, the perceptions, inadequate knowledge, low decision-making

autonomy, and dependency on the husband influenced seeking services. At the

interpersonal level, the partners and mother-in-laws made the decisions for many of the

girls. At the community level, certain traditional practices influenced the girls’ decisions
 15
while at the health system level, unfriendly services that were difficult to access

discouraged utilization of services (Shahabuddin et al., 2019). Additionally, Silumbwe et

al. (2018), in a qualitative study in Zambia, found that various community-level factors

and health system factors influenced the uptake of FP services. In Ethiopia, Jalu et al.

(2019) identified barriers to the uptake of maternal health services, including FP in the

intrapersonal, interpersonal, organizational, and the policy level. The majority of these

studies are qualitatively done in different parts of Asia and sub-Sahara Africa. They have

described different constructs in the SEM that are reported to influence the uptake of

maternal health services, including FP. These being qualitative studies, they are not

generalizable, and the association of FP uptake by different SEM level factors could not

be tested statistically. Hence these studies’ level of evidence is weak, but they generate

hypotheses for testing.

This study focused on intrapersonal, interpersonal, community, and organizational

factors, as outlined in the research questions. Figure 1 illustrates the different constructs

in each level that influence FP uptake either negatively or positively. Table 1

demonstrates how different variables were used to measure the SEM construct at

different levels.
 16
Figure 1

Conceptual Framework for Factors Associated with Family Planning Services Uptake

SEM Levels SEM constructs Influence on Family

Planning Uptake

Knowledge,
Individual perceptions, beliefs,
level and attitudes on family
planning.

Power relations that

Inter- influence decision
personal making e.g. Male
level partners and family
members dominance

FP Use

Social norms, Stigma,

Community status of women, and
level community
perceptions

Quality, accessibility,
affordability, and
Organizational availability of health
level services
Capacity of Health
workers to offer
services
 17
Table 1

Variables for Measuring Socioecological Model Constructs

SEM Levels SEM constructs Variables

Individual level Knowledge of FP Ever heard of a method
methods Heard FP messages from various
sources
Access to information Have access to newspaper, radio,
television
Access messages through a mobile
phone
community health workers
health care providers
Perceptions on family If given as a reason for not using FP
planning.
Beliefs, and attitudes If given as a reason for not using FP
Self-efficacy If given as a reason for not using FP

Interpersonal level Power relations If there was a discussion with the

partner
If the partner is aware of FP use
If given as a reason for not using FP
If the partner or another member of the
family makes the health care decisions
Community community level Having heard community leaders talk
fertility norms favorably on FP
community level
Organizational level Accessibility If given as a reason for not seeking FP
services
Quality If given as a reason for not seeking FP
services
Affordability If given as a reason for not using
Availability If given as a reason for not seeking FP
services
Interaction with health Discussed FP issues with health workers
workers Seeking of other related services
Capacity of health Provide asking clients FP needs
workers offer services
 18
The KDHS has quantitative information for each of the selected levels of the

SEM. Though not much information for the policy level is available from DHS questions,

for this level, some of the organizational factors such as the quality of services and cost

relate to policy. Hence, the policy affect FP uptake are discussed. Various research has

shown that different factors influence FP uptake; however, different studies have yielded

conflicting results on some factors such as age and knowledge of FP. Jalang’o et al.

(2017) found younger age to be positively associated with uptake of FP, whereas

Mutumba et al. (2018) found older age rather than younger to be positively associated

with FP use. Though Prata et al. (2016) and Ajero et al. (2016) reported an association

between knowledge and FP uptake, other studies have reported having knowledge and a

positive attitude did not translate into FP use among university students (Gbagbo &

Nkrumah, 2019). The extent of influence differs depending on the population and the

context. This study has added to the evidence on the association of different factors at a

different level of the SEM to the use of FP among adolescents and postpartum women.

Unless different levels are addressed, change in a complex behavior like FP uptake will

not be successful. Knowledge of the status of these factors in different regions will

contribute to the determination and implementation of appropriate measures for FP

uptake (Patton et al., 2016).

Nature of the Study

This was a cross-sectional quantitative survey. The study sought not only to

describe the factors that are associated with FP uptake but also to determine the extent to

which these factors are significant. Quantitative data are amenable to statistical
 19
manipulation of the different variables (Creswell & Creswell, 2017). Following statistical

methods, it is possible to determine if the differences in the outcome between the two

groups are by chance or are due to the factors being studied. Various statistical tests, such

as regression analysis, are available for manipulating quantitative data. Common

assumptions to ensure sound results include a large random sample that is representative

of the population for inference to be made.

Data were analyzed from the nationally representative KDHS collected in 2014.

The 2014 data set was selected since the next KDHS is yet to be conducted. The DHSs

are nationally representative population-based household studies conducted in many

countries. They have been conducted since 1984, primarily funded by the U.S. Agency

for International Development, and individual countries receive technical support from

the DHS program (International Household Survey Network, 2019). So far, more than

400 surveys in over 90 countries have been done and disseminated (ICF International.,

n.d.). The self-reported questionnaire targets women of reproductive age, 15-49 years of

age, living in the sampled households and collects a wide range of questions related to

health, including fertility, FP, maternal and child health, among others. The

questionnaires include sociodemographic such as education level, wealth status, access to

various amenities such as water, and areas of residence. Reproductive health data include

knowledge on FP methods, altitude, practice, and service utilization, including related

factors. Given this range of information, the various variables of interest were identified

and recoded as necessary. The KDHS data were collected by the KNBS with technical
 20
assistance from partners such as the ICF Micro, who have experience in conducting

similar surveys across many countries (KNBS, 2015).

The variables addressed in this study included sociodemographic characteristics

(education attainment, wealth status, literacy level, marital status and areas of residence),

intrapersonal factors (exposure to media and knowledge of FP knowledge on FP),

community factors (leaders talking positively on FP), and organizational level variables

included and the utilization of maternal and child health services as covariates. The study

describes the different adolescent and postpartum women characteristics and the

association to FP use. The factors that were found to be significantly associated with

uptake of FP were analyzed further to determine if they differed by region.

Literature Search Strategy

The Walden library was used to search various databases. ProQuest Nursing &

Allied Health Source, ProQuest Health & Medical Collection, and CINAHL &

MEDLINE, Google, and Google Scholar were used to search general papers from

professional organizations such as the world health organization and the U.S. Department

of Health Services. Keywords and combined word searches were done informed by the

research questions, the population, geographical area, key-dependent variables, and key

independent variables, and the database. Keywords searched included:

 The primary dependent variable family planning and family planning uptake

 To get insight into the social problem and benefits of FP, maternal mortality

and family planning, family planning and development, unmet need for FP
 21
 The population (priority groups) adolescent or adolescence or youth, and FP,

postpartum FP

 Geographical area: Kenya and Sub-Saharan Africa

 Independent variables: Family planning uptake determinants, and community-

based workers (factors associated with the dependent variable),

 The database demographic and health surveys

 Other concepts: FP quality of services, Rights-based FP, method mix

 Theoretical framework: socio-ecological

Searches mainly focused on studies published no later than 5 years, but older

studies were used in case more recent studies were limited. For example, seminal articles

on the SEM, the theoretical model used for this study, and articles on the historic

progress on the family program in Kenya necessitated the inclusion of older articles.

Peer-reviewed articles were preferred, but publications by organizations such as the

WHO, the U.S. Department of Health Services, and the government of Kenya were also

used. Primary searches were done using the Walden Library, and reference lists from the

selected articles were reviewed to identify any other relevant articles that had not have

been identified through the initial search.

Literature Review

Family Planning Methods

Unwanted pregnancies are associated with many poor and sometimes terrible

health and social outcomes as well as economic losses. FP is the primary public health

strategy to prevent unwanted pregnancies (Ahmed et al., 2012; Ganatra & Faundes,
 22
2016). According to the WHO (2018), various methods can be used for planning

pregnancy to achieve desired reproductive goals. Though the term FP is often used to

refer to preventing pregnancy, it does include fertility treatment to enhance conception.

According to the WHO, the following FP methods are available for use: combined oral

contraceptives, progestogen-only pills or “the minipill,” implants, progestogen-only

injectables, monthly injectables or combined injectable contraceptives, combined

contraceptive patch and combined contraceptive vaginal ring, intrauterine device (IUD):

copper-containing or levonorgestrel, male condoms, female condoms, male sterilization

(vasectomy), female sterilization (tubal ligation), lactational amenorrhea method (LAM),

emergency contraception pills (ulipristal acetate 30 mg or levonorgestrel 1.5 mg),

standard days method, basal body temperature method, two day method, sympto-thermal

method, calendar method or rhythm method, and the withdrawal (coitus interruptus)

method.

Different FP methods have different levels of effectiveness based on their use

(WHO, 2018). The combined oral contraceptives contains estrogen and progestogen, and

even though with correct use it is > 99% effective, the effectiveness drops to 92% with

the ordinary use. The progestogen-only pills can be used by breastfeeding women and is

up to 99% effective with correct use but with typical use is 90–97% effective. Implants

consist of progesterone and are about 99%. The injectables are 99% effective with correct

use and 97% effective with ordinary use. The contraceptive patch and the contraceptive

ring are relatively new, and research on the effectiveness is ongoing. The IUDs are up to

99% effective. Male condoms are up to 98% effective with consistent and correct use but
 23
on average are 85% effective as commonly used. The female condom is 90% effective

with correct use but drops to 79% with the common use. Vasectomy is more than 99%

effective after 3 months, and the tubal ligation is more than 99% effective as well. LAM

is as high as 99% effective with correct use, but it is a temporary FP method that utilizes

the natural effect of breastfeeding on fertility. Emergency contraception pills are 98%

effective when used correctly. The standard days method is a fertility awareness method

that utilizes cycle beads or other aids and is up to 95% effective when used correctly and

88% effective as commonly used. Basal body temperature method is also fertility based

that utilizes changes in body temperature and is up to 99% effective when used correctly

and consistently, but effectiveness reduces to 75% with common use. The 2-day method

is a fertility awareness method based on cervical mucous. It is 96% with correct and

consistent use and 86% with common use. The sympto-thermal method is a fertility

awareness method that is based on cervical mucous and body temperature. It is 98%

effective with correct use. The calendar method or rhythm method is a fertility awareness

method that utilizes the pattern of the menstrual cycle. It is up to 91% effective with

correct use, and effectiveness reduces up to 75% with common use. The withdrawal

method entails the man trying to keep sperm out of the vagina to prevent pregnancy. It is

up to 96% effective when used correctly and consistently, and it is 73% effective with

typical use (WHO, 2018).

FP methods can also be classified into traditional methods and modern methods

(WHO, 2018). Natural methods are based on abstaining from sex to avoid pregnancy

(Pallone & Bergus, 2009). Most organizations list the methods either as traditional or as
 24
modern without stating the criterion (Hubacher & Trussell, 2015). The methods that use

hormones or devices to control fertility seem to be universally accepted as modern

methods. However, the fertility awareness-based methods have raised controversy, with

some being classified as modern methods and others as traditional without clear

guidelines (Hubacher & Trussell, 2015; United Nations et al., 2013). Some methods use

technology to identify the fertile days to support abstinence, thus contributing to the

controversy. The WHO (2018) listed fertility awareness methods such as the sympto-

thermal method, 2-day method, basal body temperature method, and standard days as

modern methods, and they classified calendar method or rhythm method and withdrawal

(coitus interruptus) as traditional methods. Additionally, the United Nation Population

Fund (UNFPA), the Guttmacher Institute, and the DHS Program all named the LAM as a

modern method (International Household Survey Network, 2019; Singh, Darroch &

Ashford, 2014). Traditional methods are generally not as effective as modern methods in

preventing pregnancy (Pallone & Bergus, 2009).

The DHS collects information on the most common methods in a country and on

both natural and modern methods of contraception and thus provides relevant information

on the different methods that are used in a population (DHS Program, 2019). LAM is of

particular interest since it is freely available, does not need to be provided by a health

worker, and it is effective and with no side effects. All the women need is to have proper

knowledge of how to use it. It is considered a modern contraceptive method for

postpartum mothers who meet the criteria of its use (International Household Survey

Network, 2019; Singh et al., 2014). When used correctly, LAM is up to 98% effective.
 25
Three sets of criteria must be set to ensure protection: mother must have amenorrhea, be

fully or almost fully breastfeeding, and the infant must be less than 6 months. However,

there is concern that many postpartum mothers do not have the right knowledge on how

to apply this freely available method and hence miss out on the potential benefits. Fabic

and Choi (2013) found that only 26% of LAM users met the criteria for correct and valid

LAM.

Method Mix

In considering the uptake of FP, the method mix significant, as it reflects on the

right of women to make an informed choice (Bertrand et al., 2014). Method mix refers to

the distribution of the FP methods by users, and reliance on one method by a vast

proportion of the users is considered skewed. Access to diverse methods enhances the

choice a woman has and is an essential component of voluntary contraception, which is a

right (Hardee et al., 2014b). Skewed method mix may indicate inadequate access to other

methods either due to limitations in supplies and health worker’s skills, provider bias, or

even community bias (Bertrand et al., 2014). The most used method in Kenya is the

injectable among youths and adolescents as well as the older women (Dennis et al.,

2017). The longer-acting methods, such as the intrauterine copper device and the

implants, are more cost-effective and are associated with less discontinuation compared

to the short-acting methods (Benson et al., 2017; Keesara et al., 2018; Ochako et al.,

2015). Besides the user characteristics, the technology, including the logistics associated

with the use of the longer-acting methods, contribute to the reduced likelihood of

discontinuation (Hubacher et al., 2017). It would, therefore, be more desirable to have

 26
women use the longer-acting methods. Organizational factors like social franchising have

provided more access to FP services, and more women took up the more cost-effective,

longer acting, and permanent methods (Chakraborty et al., 2016). Further, in various

urban cities in Kenya, a project designed to increase FP uptake increased the overall use

of FP, and the use of LAM increased considerably (Benson et al., 2017). There were

tailored messages and improved access to a wide range of methods during the project,

which positively influenced the uptake (Benson et al., 2017). However, information on

the method mix among adolescents and postpartum women are limited in the Kenyan

context, though the DHS has information on the methods being used and the reasons.

This study will describe the method mix among adolescents and postpartum

women and relate this to different factors to determine which ones influence the use of

particular methods. The information on whether the method mix is appropriate among the

adolescents and postpartum women is critical for the FP program. Such information will

be useful to enhance the effectiveness of reaching these groups. It will also inform the

policymakers and program managers if changes are needed, for example, in terms of

access to the different methods to support a better method mix. A more comprehensive

range of method mix increases the use of FP as well as enhances the benefits by having

clients use the methods that best suit them as well as the more cost-effective methods

(Jain & Winfrey, 2017).

Unmet Need for Family Planning

Unmet need for FP has been significantly associated with the total number of

pregnancies, the number of children alive, approval of contraception by the partner, and
 27
discussion of FP within the couple (Ajong et al., 2016). Unmet need for FP is an indicator

in assessing the performance of FP programs (Vogel et al., 2015). Reducing the unmet

need for FP is one of the cost-effective public health strategies in low- and middle-

income countries with significant benefits for both mothers and infants, reducing demand

for abortion as well as vertical transmission of HIV (Zakiyah, et al., 2016). Investing in

access to modern contraception, reducing unmet need for contraception is more cost-

effective than retaining the status quo of limited access (Zakiyah, et al., 2016). Evidence

is needed to support favorable decisions to support FP uptake at all levels to reduce the

unmet need. FP use is related to the unmet need. By describing the uptake of FP among

adolescents and postpartum women, this study will provide information that could be

helpful in reducing unmet need in these groups.

Adolescents

Adolescents form a significant and growing proportion of the population in Sub-

Saharan Africa and Kenya, yet their health needs have not received adequate attention. It

is important to curb adolescent pregnancies to contribute to ending preventable maternal

deaths. Various factors such as physical immaturity, low socioeconomic status, and

sociocultural norms and practices make adolescent pregnancies in the developing world

riskier with poor outcomes. The majority of these pregnancies are unintended; some

result from early and forced marriages with unmet need for contraception (Vogel et al.,

2015).

Different researchers have studied the use of FP among youths and adolescents

and have documented different results. Gbagbo and Nkrumah (2019) did a study among
 28
young unmarried women in a tertiary institution in Ghana and established that the

students had knowledge of FP and a positive attitude; however, the knowledge and a

positive attitude did not translate into FP use because of availability and accessibility.

The emergency contraceptive was the most used since it was widely available without the

need to go to a health facility that the they would rather avoid due to perceived stigma

(Gbagbo & Nkrumah, 2019). These results suggest that there may need to restructure the

information content on FP by FP programs targeting youth. Additionally, these results

point to the need to target more than one level of the SEM in interventions to increase FP

uptake.

In another study on FP, Shahabuddin et al. (2019) explored married adolescents’

maternal health care services, including FP seeking behavior in Nepal. They interviewed

community health workers, family members, and government officials. They used the

SEM to analyze and report the findings. Intrapersonal factors such as knowledge on the

services, dependency on partners, and low autonomy in decision-making influenced their

use of FP. At the interpersonal level, the mothers-in-law, partners, and other family

members influenced whether these adolescents used FP. At the organizational level,

unfriendly, inaccessible services, and inflexible operating hours negatively influenced the

uptake of services. In contrast, supportive community environment such as the

availability of female community workers and women groups through which information

is shared positively influence the uptake of services (Shahabuddin et al., 2019). The

findings may not be generalized to Kenya due to differences in the context; however, the
 29
study provided information on variables that can be used to determine the significance of

these factors within the specific context.

Finally, Woog et al. (2015) analyzed adolescent women’s need for and use of

adolescent health services, including FP in developing countries such as Kenya. They

used the KDHS 2008–2009. The adolescents reported not using contraception due to

intrapersonal factors, not being married, and infrequent sex. However, sexually active

adolescents are at the risk of unwanted pregnancy despite the frequency of sex and

marital status. The adolescents also reported organizational factors, lack of access, health

concerns, and fear of side effects as reasons for the non-use of FP (Woog et al., 2015).

The KDHS 2008–2009 is a nationally representative survey in Kenya; the more recent

KDHS 2014 was used for this study to provide detailed information on the use of

contraceptives by adolescents in Kenya.

Postpartum Women and Postpartum Family Planning

The postpartum period is the time soon after delivery up to 6 weeks after, but for

contraception, up to 1 year is considered (WHO, 2013). It is a critical period because if

not protected, a sexually active woman is at the risk of pregnancy, leading to too close

spacing of pregnancy (Rossier et al., 2015). In a multi-country prospective study to assess

fertility intentions, contraceptive use, and unmet need for FP among postpartum women 6

weeks post-delivery, Pasha et al. (2015) found that of the 36,687 women in the study,

only 5% wanted to have a pregnancy within the first year after delivery. Despite the

majority not desiring to get pregnant, there was a huge unmet need for FP ranging from

25% to 96%. Of those using modern methods, only a small proportion was on the more
 30
effective long-acting reversible contraceptives. In this study, factors associated with high

unmet included; young age of fewer than 20 years (adolescents), low parity of two or

three, low education level, and women delivered at home (Pasha et al., 2015). Pasha et al.

(2015) analyzed several countries’ data; however, for more specific national and sub-

national data, secondary analysis of the national data is critical. This research provided

specific information on the factors associated with FP uptake among the Kenyan

postpartum women.

Rossier et al. (2015) analyzed data from 5732 women, some being from sub-

Saharan Africa. They found that 43 percent of the women had an unmet need for FP at

six weeks postpartum, while 32 percent had an unmet need at the end of amenorrhea

during the first year of delivery. There is the challenge of women relying on LAM even

when they are not protected (have not met the criteria) (Rossier et al., 2015). Though

DHS are nationally representative, the results of this study are according to regions rather

than countries. For specific information at the national and subnational levels, it is crucial

to analyze country-specific data.

An analysis of DHS data from 21 low and middle-income countries, including

Kenya, revealed high unmet need among postpartum, with more than half of the repeat

births being within an interval that was too short (Moore et al., 2015). Though the DHS

data is nationally representative, specific, and more recent and detailed information for

Kenya was obtained by analyzing the latest Kenyan DHS. Another study in Northern

Tanzania revealed that though only 11% of postpartum women wanted pregnancy within

two years of delivery, 56% were at risk of pregnancy, and 36% of the new pregnancies
 31
during the follow-up period were unwanted (Keogh et al., 2015). The study involved the

intervention of targeted counseling during the ANC and in the postpartum period.

Therefore, the results may not be generalized to a population that receives routine

services (ANC and PNC) and may not be applicable in another country.

Muamah et al. (2015) found that 15.9% of women were not protected against

pregnancy by the end of their first postpartum year. Though most postpartum women

wanted to space pregnancy by at least two years, most were using short-acting methods.

In another study, as many as 80% of the postpartum women used the less cost-effective

short-term FP methods (Moore et al., 2015). Analysis of the 2014 DHS has provided

more evidence.

The concept of unwanted pregnancy is not appreciated among married women as

it is seen as applying more to unmarried adolescents and young women (Capurchande et

al., 2017). Hence, married women may not appreciate spacing given the many

misconceptions around contraceptives and fear of side effects making some women feel

like the risks of contraception outweigh the benefits. Factors Influencing Family Planning

Uptake (Independent Variables)

Many studies have looked at the factors that influence the uptake of FP,

particularly in the developing world, where CPR remains low with high unmet needs.

Most of these studies are not nationally representative for Kenya, but they provide useful

insights on what to study for context-specific information.

 32
Intrapersonal (Individual) Factors

Intrapersonal or individual factors, as described in the SEM, influence FP uptake.

Intrapersonal factors are the intrinsic individual characteristics that influence behavior

(National Institute of Health, 2005). They include characteristics such as knowledge,

attitudes, beliefs, values, and personality traits.

Fertility and how to control it either using natural or modern methods is a

technical subject that requires correct information and skills. The subject is made more

complicated by myths, misconceptions, and wrong information (Silumbwe et al., 2018).

Like many other modern drugs, modern contraceptives have documented side effects

such as irregular bleeding and nausea, and the clients need correct information.

Information on the potential side effects and how to address them is critical. The risk of

side effects is low for modern methods and, in most cases, minor. Unfortunately,

misunderstanding the small risk and exaggeration has led to unwarranted fear of modern

methods. Women may recognize the need for spacing pregnancies, but fear of real side

effects and misconceptions greatly influences contraceptive use. Fear of side-effects is

associated with contraception non-use (Ajong et al., 2016; Nanvubya et al., 2015).

Misconceptions such as contraceptives leading to malformed babies, IUD penetrates the

body, infertility, paralysis, and need for hysterectomy are reasons for non-use of FP

(Keesara et al., 2018; Ochako et al., 2015). Some fear that though they would be willing

to use contraceptives against their partners' approval, side effects such as irregular and

excessive bleeding and loss of libido would expose them.

 33
Prata et al. (2016) conducted a household survey to document factors associated

with modern FP methods among Angolan women of reproductive age. The analysis was

stratified between adolescents and youths (15-24 years) and older women (25-49 years)

to determine the effect of age with different variables. Among all ages, the intrapersonal

factors associated with FP uptake were contraceptive accessibility perception,

contraceptive knowledge, and self-efficacy. Being exposed to FP information from

various sources such as media self-efficacy of FP and marital status was positively

associated with contraceptive use among adolescents and youths. Among the older

women receiving information about FP from a pharmacy was associated with

contraceptive use (Prata et al., 2016).

Despite increasing reports of growing widespread FP knowledge in Kenya and

other countries, some areas, particularly in the rural setting, where correct knowledge is

low (Jalu et al., 2019; Mutombo et al., 2014). In a cross-sectional qualitative study done

in the Somali region in Ethiopia, Jalu et al. (2019) explored the factors affecting health-

seeking behavior. They interviewed women of reproductive age and their partners, health

extension workers (HEWs), health care providers, and health administrators. They found

that there was limited knowledge of modern methods. The limited knowledge of

contraception and fear of modern health practices negatively influenced contraceptives

use (Jalu et al., 2019). In Nepal, Shahabuddin et al. (2019) found that women, particularly

adolescents, had little knowledge of FP, which negatively influenced FP's uptake.
 34
Additionally, though knowledge is critical in the use of FP, studies have

demonstrated a discrepancy between knowledge of FP and use. Despite immense

knowledge of FP or the need to avoid pregnancies, clients' practices on use do not march

the knowledge. In a qualitative study in Mozambique by Capurchande et al. (2017), they

found very high levels of knowledge of FP, but this did not translate to the use of FP.

Thus, knowledge alone is not enough to change behavior.

Other individual factors that negatively influence FP uptake include perceptions

of FP and fertility. In a qualitative study done in Uganda by Nanvubya et al. (2015),

women's desire for more children negatively influenced contraception. Some women

consider it their religious requirement to follow the husband's decision on whether or not

to use FP (Jalu et al., 2019).

Besides knowledge and perceptions of FP, Women's agency and self-efficacy in

deciding whether to use contraception or not is critical and relates to a woman's

empowerment (Prata et al., 2016; Wegs et al., 2016). Empowerment, as shown by access

to money, freedom of movement to seek services, and making fertility decisions, is a

critical personal factor that positively influences FP uptake (Reed et al., 2016). Being

able to make household decisions is part of women's empowerment. Those women who

make general household decisions are more likely to use modern contraceptives

(OlaOlorun & Hindin, 2014). These studies are not specific to Kenya; this study has

looked at intrapersonal factors are proxies to women empowerment and the association

with using FP in Kenya.

 35
Self-efficacy is an important aspect that influences the use of FP. It encompasses

the capability to use contraceptives correctly and consistently and the ability to negotiate

with a partner. The ability to discuss with partners, tell the husbands they want to use FP,

use FP, and use it even without the partner's approval is an integral part of empowerment

(Wegs et al., 2016). Having controlling husbands has been associated with an unmet need

for FP (Meiksin et al., 2015). Women who are empowered to make decisions concerning

FP use are more likely to utilize contraceptives (Belay et al., 2016). Dependency on

husbands and other family members, such as the mothers-in-law and limited decision-

making autonomy, negatively influences services' uptake (Shahabuddin et al., 2019).

Various factors (sociodemographic characteristics) such as women's educational and

employment status, occupation, and partner's educational status are associated with

women empowerment for FP's decision making power. Employed women and those with

higher education status are more likely to have higher decision-making power concerning

their fertility, either individually or together with their partners, than the unemployed and

those with less education (Belay et al., 2016). Women who can communicate with their

partners on FP and those with control over earnings are more likely to use contraception

(Wegs et al., 2016). Most of these studies exploring the individual factors, as noted, are

qualitative, have been done in other countries, or have covered a small geographical area

in Kenya, mainly focusing on the general population. This study used nationally

representative data to focus on adolescents and postpartum women. The KDHS has data

on why women do not use contraceptives and sociodemographic characteristics. This

 36
study described the sociodemographic characteristics of adolescents and postpartum

women and determined the association to FP use.

Interpersonal Factors

The SEM's interpersonal level focuses on the individual interaction with the

primary groups, including family, friends, and peers who are part of society and influence

their decision-making (National Institute of Health, 2005). These interactions provide

social support, identity, and role definition, which in turn influence behavior. On decision

making for FP, studies have shown that many women depend on their partners' approval

to use FP. In a study among the fishing communities in Uganda, Nanvubya et al. (2015)

found that women depended on their husbands' approval to use FP. Harrington et al.

(2016), in a qualitative study in Western Kenya, found that gender roles were involved

with many men viewing FP as the responsibility of the women since they carry the

pregnancy and take care of the children. However, the ultimate decision making still

rested with the men despite their little interest in the subject. Even when men want to use

contraception, they tend to rely on women. Ochako et al. (2015) did a secondary analysis

of the KDHS to document contraception determinants among sexually active men in

Kenya. This analysis revealed that men mainly rely on their partners for contraception

use. Factors found to influence their contraception use included; the area of residence,

marital status, religion, wealth, health care provider interaction, fertility preference,

number of sexual partners, and access to media (Ochako et al., 2015).

 37
Nevertheless, in most cases, for women to use contraception, they need the partners'

approval. In a study in Ethiopia, Jalu (et al., 2016) found male dominance in decision

making, particularly the husband's influence, as barriers to FP use.

Withers et al. (2015) did a qualitative study in the Kenyan Nyanza region. They

found that as gender roles and relations change, some men are reluctant to support FP.

They fear that FP enhances female sexual agency and promiscuity and further weakens

the male's power and role in society (Withers et al., 2015). On the one hand, some

husbands perceive FP as a woman business and do not consider it their business. On the

other hand, as a study in Nigeria revealed, some husbands fear that women who use FP

may become promiscuous, which negatively affects the FP demand (Adanikin, McGrath,

& McGrath, 2017).

Approval of contraception by the partner and discussion of FP within the couple

increases the likelihood of uptake of FP (Ajong et al., 2016; Prata et al., 2016).

Communication with a partner about FP positively influenced FP's uptake (Prata et al.,

2016). There is a significant and positive association of the utilization of postpartum FP

services husband's approval of contraception (Bwazi et al., 2014). However, some

religious beliefs discourage couples from discussing and deciding the number of children

they would want to have (Jalu et al., 2019). They consider getting children as from God,

and hence there should be no interference.

 38
Besides the partners, friends influence the use of FP. Women may be more

comfortable sharing their contraceptive issues with friends and trusted women rather than

health workers. Depending on their knowledge and perceptions of FP, the friends may

discourage FP's use, as documented in a qualitative study in Mozambique (Capurchande

et al., 2017).

The majority of these studies are qualitative and therefore documented concepts

of the interpersonal level factors influencing FP uptake that need testing in quantitative

studies. The quantitative studies were done either in other countries or in a small

geographical area in Kenya, focusing on adolescents and postpartum women. It is,

therefore, vital to get nationally specific data for the priority groups in Kenya. This study

has provided evidence-based on quantitative data that is representative nationally.

Therefore, the evidence from this study will be more reliable to support national FP

programming than that from either qualitative studies or from studies that covered a

limited geographical space.

Community Factors

According to the SEM community, factors including the social networks, norms,

or standards, which exist either as formal or informal, among individuals, groups, and

organizations influence behavior (National Institute of Health, 2005). Community factors

such as disapproval of FP use by the community members may discourage women from

using contraception (Wegs et al., 2016). Withers et al. explored men's perspectives of

gender roles and cultural norms about FP use. They found that misconceptions on the

side effects hinder men from supporting FP's use by their partners (2015).
 39
In a study done in Kenyan and Nigerian cities, many misconceptions at the

community level associated contraception with woman's potential ill health and loss of

fertility, and they negatively influenced the uptake of FP (Gueye et al., 2015). Withers et

al. 2015 found that even men who expressed their support for FP in their study spent

more time discussing FP's negative aspects rather than the positive effects. Thus it is

possible for men to verbally express support but not necessarily have enough confidence

in contraception to support their spouses. A qualitative study in Togo suggested that

sharing information with men on FP benefits such as financial gains and dispelling myths

on the side effects could positively influence contraception support (Koffi et al., 2018).

Withers et al. 2015, found that FP dialogues at the community level to address the

common myths and misconceptions and the adverse effects of the gender roles imbalance

may positively shift norms to support the use of contraception in a community (Withers

et al., 2015). An intervention in Papua New Guinea showed that by engaging the

community's gatekeepers and reaching the members, particularly the men, with

information on the benefits of birth spacing, the values and norms might transform to

prioritize FP (David & Allan, 2018). In engaging the community, it is critical to consider

their values and beliefs. For example, where they value children, highlighting FP's health

benefits to the mothers and their children enhances positive attitudes towards

contraception.
 40
Silumbwe et al. (2018) conducted a qualitative study that explored the factors that

facilitated and those that hindered FP services uptake at community and health systems

levels. The community-level factors that influenced FP uptake negatively included

experience with side effects, rumors, myths, misconceptions, community stigma, and

negative traditional and religious beliefs. However, community characteristics such as

functional structures, desire to delay pregnancy, and knowledge on contraceptives

supported FP use.

Mutumba et al. 2018, in a multicounty analysis of DHSs, found that the social

context, including socio-cultural and gender community, norms significantly shape

decision making on FP. Communities with a higher average age at first marriage, those

with higher education where women have greater autonomy of household decision

making, and wealthier communities had higher use of modern contraceptives (Mutumba

et al., 2018). Jalu et al. 2019, in a qualitative study in Ethiopia, found that pressure from

social norms and expectations continue to influence the desired family size, with

members preferring larger family sizes to conform. In another qualitative study in

Mozambique, Capurchande et al. 2017, found that traditional expectations of a large

family, which a source of pride and a guarantee of future sustenance, encouraged high

fertility (Capurchande et al., 2017). However, the changing economic times have caused

some conflict among some members who feel many children are difficult to sustain.
 41
Society's knowledge and understanding of FP influence their perception and use

of FP. Those who report having heard of FP do not necessarily have adequate knowledge

to support utilization. Information is understood through the social context, including

gender roles. Where a community has functional structures, they have been used to

positively reach the members with information to change their views on contraception

positively, thus promoting the use of contraception (Amo-Adjei et al., 2017; Silumbwe et

al., 2018).

In Capurchande's study, gender role perceptions influenced FP-related behavior

(Capurchande et al., 2017). Contraception was widely a female's responsibility, with

many men not attending counseling sessions despite being decision-makers. The females

had more knowledge of FP compared to men. The men with more information tended to

support the use of contraceptives by their partners. Unfortunately, men did not

accompany their women to the counseling sessions. FP has not primarily been their

business, and facilities have not had the infrastructure to accommodate them. When men

attempt to accompany their women, they may experience teasing from their peers with

other women stigmatizing them. Their involvement thus contradicts their gender roles.

Thus men were not involved in FP issues, yet their perceptions were critical in FP uptake

as decision-makers.
 42
Speizer et al. (2018) using data from two cross-sectional data from four urban

centers in Senegal, noted that when men were exposed to FP messages. For example,

through mass media and community outreach activities, they are more likely to discuss

FP with their partners and to support modern contraception use (Speizer et al., 2018). An

individual living in a tightly neat community rural setting is more likely to be influenced

by existing negative attitudes and perceptions of the community than a woman living a

more solitary life in the urban setting. In a mixed-method approach, Wegs et al. found

that community perceptions influence FP uptake (2016). Community discussions on FP

facilitated the shift of norms to make FP more acceptable in a community.

FP is generally looked at as for the married. In southwest Nigeria, a mixed-

methods study by Sieverding et al. (2018) revealed that unmarried adolescents face social

stigma due to the community's view of sexuality. They faced discrimination at the service

level due to provider bias. Many health workers were reluctant to provide unmarried

adolescents with longer-acting methods, and some dissuade the adolescents from using

FP altogether. These are community norms that influence the provision of services at the

organizational level since many health providers are members of the community and

share their norms and beliefs.

 43
Though the above studies provide some information on the community-level

factors that influence FP uptake, many are qualitative; hence there is a need for

quantitative studies for more robust evidence. Most of the studies are over small

geographical areas and mainly outside Kenya. A nationally representative quantitative

study is therefore critical to establish the status in Kenya. This study has provided

nationally representative information on community factors associated with FP uptake.

Organizational Factors

Though personal and community factors are critical in utilizing FP, women may

fail to use the service due to various organizational factors such as access and availability

of the services (Prata et al., 2016). In a qualitative study, Jalu et al. found that one's

residence determines services' utilization (Jalu et al., 2019). Challenges include physical

inaccessibility. The rural setting services tend to be far in many rural areas and

inconveniently located with low and costly transport compared to the urban setting

(David & Allan, 2018; Silumbwe et al., 2018). In some cases, the facilities exist, but the

commodities and supplies to support service provision are inadequate, and the opening

hours not flexible to suit the needs of different clients (Benson et al., 2017). Silumbwe et

al., 2018, found that long physical distances to the facilities, stock out of the method of

choice, and poor provider attitudes contributed to low utilization.

Besides geographical access, financial costs can be a barrier, particularly for long-

term and permanent methods. Where programs have been implemented to provide

financial support for FP methods, the uptake of these methods increased significantly

compared to areas with no support (Oyugi et al., 2017).

 44
The quality of services influences the uptake of FP services, as noted by Woog et

al., in a secondary analysis of national quantitative surveys from 70 developing countries

(Woog et al., 2015). Health workers' ability to communicate with clients may influence

the FP understanding and use by the clients. Health workers should be able to offer

culturally relevant counseling. Their ability to serve the different categories of clients

(adolescents, postpartum mothers, and other women) appropriately is also critical. In a

qualitative study, realities such as the challenging economic times and the effects of large

families on the environment when well-articulated influenced men's fertility desires

(Kock & Prost, 2017). Balanced discussions on contraception are therefore critical in

encouraging FP use. When health workers emphasize the need for a health service and

offer supportive follow-up, they are encouraged to seek services. A program in Embu that

entailed close postpartum follow-up led to an increase of FP uptake from 6% to 56%,

thus demonstrating the importance of follow-up (Warren et al., 2010). Bwazi et al., in a

hospital in Malawi, found a significant and positive association of utilization of

postpartum FP services with the provision of clear FP information (2014).

In some studies, in different parts of Africa, clients admitted having limited

discussions with the health workers and their spouses (Ajong et al., 2016; Prata et al.,

2016). Some clients felt that despite spending much time waiting to see the health

workers, they had limited interaction with them, and they were not able to ask many

questions. Thus the quality of counseling may influence the knowledge levels and

adequacy for clients to make decisions. Therefore, myths and misconceptions about FP

continue despite interactions with health workers.

 45
When women find contraception challenging to obtain, they are unlikely to use it

(Keogh et al., 2015). Integration of contraceptive services, provision of couples

counseling, and availability of skilled health care workers to provide a wide range of

methods contribute positively to the uptake of services (Amo-Adjei et al., 2017;

Silumbwe et al., 2018). Jalang'o et al. 2017 reported that ready access to contraceptives at

a health facility and favorable perception of the services encourages FP uptake (Jalang'o

et al., 2017). The approach to offering FP services can influence the uptake of particular

methods, affecting the method mix. Social franchising in Kenya increased the likelihood

of using LAPM, which is considered more cost-effective (Chakraborty et al., 2016).

Focused programs that combine different approaches such as dissemination of

information to the community and improved availability and accessibility of different

methods seem to increase the overall uptake of FP and the method mix (Benson et al.,

2017).

Though these studies provide useful insights on FP uptake issues across many

developing countries, mainly in Africa, only a few are specific to Kenya. However, even

those done in Kenya are not nationally representative and do not focus on adolescents and

postpartum women. Hence the need for more a focused study; this study focused on

determining the association between organizational factors and FP uptake in Kenya.

Covariates

Covariates are those factors that are an individual’s characteristics that are not

captured in the research questions as either dependent or independent variables but may

have an influence on the outcome of interest. Independent variables in this study are the
 46
various influence levels in SEM. Socio-demographic factors may influence FP uptake

through independent variables, hence the need to measure their potential contribution.

For example, the level of education may influence FP's uptake because an educated

woman can understand the FP messages more than a non-educated one. A higher

economic status may mean that the woman can afford the contraceptives. Various socio-

demographic characteristics have varying influences on FP uptake. Ochako et al. 2017,

found the area of residence, marital status, religion, and wealth status to influence

contraception use (Ochako et al., 2017). Jalang'o et al. 2017, found younger age, being

married, higher education level, and being employed as being associated with FP uptake

(Jalang'o et al., 2017). However, in some studies, older age rather than younger was

found to positively influence FP uptake (Mutumba et al., 2018). The realization that age

is a critical factor in FP uptake has led to the need to look at age-specific factors to unveil

specific needs, particularly for the youth (Prata et al., 2016). Johnson, 2017 analyzed

Nigeria's 2013 DHS and found socio-demographic characteristics such as education, age,

and residency area to influence the uptake of contraception (Johnson, 2017). Once

adjustment for other factors (such as education and wealth), there was little disparity in

FP uptake between the rural and urban areas of residence. However, in other studies,

one's residence has been noted to be a significant determinant of access for FP (Oyugi et

al., 2017). The residence affects the access to the FP services, with those living in the

rural setting having less access due to various factors (Jalu et al., 2019). For example,

services tend to be far in many rural areas compared to urban settings. Besides, those in
 47
the urban area are likely to have more exposure to information, including mass media and

the internet (Jalu et al., 2019)

Education level and marital status were positively associated with current modern

contraceptive use among women aged 15–24 (Prata et al., 2016). Less-educated women

are less likely to use FP (Asaarik, & Adongo, 2015; Prata et al., 2016). In another study,

Mutumba et al., 2018 found that higher education, wealthier status, urban residence, and

exposure to mass media positively influenced FP uptake (Mutumba et al., 2018). Even

among postpartum women, the level of education has a significant and positive

association with the utilization of postpartum FP services (Bwazi et al., 2014).

However, some studies have shown mixed results as far as the association of

socio-demographic characteristics with FP uptake is concerned. Ajong et al., 2016, found

some degree of association though statistically insignificant between the unmet need for

FP and level of education, religion, and number of years of cohabitation. According to

Prata et al. (2016) young people tended to have more knowledge of condoms than other

methods and thus was the preferred method.

The discussed studies provide critical insights on the covariates that may be

associated with FP uptake. They demonstrate that it is crucial to study the covariates to

establish if they are confounders to the independent variables of interest. However, these

studies are not specific to Kenya, and some are qualitative, hence not suitable for

inference. This study will provide information that is specific to Kenya. Establishing the

association between the covariates with FP uptake will provide valuable evidence on

addressing the FP needs of the different groups in the Kenyan population.

 48
Utilization of care, particularly maternal and child health services such as

delivery, ANC, PNC, and immunizations for the infant, may be related to personal

characteristics such as socioeconomic status, attitudes, and perceptions. They could also

be related to organizational factors such as availability and quality of services. They have

been discussed together with organizational factors due to their close association.

However, they will be considered covariates, and strictly speaking, they do not fit in

either the organizational level or the intrapersonal level in the SEM.

Definitions

Family Planning

Family planning is the term used to refer to the preparation, knowledge, and

methods that assist people to plan and attain their desired family size and to determine the

spacing of pregnancy (WHO, 2018). It involves making decisions. Therefore, FP services

include the provision of counseling to vail the necessary information to make informed

choices. Though the term FP often refers to pregnancy prevention, strictly speaking, it

does include fertility treatment to enhance conception. In this paper, the term FP refers to

pregnancy prevention. Some methods of FP are modern, while others are traditional.

Contraceptives

Contraceptives are the methods used to prevent pregnancy for achieving the

desired family size and spacing (WHO, 2018). Thus contraceptives are used to assist

individuals and coupes to implement their FP decisions. This relationship may explain

why the terms contraception and FP are used interchangeably in many write-ups. The

majority of the available contraceptive methods are for women, with men having a
 49
limited choice of modern methods, mainly limited to male condoms and sterilization

(vasectomy) (WHO, 2018). The limited availability of FP methods for men may explain

why most of the FP discussion refers to women.

Modern Methods of Family Planning

The definition of modern methods has varied across organizations; in most cases,

organizations list the methods without a clear criterion (Hubacher & Trussell, 2015). The

hormonal methods and devices, such as the intrauterine contraceptive device (IUCD)

used to control fertility, seem universally accepted as modern methods. However,

methods based on fertility awareness have raised controversy. World health organization

has listed fertility awareness methods such as the sympto-thermal method, two-day

method, basal body temperature method, standard days method, and the LAM as modern

methods. In contrast, it has classified the calendar method or rhythm method and

withdrawal (coitus interruptus) as traditional methods (WHO, 2018). Hubacher and

Trussell, 2015 propose a definition of modern methods as a product or medical procedure

that interferes with reproduction from acts of sexual intercourse (Hubacher & Trussell,

2015). With this definition, a method such as LAM does not qualify as a modern method.

This study considered the following methods as modern methods; the combined pill, the

mini-pill, the IUD, injectables, condoms, female sterilization, male sterilization, implants,

lactational amenorrhea, and female condom as classified by the DHS program

(2019).definition of modern methods has varied across organizations; in most cases,

organizations list the methods without a clear criterion (Hubacher & Trussell, 2015). The

hormonal methods and devices, such as the intrauterine contraceptive device (IUCD)
 50
used to control fertility, seem universally accepted as modern methods. However,

methods that based on fertility awareness have raised controversy. World health

organization has listed fertility awareness methods such as the sympto-thermal method,

two day method, basal body temperature method, and standard days method as well as

the LAM as modern methods. In contrast, it has classified the calendar method or rhythm

method and withdrawal (coitus interruptus) as traditional methods (WHO, 2018).

Hubacher and Trussell, 2015 propose a definition of modern methods as a product or

medical procedure that interferes with reproduction from acts of sexual intercourse

(Hubacher & Trussell, 2015). With this definition, a method such as LAM does not

qualify as a modern method. In this study, the following methods are considered as

modern methods; the combined pill, the mini pill, the IUD, injectables, condoms, female

sterilization, male sterilization, implants, lactational amenorrhea, and female condom as

classified by the DHS program (2019).

Contraceptive Prevalence Rate

CPR is a critical indicator used to measure the access of FP in a population.

International frameworks such as the millennium development goals use CPR, which

measures married women's proportion on an FP method (United Nations, 2017). The

restriction to married women or those in unions allows for comparable data across the

countries (United Nations, 2017). Often CPR measures the modern methods used by

married women, but CPR for other categories of women is calculated and specified as

such. For example, CPR for all sexually active or CPR for all methods (including the
 51
traditional ones) is determined. In this study, CPR for the various categories will be

calculated and specified appropriately.

Unmet Need for Family Planning

Unmet need refers to the proportion of sexually active women of reproductive age

who are capable of becoming pregnant, who want to limit (no more children) or to space

(postpone pregnancy) their children, but are not using contraception (Kennedy et al.,

2011). It has been used as a measure of FP programs’ effectiveness and was the

millennium development goals indicator 5.6 for monitoring target 5.B, which aimed at

achieving universal access to reproductive health by 2015 (United Nations 2015).

Method Mix

Method mix refers to how the distribution of FP among users. It is a proxy

indicator of the availability of different methods to the women and is therefore critical in

supporting FP's informed choice (Bertrand et al., 2014).

Adolescents

Young people refer to the age group between 10 and 24 years, which is further

divided into adolescents and youths; adolescents encompass ages 10 to 19, while 20 to 24

are young adults (Patton et al., 2016). Adolescence is further divided into early (10 to 14)

and late (15 to 19). The DHS contains reproductive health data for women aged 15 and

49 years. This study considered adolescents between 15 and 19 years (DHS Program,

2019).
 52
Postpartum Women and Postpartum Family Planning

The postpartum period and the postnatal period are closely related, and the two

terms are often used interchangeably. However, strictly postpartum refers to the mother's

issues, and postnatal refers to those concerning the baby (WHO, 2010). The postpartum

period is the period soon after delivery up to six weeks after delivery. Hence, postpartum

women are those within six weeks of delivery of a young one. However, for FP, up to one

year is considered (WHO, 2013). Since this study focuses on FP, postpartum women will

be considered as those within one year since the delivery of a young one, regardless of

whether the young one is alive or not.

Assumptions

This study's primary assumption is that the sample is nationally representative,

given the large sample size, the experience of the organizations conducting the study, and

the high response rates with minimal missing data. This study has analyzed KDHS 2014

data, a population-based nationally representative survey conducted through the

cooperation of many international organizations and Kenya's government (International

Household Survey Network, 2019). Its overall objective was to evaluate and monitor the

health of the Kenyan population. Some of the specific objectives included estimating the

fertility and measure changes in fertility and contraceptive prevalence. The KNBS was

the leading implementing organization and thus led to the survey's planning and the

analysis and dissemination of the survey results. The KNBS, supported by other

organizations, was responsible for the survey's operations and received technical

assistance from the ICF International staff (Kenya National Bureau of Statistics, 2015).
 53
The questions' response rate was high; of the selected demographic and health questions,

the average missing response was less than 1%, and missing anthropometry information

for children was 2.41%. That of women was 5.61%. These proportions indicate high data

validity given the high response rate.

Scope and Delimitations

The study participants were women aged 15 to 49 years found in the pre-selected

households; the total number of women age 10 to 54 who had slept in the households that

were visited the night before the interview was 32,247. Women younger than 15 years

and those older than 49 were excluded. There was a high response rate of 96.1%. This

high response rate may be attributed to the quality training that the interviewers had

received and the mass publicity the survey received in the country.

This study analyzed adolescents and postpartum women. The analysis will

provide information that is specific to these priority groups in Kenya. The KHDS was not

per any theory. However, the secondary analysis used SEM.

Significance, Summary, and Conclusions

FP is a well-known public health intervention that reduces the unmet need for

limiting and spacing pregnancies. Reducing the unmet need for FP has been highlighted

as one of the cost-effective public health strategies in low and middle-income countries

with significant benefits for both mothers and infants, reduces demand for abortion as

well as vertical transmission of HIV (Zakiyah et al., 2016). For contraception to

significantly reduce maternal mortality, it must reach the most at risk and those with an

unmet need (Ganatra & Faundes, 2016). A literature review has shown that adolescents
 54
and postpartum women are priority groups due to their high unmet need for FP and the

high risk of poor pregnancy outcomes they carry. It takes political (organizational

factors) as well as socio-cultural (individual and community/societal factors) decisions to

have contraception used or not used (Stiegler & Susuman, 2016). A unique challenge is

associated with reaching postpartum women and adolescents, resulting in marked

disparities and inequities within these groups (Dennis et al., 2017; Sonalkar et al., 2013;

Vogel et al., 2015). Besides the health benefits of FP, there are related development

gains. Various international and national targets for health and development, including

the sustainable development goals, have been set. To achieve these targets by 2030, the

majority of countries, including Kenya, need to accelerate the uptake of FP (Choi et al.,

2015). Evidence is needed to support favorable decisions to support FP uptake at all

levels. Though there are many studies internationally, regionally, and nationally, there

lacks nationally representative data in Kenya for adolescents and postpartum women

despite their high unmet need for FP. Many studies have not examined the factors

contributing to FP use holistically by considering all SEM levels. This study will bridge

this gap and provide information that will support better programming for the adolescents

and the postpartum women, to enhance their FP uptake.

Increased use of FP can improve the health of women and their children and

empower adolescent girls to avoid unintended pregnancy, thus improving their

opportunity to pursue education. Education will enhance the girls' chance of being

economically independent. Improved health and economic status will, in turn, enhance
 55
the general quality of life (Canning & Schultz, 2012). Additionally, fertility decline

reduces youth dependency, further strengthening economic growth.

Increased uptake of FP can prevent unintended pregnancies and all their

consequences, thus saving millions of dollars that are made available for other social

services and economic development, thus improving the general quality of life for all, not

just women (Canning & Schultz, 2012). These gains are very significant positive social

changes in society.

KDHS secondary analysis tested the significance of various factors and

determined their association with FP uptake for adolescents and postpartum women. The

KDHS is a nationally representative survey done using international standards to ensure

quality information is collected. The sampling frame used is from the country's census,

following international standards to ensure its internal and external validity. The analysis

in this study followed statistically sound methods to ensure quality information. The

reporting is relevant to the Kenyan context.

 56
Section 2: Research Design and Data Collection

Introduction

This study examined the factors that contribute to FP uptake among priority

groups (adolescents and postpartum women) in Kenya. The primary dependent variable

was the uptake of FP, and various independent factors were analyzed. The uptake of FP

has been expressed as the proportion of adolescents or postpartum women who reported

to have been using a modern method of FP during the survey. The independent variables

included the intrapersonal, interpersonal, community, and organizational factors.

Sociodemographic characteristics and characteristics on utilization of services were

considered as covariates. Analysis was done to establish their association with FP uptake

among adolescents and postpartum women and how they differ by regions. Data from the

KDHS were used. The results of this study will provide important information on

improving FP programming for adolescents and postpartum women in Kenya. The results

of this study will also provide feedback that can inform subsequent KDHS questions in

the future. It will provide a baseline for comparison on the factors associated with FP

among adolescents and postpartum women for future KDHS data as well as future

surveys on the status of factors influencing FP uptake across the regions for postpartum

women and adolescents.

Research Design and Rationale

Study Variables

This quantitative study sought to establish the factors that are associated with the

uptake of FP among adolescents and postpartum women. The dependent variable was FP
 57
uptake, which was measured by the current use of FP. From the data, the proportion of

adolescents and postpartum women of modern methods of FP was determined. The

independent variables were the various factors that influence FP uptake, which were

organized according to the different SEM levels. The intrapersonal factors included

knowledge on FP and exposure to different sources of FP information. Perceptions on FP

were extrapolated from reasons given for not using FP. Interpersonal variables included

partner and other family involvement in decision-making as well as influence from

friends. Community-level variables included mean age at marriage, mean age at first

birth, and mean age of sexual debut, household decision-making norms, community

perception on the ideal number of children, and community partner violence attitudes. At

the organizational level, the variables included access to the services (distance and cost),

counseling on FP, availability, and utilization of maternal and child health services such

as ANC, delivery, postnatal care and immunizations. Covariates included

sociodemographic characteristics such as education level and wealth status and use of

maternal and child health services (seeking care during pregnancy, delivery, and

postpartum period, and immunizations for the infant).

Research Design

This study was a quantitative, correlational research utilizing a cross-sectional

survey design. Its aim was to answer the question regarding the factors that are associated

with FP uptake (use) by adolescents and postpartum women. Secondary analysis of the

KDHS has provided information on the different factors that influence the uptake of FP

in Kenya. The chi-square of association was used to determine which factors significantly
 58
influence the uptake of FP. Logistic regression model was used to determine the

contribution of various factors. The factors were grouped according to the SEM levels.

Models were created for adolescents and for the postpartum mothers.

The DHS is a massive data set that required time and sufficient knowledge and

expertise in statistics as well as FP. It was critical to identify the relevant questions that

answer the research questions from the different sections of the survey. Understanding

the way the questions were asked and the coding of each variable was required to ensure

proper interpretation for validity. Recoding was necessary for the analysis to answer

particular specific research questions. No payments were required to obtain the data.

FP access is used as a proxy indicator of access to reproductive health services.

FP access is a priority that was in the millennium development goals and is now in the

sustainable development goals, and CPR is used as the indicator (United Nations, 2017).

Despite much progress made in women getting FP and other maternal health services,

there is a need to go beyond the traditional programming for universal access to be

realized (Kissoon et al., 2015). One area of concern is inequity across and within

countries where some geographical areas or some segments of the population are left

behind (United Nations, 2017). FP programs need to be evidence-based to promote

equitable access to all. The KDHS is done scientifically and provides nationally

representative quantitative data on women of reproductive age and now covers all

geographical areas. This study has been done according to the scientific standards, the

data have provided specific information on adolescents and postpartum women, thus

filling critical gaps on these priority groups.

 59
Methodology

Population

The KDHS 2014 collected information from women aged 15–49 years and from

men aged 15–54 years. According to the 2009 population census, women of reproductive

age were 9,375,784 with 2,045,890 being 15 to 19 and 2,020,998 being 20 to 24 years of

age (Kenya National Bureau of Statistics, n.d.). The number of households that were

enumerated was 40,300 up from 9,936 households in the previous DHS

(PopulationPyramid.net, n.d.).

The total population as per the latest census is 47,564,296, of which 23,548,056

were males, 24,014,716 were females, and 1,524 were intersex (Kenya National Bureau

of Statistics, 2019). The population grew from 37.7 million in 2009 to 47.6 million in

2019. The population between 15-24 years was 19.61% of the population, and 34.27%

were 25-54 years old (CIA World Factbook, 2019). From these data, the female

adolescents and youths (15-24 years) were about 4,752,896. The KDHS 2014 interviewed

6,078 females who were 15 to 19 years old and 5,405 females 20 to 24 years of age

(Kenya National Bureau of Statistics, 2015). Despite the increase in population, this is a

sample size that will still be representative given the sampling design applied and the

increased number of households compared to the previous surveys.

Sampling Strategy and Procedure

The KDHS utilized a multistage sampling strategy. First, clustering was done

following geographical representative units, then probability random sampling within the

clusters followed. Thus, samples were selected independently from each sampling
 60
stratum. During the first stage, a total of 1,612 enumeration areas were selected with an

equal probability of being selected from the sampling frame. For the second stage, the

listed households within the clusters served as the sampling frame. From each cluster, 25

households were selected randomly.

There was an increase in the number of households from 9,936 in the previous

2008 KDHS to 40,300. The increase was due to the need to capture data at the county

level, which are much smaller units, whereas previously the segregation was up to the

provincial level (now considered as regions, in the 2014 KDHS). To ensure data quality,

given this considerable increase, only priority indicators at the county level were

collected; hence, there were full questionnaires and short questionnaires. The sample was

divided into halves; one half received the full Household Questionnaire, the full

Woman’s Questionnaire, and Man’s Questionnaire, and the other half received the short

Household Questionnaire and the Short Woman’s Questionnaire. In each cluster, one in

every two households was selected for the full questionnaires, and the remaining

households were selected for the short questionnaires. The short questionnaires were a

subset of the long questionnaires. Data collected in both the long and short questionnaires

can produce estimates at the national, rural/urban, and regional levels, but the data only in

the long questionnaires is not suitable for making estimates at the county level.

Sampling Frame

The sample for the KDHS was drawn from the master sampling frame, the Fifth

National Sample Survey, and Evaluation Programme. The frame contained a total of

5,360 clusters, which were divided into four equal subsamples. The clusters were drawn
 61
using a stratified probability proportional to size sampling methodology based on the

2009 Kenya Population and Housing Census. For this survey, two subsamples of the

frame that were developed in 2013 and updated in 2014 were used. Each of the 47

devolved units of administration called counties was stratified into urban and rural strata.

Two of the counties, Nairobi and Mombasa, have only urban areas; hence, the total

sampling strata was 92. The design aimed at producing nationally representative

indicators at the national level for urban and rural areas separately, at the regional (former

provinces) level, and for selected indicators at the county level. The sample had 40,300

households from 1,612 clusters spread across the country, with 995 clusters in rural areas

and 617 in urban areas. In each sampling stratum, a two-stage sample design was used to

select the samples independently. For the second stage of selection, the households from

listing operations served as the sampling frame, and 25 households were selected from

each cluster.

Sample Size

Many factors are considered in determining the sample size, such as the

methodology and the effect size or the prevalence of the outcome of interest. Descriptive

studies require a larger sample size compared to experimental studies, and a one-tailed

hypothesis requires less sample compared to a two-tailed hypothesis. Simple random

sampling requires fewer sample subjects compared to other designs. The outcome of

interest is also an important consideration (Kaliyadan & Kulkarni, 2019). The KDHS was

a descriptive cross-sectional study that had a two-sided hypothesis, utilized multistage

sampling, and the many outcome variables were categorical.

 62
Another consideration for the sample size was the need to have regional

representation as per the administrative units. With the promulgation of a new

constitution in 2010, the counties were much smaller units compared to the previously

used provinces; hence, a bigger sample size was needed. The determination of a

nationally representative sample was done with all these factors taken into consideration.

The sample size for women age 15–49 years was 32,172, with 16,855 for the short

questionnaires and 15,317 for the long questionnaires. The sample for men aged 15 to 54

years was 14,217 14. According to the KDHS report, 6,078 females who were 15 to 19

years old were interviewed (Kenya National Bureau of Statistics, 2015). Since the sample

was nationally representative, it was used as such to retain generalizability. However, to

be sure that the sample size for the subgroups of interest (adolescents and postpartum

mothers) were within the minimum acceptable sample size, sample size calculation was

done. The following formula was used (Suresh & Chandrashekara, 2012):

N = Z2α/2 * P* (1-p) * D
E2

N= 1.96 x 0.532 x (1- 0.532) x 1.503/ 0.05322 = 1.96x 0.532 x 0.468 x 1.503/

0.00283=0.73345/0.00283 = 259.16 = 260

N is the sample size. P is the prevalence or proportion of event of interest for the

study. According to the 2014 DHS the proportion of currently married women of

reproductive age using a modern method is 53.2%. E is the Precision (or margin of error);

generally, E is 10% of P. Zα/2 is normal deviate for two-tailed alternative hypothesis at a

level of significance; for 5% level of significance, Zα/2 is 1.96. D is the design effect,
 63
which reflects the sampling design; for cluster random sampling is taken as 1.5 to 2. The

design effect for the 2014 KDHS, for women currently using a modern method is given

as 1.503.

Procedure

The self-reported questionnaire targeted women of reproductive age, 15–49 years

of age, and men aged 15–54 years. The interviewers visited only the preselected

households for data collection. Information was collected for individuals who had slept in

the household the night before the survey. The Household Questionnaires and the

Woman’s Questionnaires were administered in all households, but the Man’s

Questionnaire was administered in every second household. The resulting data were

weighted to be representative at the national, regional, and county levels.

Identification of participants eligible for individual interviews was done using

information from the household questionnaire. The household questionnaire was used to

list all the members of the household who regularly stayed there and visitors who stayed

in the household the night before the survey. The questionnaire collected personal details

such as the sociodemographic characteristics and the relationship to the head of the

household. Household characteristics such as the source of water, type of toilet facilities,

materials used for the floor and roof of the house, and ownership of various durable

goods were collected. The data were collected face to face.

After the eligible men and women in the household were identified, the relevant

questionnaire was administered. Men aged 15–54 years living in every second household

in the sample were interviewed using the Man’s Questionnaire. The Man’s Questionnaire
 64
collected information similar to that contained in the Woman’s Questionnaire. However,

it was shorter because it did not contain questions on maternal and child health, nutrition,

mortality, female circumcision or fistula. These excluded questions are not of interest in

answering the research questions for this study.

The questionnaires were prepared in English. After finalization they were

translated into 16 local languages. The translated questionnaires were pretested to detect

any possible to ensure proper flow as well as to gauge the length of time required for

interviews.

Gaining Access to the Data Set

The KDHS data are available online on the DHS website. The requirement is to

register as a user of the website and then request for the data. The information required is

the purpose of the data, the name of the project, and the specification of the particular

survey data requested. The requesting person agrees to abide by several conditions,

including maintaining the confidentiality and only use the data for the purpose stated in

the request. The request is then processed, and feedback is received via email within 48

hours. The permission is granted via email with an attached letter authorizing one to

download. This process was done for this study. There were no charges; the data are free

to access.

Instrumentation and Operationalization of Constructs

The 2014 KDHS questionnaires were developed on the bases of the previous

survey questionnaires. The DHSs are nationally representative population-based

household studies conducted in many countries, including Kenya (International

 65
Household Survey Network, 2019). They have been conducted since 1984, primarily

funded by the primarily by the U.S. Agency for International Development and individual

countries receive technical support from the DHS program. In Kenya, previous DHSs

were done in 1989, 1993, 1998, 2003, and 2008–2009 (International Household Survey

Network, 2019).

For this survey, the questionnaires were developed through consultation of many

stakeholders; a two-day workshop was held. Desk reviews were done to ensure that the

country’s needs were being met. This included considering the objectives of the relevant

population and health policies and programs. The objectives of a new population and

national development policy released by the government in 2012 were considered. Such

objectives included reducing population growth rate to be in tandem with economic

growth and social development goals envisioned in the Kenyan Vision 2030, fertility and

mortality rates reduction, provision of equitable and affordable quality reproductive

health services, including FP among other objectives. The 2010 Kenya constitution

devolved the health function to the county governments, with distinct functions being

assigned to the national and county governments. The Kenya Health Policy 2014–2030

objectives and principles, which include equity in the distribution of health services and

interventions, were considered in the development of the instrument. Hence, there was

the need to have data that could be segregated to the county levels unlike in the past.

Priority indicators were selected based on policies and programs' objectives and

with the input of the stakeholders. Questionnaires were developed for the household, for

the individual woman, and individual man. The household questionnaire had information
 66
on identification, usual members and visitors in the selected households, background

information on each person listed, such as relationship to head of the household, age, sex,

marital status, survivorship and residence of biological parents, and highest educational

attainment. It also had characteristics of the household's dwelling unit, such as water

source, type of toilet facilities, materials used for the floor, roof, and house walls. The

ownership of various durable goods (these items are proxy indicators of the household's

socioeconomic status) was also included. The individual woman questionnaires collected

information on background characteristics (education, marital status, media exposure),

reproductive history, FP knowledge and practices, sexual activity, husbands' background

characteristics, awareness, and behavior regarding HIV and other sexually transmitted

infections. The individual man questionnaires had information on the respondent's

background characteristics, reproduction, contraception, marriage, sexual activity,

fertility preferences, employment and gender roles, HIV/AIDS, and other health issues.

Reliability and Validity of the Study Instrument

Reliability refers to the consistency and stability of measurements, while validity

has to do with measuring what is intended. Care has been taken over the years by the

DHS program to ensure that the survey questionnaires are reliable and valid in their

measurements.

The KDHS used self-reported questionnaires, targeted women of reproductive

age, 15–49 years of age, and men aged 15–64, living in the sampled households, and

collected a wide range of health questions, including fertility, FP, maternal and child
 67
health among others. The questionnaires included sociodemographic factors such as

education level, wealth status, access to various amenities such as water, and residential

areas. Reproductive health data include knowledge on FP methods, altitude, practice, and

service utilization, including related factors. The man’s questionnaire had questions

similar to the woman’s but did not include the maternal and child health, nutrition, adult

and maternal mortality, or female circumcision or fistula experience.

Data quality in a survey can be affected by either sampling or non-sampling

errors. The survey methods, the respondents, and the instruments or measurements can

introduce measurement error (Boo & Froelicher, 2013). It is challenging to evaluate non-

sampling errors statistically. Multistage sampling was done to minimize sampling errors

to ensure the data was nationally representative, as detailed in the sampling section.

Sampling errors can be computed statistically. For the DHS, sampling errors are

computed in either Integrated System for Survey Analysis or Statistical Analysis System,

using ICF Macro programs (Kenya National Bureau of Statistics, 2015). The Taylor

linearization method of variance estimation for survey estimates was used for means,

proportions, or ratios. Simultaneously, the Jackknife repeated replication method was

used for variance estimation of complex statistics such as fertility and mortality rates.

To ensure non-sampling errors were minimized, stringent measures were taken,

including training all personnel involved in data collection and processing and strict data

cleaning procedures.

One week-long training was conducted to ensure that the interviewers understood

the study instruments and used them as intended for validity. The training focused on the
 68
survey design concepts and the content of the questionnaires. Adult teaching techniques

were used. The trainers continued to support the process; they led the pre-test exercise

and served as coordinators during data collection. After the training pre-test of the

questionnaires, including the translated ones, was done. Pre-testing was done but within

the clusters not included in the survey data collection. The lessons learned from the

exercise were used to improve the questions and enhance the logistics of the data

collection. For quality data collection, several personnel categories were recruited and

trained thoroughly to ensure a full understanding of the study design, the questionnaires,

and the logistics involved in data collection. These personnel included 48 supervisors, 48

field editors, 144 female interviewers, 48 male interviewers, 28 quality assurance

personnel, and 20 reserves.

Completed questionnaires were sent to the Data Processing Centre in Nairobi.

Here, cluster and household numbers were verified to ensure that they were consistent

with the sampled list and that each cluster had questionnaires for 25 households. Training

for data entry was done before starting the entry. For verification, all data were double

entered using the CSPro software. As entry is done, secondary editing, including further

data cleaning and validation, was done simultaneously.

Low response rates and missing data can reduce the sample size and hence reduce

the validity. The selected sample consisted of 39,679 households, and 36,812 were found

occupied at the time of the data collection, and a total of 36,430 were successfully

interviewed. Thus, an overall household response rate of 99percent. The majority of the

households that were not interviewed were due to structures found vacant or destroyed,
 69
and others where occupants were absent for a long time. A total of 31,079 women age 15

– 49 were interviewed out of 32,172 of the eligible women giving a response rate of 97%.

14,741women out of the eligible 15,317were interviewed using the long questionnaire

giving a response rate of 96%. 12,819 men out of 14,217 eligible men were interviewed,

giving a response rate of 90%. For the short questionnaires, 16,338 women out of 16,855

eligible women were interviewed, translating into a 97% response rate. The response rate

was lower in urban areas, particularly for men, due to failure to find them at home despite

several visits. Missing data on variables of interest can affect the data quality and validity

of the results. In this survey, great care was taken to ensure minimal missing data.

The questions have been studied carefully to understand what they measured and

the scale to ensure that the questionnaires sufficiently answer the research questions. The

data codebook has also been studied to ensure that the variables can be analyzed to

answer the intended questions.

Operationalization of Variables

When conducting secondary analysis, it is critical to ensure that data are

operationalized and coded appropriately to allow for the intended analysis. The database

evaluation was done by reviewing the survey's purpose and reports and examining the

manual's codebooks and operations manual. One of the main objective s of the KDHS

was to collect FP information, including determining the fertility rates and some factors

related to FP utilization.

Data Analysis Plan

SPSS (v. 24) was used for analysis to answer the research questions.
 70
Research Question 1: To what extent are intrapersonal, interpersonal, community,

and organizational factors (independent variables) associated with the uptake of family

planning (dependent variables) among adolescents in Kenya?

H01: Intrapersonal, interpersonal, community, and organizational factors are not

associated with the uptake of family planning among adolescents in Kenya.

Ha1: Intrapersonal, interpersonal, community, and organizational factors are

associated with the uptake of family planning among adolescents in Kenya.

Research Question 2: To what extent are intrapersonal, interpersonal, community,

and organizational factors (independent variables) associated with the uptake of family

planning (dependent variables) among postpartum women in Kenya?

H02: Intrapersonal, interpersonal, community, and organizational factors are not

associated with the uptake of family planning among postpartum women in Kenya.

Ha2: Intrapersonal, interpersonal, community, and organizational factors are

associated with the uptake of family planning among postpartum women in Kenya.

Research Question 3: To what extent are the intrapersonal, interpersonal,

community, and organizational factors (independent variables) associated with family

planning uptake (dependent variables) among postpartum women and adolescents differ

by counties in Kenya?

H03: Factors associated with family planning uptake among postpartum women

and adolescents in Kenya do not differ by counties.

Ha3: Factors associated with family planning uptake among postpartum women

and adolescents in Kenya differ by region.

 71
Descriptive analysis was carried out first for the socio-demographic

characteristics such as age, level of education, wealth status, use of FP, and the different

factors of interest such as exposure to FP messages. Splitting into adolescents and older

women was done, followed by splitting by regions for various variables. The chis square

test of association was used to test which factors (variables) were significantly associated

with FP use for adolescents and postpartum women. Logistic regression analysis was

done to assess the contribution of different factors to the uptake of FP.

According to the SEM framework (intrapersonal, interpersonal, community, and

organizational levels), the results have been reported for the different independent factors

for the adolescents and postpartum mothers. For each significant variable, the odds ratios

were examined to get the probability to use or not to use FP.

Interpretation of Results

The dependent variable is family planning uptake, which is measured by getting

the proportion of adolescents and postpartum women who reported being on a method

during the interview. The pattern of use, including those who have ever used and future

intention to use a method, was described. The discontinuation pattern was described as

well as the reasons for the discontinuation. FP use was related to the different

independent variables and covariates, and the association reported among adolescents and

postpartum women. The distribution of the different factors was analyzed and discussed

in the report.
 72
Threats to Validity

The respondents may want to give answers that are socially acceptable and may

be affected by recall bias. The instrument needs to be reliable for accurate data. The

wording and order of questions in the instrument and the timing of data collection can

influence the responses' reliability and validity (Bierman & Bubolz, 2003). The

questionnaires were pre-tested, and the interviewers trained well to minimize any errors.

Ethical Procedures

Secondary analyses of nationally representative surveys are cost-effective, raise

few ethical issues, and are a fast way of doing nationally representative researcher (Boo

& Froelicher, 2013). The DHS Program follows strict standards to protect respondents'

and household members' privacy in all DHS surveys (DHS Program, 2019). The ICF

Institutional Review Board (IRB) reviews the country's specific survey protocols,

including the questionnaires, and gives ethical approval. Additionally, an IRB in the host

country reviews and gives ethical review independently. ICF IRB ensures compliance

with the U.S. Department of Health and Human Services regulations to protect human

subjects (45 CFR 46). The host country IRB ascertains that the survey complies with the

host's laws and norms country. Before each interview, an informed consent statement is

read to the respondent, and they are allowed to accept or decline to participate. In Kenya,

the survey materials are translated into various relevant languages to ensure the

respondents understand. The statement outlines the interview/test's purpose, the expected

duration of the interview, the interview procedures, any potential risks and benefits to the

respondent, the contact information for more information about the interview, and
 73
emphasizes the voluntary nature of participation. Privacy and confidentiality are strictly

observed due to some questions' sensitive nature, such as those on sexual activity. After

data processing, the identifiers are destroyed. The geographic coordinates of each survey

are displaced at a random distance and in a random direction. Where testing such as for

HIV or anemia is done, treatment and referral services are made available as appropriate.

For data access, there are well laid down procedures and declarations to ensure the data's

further ethical use. For this secondary analysis, the necessary approvals for the data's

access have been made, and the Walden IRB approval was obtained before any analysis

was done.

Summary

This study is a cross-sectional quantitative research that involved secondary

analysis of the 2014 KDHS. It has determined the independent variables and covariates

significantly associated with FP uptake among adolescents and postpartum women in

Kenya. The logistic regression model was used to establish how different factors

contribute to the uptake of the FP and by the adolescents and postpartum women. The

independent variables were categorized according to the SEM levels of influence and

association of different FP use factors described.

Results are organized into descriptive analyses of various characteristics and

inferential statistics describing the relationship between the dependent and independent

variables. The multi-variable analysis has reported the degree to which various variables

contribute the uptake of FP. The findings are organized by research questions to answer

the hypothesis associated with each research question.

 74
Section 3: Presentation of the Results and Findings

Introduction

This study examined the factors that contribute to FP uptake among adolescents

and postpartum women in Kenya. The analysis sought to answer three research questions

and hypotheses related to the extent that intrapersonal, interpersonal, community, and

organizational factors are associated with the uptake of FP among adolescents and

postpartum women in Kenya as well as the differences across the regions. The primary

dependent variable was the uptake of FP, and various independent factors were included.

The independent variables were categorized according to the socioecological levels

(intrapersonal, interpersonal, community, and organizational factors). Sociodemographic

characteristics were considered as covariates. These characteristics are described and

their association with FP uptake among adolescents and postpartum women including

how their distribution by regions reported. Data from Kenya DHS 2014 were used. Based

on the findings, recommendations to inform subsequent KDHS questions on FP in the

future are made. The results provide a baseline for comparison on the factors associated

with FP among adolescents and postpartum women and their distribution across the

regions using future KDHS data.

Data Collection of Secondary Data Set

This study involved the secondary analysis of the KDHS 2014 data. It is a

population-based survey that was conducted through the cooperation of many

international organizations and the government of Kenya (International Household

Survey Network, 2019). Its overall objective was to evaluate and monitor the health of
 75
the Kenyan population. Some of the specific objectives included estimating the fertility

and measure changes in fertility and contraceptive prevalence. The KNBS was the

leading implementing organization in partnership with the Ministry of Health and thus

led in the planning of the survey and the analysis and dissemination of the survey results.

The KNBS was also supported by other organization and received technical assistance

from the ICF International staff but was responsible for the operations of the survey

(Kenya National Bureau of Statistics, 2015). The other organizations that partnered with

the KNBS included the National AIDS Control Council, the National Council for

Population and Development, and the Kenya Medical Research Institute). The survey

was conducted from May 2014 to October 2014.

Sampling and Sample Characteristics

A multistage sampling strategy was used first clustering following geographical

representative units, followed by probability random sampling within the clusters. Thus,

samples were selected independently from each sampling stratum. During the first stage,

a total of 1,612 enumeration areas were selected with an equal probability of being

selected from the sampling frame. In the second stage, the listed households within the

clusters served as the sampling frame with 25 households being selected randomly from

each cluster. During data collection, the interviewers visited only the preselected

households for data collection. Information was collected for individuals who had slept in

the household the night before the survey. The self-reported questionnaire targeted

women of reproductive age, 15–49 years of age, and men aged 15–54 years. The total

number of eligible women was 32,172 out of whom 96.6% were interviewed, giving a
 76
total of 31, 079. The response rate to the questions was high; the average missing

response was less than 1%, and missing anthropometry information for children was

2.41% and 5.61% for women. These proportions indicate high data validity given the

high response rate. This high response rate may be attributed to the quality training that

the interviewers had received and the mass publicity the survey received in the country.

For this study, the analysis focused on adolescents and postpartum women, and various

factors were tested to determine if there was a statistically significant association with

uptake of FP. The factors were organized into the various SEM levels.

Analysis

Discrepancies from Initial Plan

Though not in the initial analysis plan, analysis for adolescents included a

comparison of their characteristics to those of older women rather than just focusing on

the group itself. This approach was adopted to give more information on differences

between the adolescents and the older women, which could point to inequities. Hence, the

results provide critical information for the policy makers and program implementers to

support the empowerment of adolescents.

Some factors such as the personal and community perceptions on FP (myths and

misconceptions) were not evaluated because there were no suitable variables in the data

set. To facilitate analysis, recoding was done for several variables. To get the sample for

adolescents the categorical age variable was recoded with 15 to 19 years as adolescents

and the rest as older women. For postpartum women, the variable “births in the past year”

was used with those who had no birth within the past year being coded as other women
 77
and those with a birth within the year being coded as the postpartum women. First

bivariate analysis was done using the chi-square test to determine the factors that were

significantly associated with FP uptake, both for the adolescents and for the postpartum

women. The assumptions of the chi-square test were met in all the cases where the results

are reported. Thus, variables used in the chi-square tests were all independent

observations (mutually exclusive), frequencies, categorical measured as nominal,

measured at one point in time, and no cell had an expected value of less than one. The

factors were organized by the SEM levels.

To characterize the sample, various sociodemographic characteristics were

analyzed including marital status, education level, literacy, area of residence and the

wealth index. Characteristics related to utilization of services such as place of delivery

and seeking various services are as a results of the interaction of the different factors in

the SEM level. They were included as covariates to provide more insights about the study

population and to test if they are associated with use of FP. Their results are reported with

the rest of the results.

Additionally, for the sociodemographic characteristics, marital status was coded

into those who had never been in a union, those currently in union, and those who were

no longer in a union. The education achievement was coded into those who have primary

school and below, those with secondary level, and those with aabove secondary. On the

wealth index, the recoding was done into poor, middle, and rich. For literacy, the

recoding was done to cannot read, able to read, and not applicable (where no cards were

available or the visually impaired).

 78
In order to test the association between use of modern methods and for inclusion

in the binary logistic model, some characteristics were also recoded further. Level of

education was recoded into primary level and below (incomplete secondary education

and below) and those with secondary level and above. On wealth status the poorest and

poorer were coded as poor and the middle, richer, and richest were coded as not poor. On

literacy, those who could not read were one category, and those who could read parts of a

sentence and those who could read whole sentences were classified as can read with those

where no card for reading or the blind were classified as missing.

Some of the independent variables and covariates were also recoded. For the

place of delivery those who delivered in any facility whether public or private were coded

as having delivered in a facility while those who delivered either in their home or any

other home were categorized as having delivered at home. For taking their infants for

immunization the first DPT was used. Those who reported not to have taken them

remained as no, those who reported to have taken them were all grouped as having been

taken, and any other response such as don’t know was considered missing. All those who

reported to have attended ANC whether once or more were coded as having attended

(yes), and those who reported not to have attended were coded as no and the rest as

missing.

The dependent variable was the current use of FP. It was recoded to those using

modern FP methods and those not using. Those who used pills, IUD, injectables,

diaphragm, condom, male and female sterilization, implants, lactational amenorrhea, and

foam were classified as using a modern method while the rest were classified as not using
 79
a modern method. Binary logistic regression model was used to determine the

contribution of various characteristics in the use of FP. The assumptions for binary

logistic model were met; the dependent variable was categorical and dichotomous and

nominal in nature, the independent variables were categorical and nominal, the

observations were mutually exclusive, and any variable that had categories with cases

less than 15 was not included in the model.

Results

Adolescents

To answer the first research question regarding the extent to which intrapersonal,

interpersonal, community, and organizational factors are associated with the uptake of FP

among adolescents in Kenya, first baseline sociodemographic and other descriptive

characteristics of the adolescents were analyzed. Second, the bivariate analysis using chi-

square test was done. Finally, the significant factors that met the assumptions of the

logistic regression model was done.

Baseline Descriptive and Demographic Characteristics of the Adolescents

The total number of respondents among women of reproductive age was 31,079;

out of these 6,078 (19.6%) were adolescents, and the postpartum women were 4,338

(14.0 %) with 495 being adolescents; thus, 8.2% of the adolescents were in the

postpartum period.

The Rift valley region had the highest number of adolescents (27.9%) followed by

Eastern at 107%. Nairobi region has the lowest with 2.1%. Most of the adolescents

(85.8%) had never been in any marital union compared to only 13.4% of the older
 80
women who had never been in a marital union. Eight hundred and sixty-four (864) of the

adolescents reported to have been in union, and over 2,000 (36.7%) had had sexual

encounters. More than half of the adolescents (68%) lived in the rural areas compared to

61.3% of the older women. More than half of the adolescents 57.6% had primary level

education, and an even higher percentage of older women (65.2%) have a similar level.

More adolescents (40.8%) had secondary level education compared to the older women

(24.5%), but only 1.6% went beyond secondary school compared to 10.4% of the older

women. The proportion of adolescents in the poor wealth quintile is higher than that of

the older women, whereas the proportion of the older women in the rich wealth quintile

are higher than that of the adolescents. A higher percentage of the adolescents (85.3%)

were able to read compared to 69.8% of the older women. The sociodemographic

characteristics of the adolescents and those of the older women are shown in Table 2.

The mean age at first birth for all the respondents was 19.42 years, with that of the

adolescents being 16.43 years and that of the older women being 19.54 years. The mean

age at first birth for the postpartum women was 19.42 years 19.37 and 19.44 years for the

other women. Four-point five percent of the adolescents were pregnant during the survey

time compared to 7.3% of the older women. The mean age of adolescents at first birth is

16.4 years, mean age at first sex is 14.5 years and mean age at first cohabitation is 16.1

years. The mean age of older women at first birth is 19.5 year mean age at first sex ix

15.4 years and their mean age at first cohabitation is 19.1 years.
 81
Table 2

Sociodemographic Characteristics of the Respondents

Variable Adolescents Older women

Frequency Percent Frequency Percent
Region (n = 6078) (n = 25001)
Coast 807 13.3 3095 12.4
North 358 5.9 1306 5.2
Eastern
Eastern 1032 17.0 4215 16.9
Central 482 7.9 2632 10.5
Rift Valley 1698 27.9 7361 29.4
Western 676 11.1 2164 8.7
Nyanza 900 14.8 3354 13.4
Nairobi 125 2.1 874 3.5
Marital Status n = 6078 n = 25001
Never in 5214 85.8 3361 13.4
union
In union 788 13 18248 73
No longer 76 1.2 3392 13.5
in union
Residence n = 6078 n = 25001
Urban 32.0 38.7
Rural 68.0 61.3
Educational n = 6078 n = 25001
attainment
Primary 57.6 65.2
school and
below
Secondary 40.8 24.5
Above 1.6 10.4
secondary
Wealth Status n = 6078 n = 25001
poor 46.6 41.6
middle 21.5 18.6
rich 31.9 39.9
Literacy n = 6057 n=
24957
Cannot read 14.5 29.9
Able to read 85.3 69.8
Not 0.1 0.3
applicable
 82
On bivariate analysis, there was a statistically significant difference between the

adolescents and older women for all the sociodemographic characteristics tested as

follows: marital status, chi-square with 2 degrees of freedom equals 12814.6 p is less than

.001; the areas of residence between the adolescents and older women is statistically

significant; chi-square with 1 degrees of freedom equals 91.91 p is less than 00; chi-

square with 2 degrees of freedom equals 952.680 p is less than 001; wealth status, chi-

square with 2 degrees of freedom equals 132.2 p is less than 001; and for literacy, chi-

square with 2 degrees of freedom equals 597.01 p is less than 001.

Family Planning Use Among Adolescents and Associated Characteristics

Only 7.8% of the adolescents were found to be using modern methods of FP

compared to 42.2% of the older women. The mean age at first sex of adolescents using a

modern method was 16.58 years, and those not using a modern method was 16.18 years.

The most popular modern method among adolescents and older women was the

injectable. Some methods such as sterilization and the lactational amenorrhea were not

being used by adolescents. Table 3 shows the pattern of use, and Table 4 shows the

current use by the specific method.

Table 3

The Pattern of Family Planning by the Adolescents and Older Women

Adolescents Older women

Percent (n =6,078 Percent (n = 25,002
Currently using 8.6 46.0
Used since last birth 49.7 33.9
Used before last birth 1.3 5.3
Never used 40.5 14.7
 83
On the intention to use, 63.2% of the adolescents reported that they intended to

use some FP later compared to 49.9% of the older women. Twelve percent of the

adolescents were not sure whether they would use in the future, and 24.2% did not intend

to use FP in the future. The injectables were the preferred future method by most of the

adolescents (50% of those on any method), followed by condoms (23.3% of users).

Among the older women, 45.7% reported the injectable to be their preferred future

method, followed by implants.

Table 4

Current Use of Family Planning by Adolescents and Older Women by Specific Method

Method Adolescents Older Women

Frequency percent Frequency percent
Not using 5558 91.4 13489 54.0
Pill 24 .4 1415 5.7
IUD 2 .0 612 2.4
Injections 259 4.3 5257 21.0
Condom 124 2.0 662 2.6
Periodic abstinence 38 .6 775 3.1
Withdrawal 7 .1 136 .5
Other 1 .0 50 .2
Implants/Norplant 60 1.0 1909 7.6
Female condom 4 .1 7 .0
Other modern method 1 .0 14 .1
Female sterilization 0 0 652 2.6
Male sterilization 0 0 3 .0
Lactational amenorrhea (LAM) 0 0 20 .1
Total 6078 100.0 25001 100.0

Discontinuation of Family Planning Among Adolescents

Respondents were asked on the last method of FP they had discontinued within

the last 5 years of the survey. Among the adolescents, 85.9% had discontinued a modern

method with more than half discontinuing the injection and 22.2% discontinuing the pill.

Table 5 shows discontinuation rates by method, and Table 6 shows the reasons for
 84
discontinuation. The reasons for discontinuation were classified into two: those

associated with the method (side effects, health concerns, need for a more effective

method and inconvenience to use) and the rest. Among the adolescents, 22.7% had

discontinued the use of FP due to method related reasons. The association between

discontinuation of a modern method and method related reasons was not statistically

significant; chi-square equals 1.60 with 1 p equals .206.

Table 5

Family Planning Methods Discontinuation by Adolescents and Older Women by Specific

Method

Adolescents Older women

Method Percentage n = 149 n = 4330
Pill 12.8 22.2
Injections 27.5 52.9
Condom 40.3 6.4
Periodic abstinence 12.1 7.7
Withdrawal 2.0 1.7
Norplant 4.0 5.4
IUD - 2.2
Other modern method 1.3 0.3
0thers - 1.2

Table 6

Reason for Discontinuation of Family Planning Methods Discontinuation by Adolescents

and Older Women by Specific Method

Reason Adolescents Older Women

Percent n = 132 Percent n = 4057
Became pregnant 9.8 6.5
Wanted to become pregnant 24.2 30.1
Husband disapproved 5.3 1.2
Side effects 15.2 30.4
Access, availability 1.5 1.1
Wanted more effective method 4.5 8.0
Inconvenient to use 3.0 2.6
Infrequent sex, husband away 29.5 8.5
Others 6.9 6.5
 85
Reasons for not Using Family Planning Methods

Various reasons were given for not using FP. At the intrapersonal level, a few of

the respondents’ themselves reported being opposed to using FP methods; only 1.1% of

the adolescents and 4.4% of the older women. Fear of side effects and health concerns

were also cited by 5.6% of the adolescents and 20.8% of older women. Methods being

inconvenient to use was given by 0.2 of the adolescents and 0.9% of the older women.

In comparison, fear of interfering with the body’s processes was cited by 1.7% of

adolescents and 2.9% of older women. Not knowing of a method was cited by 1.5% of

adolescents and 2.4% of older women, while not knowing any source of the method was

given to not use FP by 0.8% of the adolescents and 1.0% of older women. (how can I do

cross tabs to see if there is any significant difference between the adolescents and older

women)

At the interpersonal level, various reasons were given for not using FP methods.

Husbands or partner’s opposition was cited by 0.9% of the adolescents and 4.0% of the

older women. At the community level, an equal percentage (0.1%) of adolescents and

older women indicated that they did not use FP methods due to opposition from other

people and a similar proportion between both groups cited religious prohibition as the

reasons. At the organizational level, various reasons were given for not using FP; lack of

access or the services being too far, 0.1% of the adolescents and 0.3% of the older

women; high cost, 0.2% of the adolescents and 0.8% of the older women; unavailability

of the preferred method, 0.1% of the adolescents and 0.3% of older. Table 7 summarizes

the various reasons given by adolescents for not using FP.

 86
Table 7

Reasons for not Using Family Planning Methods by Adolescents

Variable Adolescents Older Women

Frequency Percent Frequency Percent
Respondent opposed No 1389 98.9 3444 95.6
Yes 16 1.1 160 4.4
husband/partner opposed No 1393 99.1 3461 96.0
Yes 12 .9 143 4.0
others opposed No 1403 99.9 3600 99.9
Yes 2 .1 4 .1
religious prohibition No 1391 99.0 3487 96.8
Yes 14 1.0 117 3.2
knows no method No 1384 98.5 3516 97.6
Yes 21 1.5 88 2.4
knows no source No 1394 99.2 3567 99.0
Yes 11 .8 37 1.0
fear of side effects/health No 1326 94.4 2856 79.2
concerns Yes 79 5.6 748 20.8
lack of access/too far No 1404 99.9 3592 99.7
Yes 1 .1 12 .3
costs too much No 1402 99.8 3576 99.2
Yes 3 .2 28 .8
interferes with body No 1402 99.8 3570 99.1
Yes 3 .2 34 .9
inconvenient to use No 1381 98.3 3499 97.1
Yes 24 1.7 105 2.9
preferred method not No 1404 99.9 3592 99.7
available Yes 1 .1 12 .3
 87
Information on Family Planning for Those Who Obtained Family Planning Methods

For the respondents who took an FP method, information was given to them on

FP. Among the adolescents, 47.7% of them were told about side effects compared to

54.3% of the older women. Three-point eight percent of the adolescents were told about

side effects by health or FP worker adolescents compared to 11.8% of the older women.

Information on any other method (besides the one they took) was given to 59.6% of the

adolescents compared to 69.6% of the older women. Health or FP worker told 15.3% of

adolescents about other FP methods compared to 26.1% of the older women. See Table 8

for a summary of these results.

Table 8

Information on Family Planning given to Adolescents who Obtained Family Planning

Methods

Variable Adolescents Older Women

Frequency Percent Frequency Percent
Told about side effects No 79 52.3 1793 45.7
Yes 72 47.7 2128 54.3
Told about side effects by No 76 96.2 1584 88.2
health or FP worker Yes 3 3.8 211 11.8
Told how to deal with side No 11 15 287 14.7
effects Yes 64 85 2051 85.3
Told about other FP No 61 40.4 1194 30.4
methods Yes 90 59.6 2731 69.6
Told about other FP No 50 84.7 870 73.9
methods by health or FP Yes 9 15.3 308 26.1
worker

Association of Various Factors with the Uptake of Family Planning Among

Adolescents in Kenya

Various sociodemographic characteristics were tested for association with the use

of FP using the chi-square test. For literacy, 95.5%. of the adolescents using a modern
 88
method can read compared to 91.8% of those not using. The association between being

able to read and using a modern method is statistically significant; X2 (1, N = 6049) =

8.32, p = .004. Of the adolescents using a modern method of FP, 12.9 % had education

attainment of secondary education and above compared to 8.6% of those not using a

modern method. The association between having a secondary education level education

and higher and use of modern method is statistically significant; X 2 (1, N = 31079) =

9.69, p = .002. Of those using modern methods, 58.2% were in the not poor category

compared to 52% of those not using. The association between wealth status and use of

modern method was statistically significant X2 (1, N = 6078) = 4.8, P =.028. of those

using modern methods, 53% were in union currently or had ever been compared to 10.9%

among those not using. The association of ever having been in union and using FP is

significantly significant; X2 (1, N = 6078) = 632.65, p < .01. Of those using modern

methods, 40.3% were living in the urban areas compared to 31.3% of those not using.

The association between using a modern method and place of residence is statistically

significant; X2 (1, N = 6078) = 16.12, p < .01.

There was a statistically significant difference between the mean age of

adolescents at first birth of those using a modern method and those not using a modern

method; t 896 degrees of freedom is equal to -2.287 p equals .022, two-tailed. The mean

age at first cohabitation of adolescents using a modern method was 16.12 years and that

of those not using a modern method was 16.35 years. The difference in the mean age of

adolescents at first cohabitation of those using a modern method and those not using a
 89
modern method is not statistically different; t 586.03 degrees of freedom is equal to -.503

p equals .615, two-tailed. The results are summarized in Table 9.

 90
Table 9

Summary of Chi-Square Results for Sociodemographic Characteristics and Use of

Modern Methods of Family Planning Among Adolescents

Sociodemographic Not using a modern Using a modern Statistical results using chi-square
Characteristics method method
Wealth
Poor 2634 (47%) 198(41.8%) X2 (1, N = 6078) = 4.8 , P=.028
Not poor 3246 (53%) 276(58.2%)

Education Level
Below 5121 (91.4%) 413(81.1%) X2 (1, N = 6078) = 9.69, P=.002
Secondary
Secondary and above 483 (8.6%) 61(18.9%)

Literacy level
Cannot read 457 (8.2%) 21(4.5%) X2 (1, N = 6049) = 8.32, P=.004

Can read 5121 (91.8%) 450(95.5%)

Marital status
Never been in union 4991 (89.1%) 223(47%) X2 (1, N = 6078) = 632.65, p <
Ever been in union .01
613 (10.9%) 251(53.0%)

Place of residence
Urban 1756 (31.3%) 191(40.3%) X2 (1, N = 6078) = 16.12, p < .01

Rural 3848 (68.7%) 283(59.7%)

 91
Intrapersonal Characteristics

Ninety-four point six percent of adolescents reported knowing of a modern

method of FP while 97.4 of older women said they knew a modern method of FP. Among

the adolescents 8.2% of those who knew a modern method, were actually using one, none

of those who reported not knowing a method was using. The association between

knowing a modern method and using one is statistically significant; chi-square equals

29.23 with 1 degree of freedom p is less than 0.01.

The survey assessed the means by which the respondents were getting FP

messages over the last few months preceding the survey. They evaluated the radio, TV

and newspapers/magazines. Majority of the respondents had heard FP messages through

the radio with 59.7% of adolescents and 69.3% of older women having heard. Of the

adolescents using a modern method, 76.2% had heard FP messages through the radio.

The association between hearing the messages through the radio and using a modern

method was statistically significant; chi-square equals 25.48 with 1 degree of freedom p

is less than .001. Those who reported having heard FP messages on TV were fewer than

those hearing from the radio; 32% of the adolescents and 38.4% of the older women had

heard. Of the adolescents using modern FP method 43.8% had heard FP messages

through the TV. The association between hearing FP messages on TV and using and

using a modern method is statistically significant; chi-square equals 114.42 with 1 degree

of freedom p is less than .001. Less than a quarter of the respondents had seen FP

messages in the newspapers/magazines with 23.4% of adolescents and 23.9% of older

women having seen. Of the adolescents using modern FP methods, 31.9% had seen FP
 92
messages in the newspaper/magazine. The association between seeing FP messages in the

newspaper/magazine and using a modern method was statistically significant; chi-square

equals 19.04 with 1 degree of freedom p equals .03.

Thirty-five point seven percent (35.7%) of the adolescents reported to have seen

FP informational materials compared to 45.2% of older women. The association between

using a modern method and having seen FP informational materials is statistically

significant; chi-square equals 33.92 with 1 degree of freedom p is less than 0.01.

Nine point two percent (9.2%) of the adolescents received FP messages through

social media compared to 11.0% of the older women. The association between receiving

messages through social media and using a modern method is not statistically significant;

chi-square equals 2.84 with 1 degree of freedom p equals 0.09. Four point eight percent

of the adolescents received FP messages through the mobile via text/email compared to

6.3% of the older women. The association between receiving FP messages via text/email

is statistically significant; chi-square equals 3.86 with 1 degree of freedom p is equal to

0.05.

Slightly over half of the adolescents (52.8%) reported being the main decision

makers on FP compared to 73% among the older women. Among the adolescents using a

modern method 52.9% reported to be the main FP decision makers compared to 50%

among those not using a modern method. The association between using a modern

method and making own decision on FP is not statistically significant; chi-square equals

0.01 with 1 degree of freedom p equals 0.94.

 93
Interpersonal Level

More than half of the adolescents, 58.5% of adolescents reported that their

Husband / partner approves use of FP compared to 68.7 of the older women. Of the

adolescents using a modern method, 89% reported that their husband/partner approved

the use of FP. The association between using a modern method and having a

husband/partner who approves the use of FP is statistically significant; chi-square equals

47 with 1 degrees of freedom p is less than 0.01. Thirty-eight percent of the adolescents

reported having never talked to their husband / partner on FP compared to 30% of the

older women. Of the adolescents using the modern method, 90.1% reported of having

talked to their husbands/partners on FP. The association between using a modern method

and talking to the husband/partner on FP is statistically significant; chi-square equals 42.1

with 1 degree of freedom p is less than 0.01. Ninety-one percent of the adolescents

reported that their husband/partner knew you they were using FP compared to 90.7% of

the older women.

Community Factors

Seventeen point four percent of the adolescents heard FP messages at public

forums compared to 36.5% of the older women. Of the adolescents using modern

methods, 27.6% had heard FP messages in a public forum. The association between using

a modern method and hearing FP messages in a public forum is statistically significant;

chi-square equals 16.33 with 1 degree of freedom p is less than 0.01.

Nineteen point six percent (19.6%) of the adolescents reported having heard

political, religious, or community leaders talk favorably at bout FP compared to 32.8% of

 94
the older women. The association between having heard political, religious, or

community leaders talk favorably at bout FP is statistically significant; chi-square equals

16.85 with 1 degree of freedom p is less than 0.01.

Organization Factors

Very few of the respondents were visited by a FP worker; 4.8% of adolescents

and 7.8% of older women, in the preceding last 12 months. Of the adolescents using a

modern method of FP, 11% were visited by a FP worker. Only 3.5% of the adolescents

were visited by a health worker to discuss FP compared to 8.2% of the older women. The

association between using a modern method and having been visited by a health worker

to discuss FP is statistically significant; chi-square equals 14.19 with 1 degree of freedom

p is less than 0.01. The association between being visited by a FP worker and using a

modern method of FP is statistically significant; chi-square equals 19.23 with 1 degree of

freedom p is less than 0.01.

Eleven point three percent (11.3%) of the adolescents were told about an FP

method at the facility compared to 28.6% of the older women. Of the adolescent using a

modern method, 33.3% were told about a method at the health facility. The association

between being told of an FP method at the facility and using a modern method is

statistically significant; chi-square equals 67.56 with 1 degree of freedom p is less than

0.01. Forty-four point three percent (44.3%) of the adolescents reported to have visited

the health facility within the preceding 12 months of the interview compared to 68.7% of

the older women. Of the adolescents who were using a modern method of FP 65.7% had

visited the health facility. The association between visiting the health facility and using
 95
FP method is statistically significant; chi-square with 1 degrees of freedom equals 42.18

p is less than 001. Thirty-six point seven percent (36.7%) were asked of their FP needs

after the delivery of their last baby compared to 48.8% of the older women. Of the using

a modern method 49.6% were asked of their FP needs after delivery compared to 30.1%

of those not using a modern method. The association between using a modern method

and having been asked of the FP needs after delivery is statistically significant; chi-

square equals 14.24 with 1 degree of freedom p is less than 0.01.

Utilization of Services (Covariates)

The survey enquired on various services utilization aspects for the last delivery

within the preceding five years. Among the adolescents 65.1% delivered in a health

facility compared to 58.9% of the older women. Of the adolescents who were using a

modern method of FP, 72.3% delivered in a facility. The association between using a

modern method and having delivered in a facility is statistically significant; chi-square

equals 11.74 with 1 degree of freedom p is less than 0.01.

Five point eight percent of the adolescents had delivered by caesarean section

(C/S) compared to 7.9% among the older women. Among adolescents had delivered by

C/S 36.5% were using a modern method compared to 36.9% among those who did not

deliver by C/S. The relationship between having delivered by C/S and the use of a

modern method is not statistically significant; chi-square equals .003 with 1 degree of

freedom p equals 0.96.

Eighty-seven percent (87.6%) of the adolescents reported having taken their

infants for the first DPT immunization compared to 93.2% of the older women. Among
 96
the adolescents using a modern method, 97.8% had taken their infants for the first DPT

compared to 81.5% among those not using a modern method. The association between

using a modern method and having taken the infant for DPT immunization is statistically

significant; chi-square equals 49.61 with 1 degree of freedom p is less than 0.01.

Sixty-five percent of the adolescents reported having taken their infants for PNC

and 63.2% of the older women had taken babies for the PNC. Among the adolescents

using a modern method 71.1% had taken their infants for PNC compared to 61.9% of

those not using a modern method. The association between using a modern method and

having taken the infant for PNC is not statistically significant; chi-square equals 3.41

with 1 degree of freedom p equals 0.07.

Among the adolescents, 92.8% had attended ANC during their last pregnancy. Of

those using a modern method 96.3% had attended ANC compared to 90.7% of those not

using a modern method. The association between using a modern method and having

attended ANC is statistically significant; chi-square equals 9.73 with 1 degree of freedom

p equals 0.002. Sixty-five percent had delivered in a health facility. Of the adolescents

using a modern method 72.3% had delivered in a facility compared to 60.9% of those not

using a modern method. The association between using a modern method and having

delivered in a facility is statistically significant; chi-square equals 11.74 with 1 degree of

freedom p equals 0.001. From these results the Ho 1 that intrapersonal, interpersonal,

community, and organizational factors are not associated with the uptake of FP among

adolescents in Kenya was rejected.

 97
Extent of Association of Family Planning Use with Various Characteristics Among

Adolescents

To determine the extent of various variables in contributing to the uptake of FP

among adolescents, a binary logistic regression was performed. The logistic regression

model was statistically significant, χ2(11) = 44.02, p < .001. The model explained 15.0%

(Nagelkerke R2) of the variance in the use of FP and correctly classified 67.5% of cases.

The sensitivity was 29.8%, and the specificity was 87%. The following predictor

(independent) variables were statistically significant: the place of delivery (p = .029),

having seen FP informational materials (p = .011), having been visited by a health worker

to discuss FP (p = .029), having heard FP messages through the T.V (p = .018), and

having been asked of their FP needs after delivery (p = .003). The adolescents who

delivered at a facility had increased odds of using FP compared to those who did not, OR

= 1.78; 95% CI [1.06, 2.98]. Those who had seen FP informational materials had higher

odds of using FP than those who had not seen, OR = 1.93; 95% CI [1.16, 3.20]. Having

seen FP messages on T.V increased the odds of using FP, OR = 1.89; 95% CI [1.12,

3.21], while having been asked of one’s FP needs after delivery increased the odds of

using FP, OR = 2.04, 95% CI [1.27, 3.28]. Table 10 summarizes the significant

associations.
 98
Table 10

Extent of Association of Family Planning Use Among the Adolescents and Various
Characteristics

Characteristics 95% C.I.

Wald df Sig. Exp(B) EXP(B) Wald
Organizational level
Place of delivery 4.762 1 .029 1.778 1.060 2.983
Visited by health worker to discuss FP 4.703 1 .030 .256 .075 .877
Asked about your FP needs after 8.791 1 .003 2.044 1.274 3.278
delivery
Intrapersonal level
Seen FP informational material 6.414 1 .011 1.926 1.160 3.198
Heard FP on TV last few months 5.610 1 .018 1.894 1.116 3.211

These results indicate that among the adolescents in Kenya, factors at the

intrapersonal and organizational level of the SEM contribute most significantly to the

uptake of FP.

Postpartum Women

To answer the second research question regarding the extent to which

intrapersonal, interpersonal, community, and organizational factors are associated with

the uptake of FP among the postpartum women in Kenya, first baseline socio-

demographics and other descriptive characteristics of postpartum women were analyzed.

Second the bivariate analysis using chi-square test was done and finally the significant

factors that met the assumptions of the logistic regression model was done.

Baseline Descriptive and Demographic Characteristics of the Postpartum Women

Of all the respondents 14% (4338) of the women had delivered within one year of

the interview, thus for the purposes of this study, they are considered to be within the

postpartum period. Those who did not deliver within one year prior to the study were
 99
classified as other women. Of these postpartum women, 11.4% were adolescents while

among the other women 29.5% were adolescents. Majority of the pot-partum women

(29%) were in the age group 20 to 24 years old followed by age group 25 to 29 years

(28.1%). The postpartum women mean age was 29.28 years while that of the other

women was 26.84 years.

The highest percentage of postpartum women was from the Rift Valley region

(33.8%) followed by the Eastern region (14.7) and 66.9% were living in the rural areas.

The other women had a similar pattern with the Rift Valley region having the majority at

28.4% followed by Eastern region with 17.2% and 61.9% were from rural areas. Of the

postpartum women, 83.8% were in marital unions compared to 57.6% of the other

women. Majority (72.6%) of the postpartum women had primary level education and

below with only 6.8% having an education level of beyond secondary school. The other

women had a similar pattern with 62.2% having a primary school education level and

below and 8.9% having education attainment above secondary More than half (55.6%) of

these women were in the poor wealth quintile with 16.7% being in the middle and 27.7%

in the rich status. The other women had 40.5% in the poor wealth status, 19.5% in the

middle and 40.0% in the rich wealth status. On literacy, 62.6% of the postpartum women

could read and 74.5% of the other women were able to read. The social-demographic

characteristics of the respondents are summarized in Table 11.

The mean age of postpartum women at first birth was 19.37 years and the mean

age at first cohabitation is 18.7 years. The mean age of other women at first birth is 19.4

years and their mean age at first cohabitation is 19.1 years.

 100
The difference between the postpartum women and the other women for various

characteristics was done using the chi-square test of independence. The difference

between the two groups of women marital status was statistically significant; chi-square

equals 1122.73 with 2 degrees of freedom p is less than 0.01. The difference between the

postpartum and the other women in education attainment was statistically significant; chi-

square equals 174.92 with 2 degrees of freedom p is less than 0.01. There was a

statistically significant difference between the wealth status of the postpartum women

and that of the other women; chi-square equals 363.14 with 2 degrees of freedom p is less

than 0.01. on literacy, the difference between the two groups of women was statistically

significant; chi-square equals 270.2 with 2 degrees of freedom p is less than 0.01.
 101
Table 11

Sociodemographic Characteristics of Postpartum and Other Women

Variable Postpartum women Other women

Percent (N = 4,332) Percent (N = 26,682)
Region Coast 13.3 12.4
North Eastern 7.3 5.0
Eastern 14.7 17.2
Central 6.1 10.7
Rift Valley 33.8 28.4
Western 9.1 9.1
Nyanza 13.2 13.8
Nairobi 2.4 3.3
Marital Status Never in union 9.0 30.6
In union 83.8 57.6
No longer in union 7.1 11.8
Residence Urban 33.1 38.1
Rural 66.9 61.9
Educational Primary school and below 72.6 62.2
attainment
Secondary 20.6 28.8
Above secondary 6.8 8.9
Wealth Status Poor 55.6 40.5
middle 16.7 19.5
Rich 27.7 40.0
Literacy (n = 4,332) (n = 26,682)
Cannot read 37.2 25.2
Able to read 62.6 74.5
Not applicable .2 .3

The difference in the distribution of the postpartum women and other women by

area of residence (urban and rural) is statistically significant chi-square equals 39.21 with

1 degree of freedom p is less than 0.01. The difference in the distribution of the

postpartum women and other women through the regions is statistically significant chi-

square equals 176.88 with 7 degrees of freedom p is less than 0.01.

Use of Modern Family Planning Methods by Postpartum Women

Among the postpartum women 36.3% were using modern methods of FP during

the interview compared to 35.3% of the other women. The difference between the
 102
postpartum women and the other women in the use of modern methods is statistically

significant; chi-square equals 59.75 with 1 degree of freedom, p is less than 0.01. As

many as 62.8% reported intention of using FP later. The most used method both by the

postpartum women (22.5%) and the other women (17.0%) was the injectable followed by

pills. Only 1.1% the postpartum women were using the IUD while 6.2% of were using

implants and 0.4% were using the LAM. Over 40% reported that they had never used FP

before. Various reasons were given for not using FP by the postpartum women. Less than

one percent of the women not using FP cited cost or lack of access as the reasons for not

using.

Table 12

Pattern of use Among Postpartum and Other Women

Postpartum women Other women

Percent (n = 4,338) Percent (n = 26,741)
Currently using (all methods) 38.7 38.7
Used since last birth 32.3 37.7
Used before last birth 10.1 3.6
Never used 18.9 19.9

Table 13

Use of Family Planning Modern Methods Among Postpartum and Other Women

Method type Postpartum Other women

women
Frequency percent Frequency percent
Pill 160 10.1 1279 13.6
IUD 49 3.1 565 7.0
Injections 977 62.1 4539 48.1
Condom 69 4.4 717 7.6
Female sterilization 31 2.0 621 6.6
Male sterilization 3 0.03
Implants/Norplant 268 17 1701 18
Lactational amenorrhea (LAM) 17 1.1 3 0.03
Female condom 2 0.1 9
Total 1573 100 9437 100
 103
Methods Discontinuation

Among the postpartum women 15.7% reported having discontinued a method

within the previous five years of the survey with 87.7 % discontinuing use of a modern

method. There was a statistically significant difference between the discontinuation of

modern methods and the other methods; chi-square equals 74.402 with 1 degree of

freedom p is less than 0.01. Injectables were the most discontinued methods at 10.4%

followed by the pill at pill 4.1%. Majority of the reasons given for discontinuation was

the desire to become pregnant (46.5%) or having gotten pregnant 19.6%. Reasons for

discontinuation were grouped according to the method related ones (side effects, health

concerns, desires for a more effective method and inconvenience use) and those not

method related. Nearly a quarter of the discontinuation (24.1%) was due to method

related reasons. The association between discontinuation of a modern method and a

method related reason is statistically significant; chi-square equals 29.7 with 1 degree of

freedom and p is less than 0.001.

Table 14

Type of Method discontinued Among Postpartum Women and Other Women

Type of method discontinued Postpartum women Other women

Percent (n = 4338) Percent (n = 26741)
Modern method 15.7 11.2
Other methods 84.3 88.8

Association of Various Factors with the Uptake of Family Planning Among

Postpartum Women in Kenya

Sociodemographic Characteristics

Among the postpartum women who were using a modern method 24.7% had a
 104
secondary level of education and above compared to 13.5% who were not using a modern

method. The association between having a secondary education level and above and use

of a modern method of FP is statistically significant; X2 (1, N = 4338) = 123.36, p > .001.

The mean age of postpartum women at first birth of those using a modern method

was 19.63 years and that of those not using a modern method was 19.22 years. There was

a statistically significant difference between the mean age of postpartum women at first

birth between those using a modern method and those not using a modern method; t 4336

degrees of freedom is equal to -3.760 p is less than .01, two-tailed. The mean age at first

cohabitation of postpartum women using a modern method was 19.35 years and that of

those not using a modern method was 18.35 years. The difference between the mean age

of postpartum women at first cohabitation of those using a modern method and those not

using a modern method is statistically significant; t 3944 degrees of freedom is equal to -

8.188 p is less than 0.01, two-tailed. The chi-square results of the association of socio-

demographic characteristics and the use of FP are summarized in Table 15.

 105
Table 15

Chi-square Results of the Association of Sociodemographic Characteristics and Use of

Family Planning

Sociodemographic Not using a modern Using a modern method Results

characteristics method
Wealth
Poor 1773 (64.1%) 638(40.6%) X2 (1, N = 6078) = 4.8
Not poor 992 (35.9%) 935(59.4%) , P=.028
Education level
Below secondary 2391 (86.5%) 2391(75.3%) X2 (1, N = 4338) =
Secondary and above 374 (13.5%) 374(24.7%) 225.48, p < .01
Literacy level
Cannot read 1073(37.8%) 145(9.3%) X2 (1, N = 4338) =
Can read 1715 (62.2%) 1420(90.7%) 408.43, p < .01
Marital status
Never been in union 302 (10.9%) 90(5.7%) X2 (1, N = 4338) =
Been in union 2463 (89.1%) 1483(94.3%) 32.99, p < .01
Place of residence
Urban 781 (28.2%) 655(41.6%) X2 (1, N = 4338) =
Rural 1984 (71.8%) 918(58.4%) 81.23, p < .01

Association of Different Characteristics with Use of Family Planning Among

Postpartum Women

Intrapersonal Level

There was almost universal knowledge of a modern method with 96.4% of the

postpartum women and even among the other women with 97% of them reporting they

knew of a method. The survey enquired on who had heard/seen FP messages over the

radio, TV, or newspaper/magazine over the past few months preceding the interview.

Among the postpartum women 61.7% had heard FP messages over the radio, 28.5%

through the TV, and 17.7% through the newspaper/magazine. Of the postpartum women

using a modern FP method, 78.7% had had FP messages through the radio compared to

52.1% of those not using a modern method. There is a statistically significant association
 106
between hearing FP messages through the radio and using a modern method of FP; chi-

square equals 144.53 with 1 degree of freedom p is less than 0.01. Of the postpartum

women using modern methods, 42.1% had heard FP messages through the TV compared

to 20.7%. among those not using a modern method. There is a statistically significant

association between the use of modern methods and having heard FP messages through

the TV; chi-square equals 95.67 with 1 degree of freedom p is less than 0.01. Of the

postpartum women using a modern method 25.2% had seen FP messages through a

newspaper/magazine compared to 13.5% among those not using. The association

between having seen FP messages in the newspapers/magazine is statistically significant;

chi-square equals 32.30 with 1 degree of freedom p is less than 0.01.

Among the postpartum women, 37.6% had seen FP informational materials. Of

those using modern 51.7% had seen FP informational materials compared to 29.6% of

those not using a modern method. The association between having seen FP informational

materials and using a modern method is statistically significant; chi-square equals 100.41

with 1 degree of freedom p is less than 0.01.

Seven point eight percent of the postpartum women had received FP messages

through social media and 5.1% had received FP messages through mobile via text or

email. Of the postpartum women using a modern method 11.8% had received FP

messages through social media compared to 5.5% of those not using a modern method.

The association between receiving FP messages through social media and using a modern

method is statistically significant; chi-square equals 26.16 with 1 degree of freedom p is

less than 0.01. Among the postpartum women using modern methods 7.7% had received
 107
FP messages through mobile via text or email compared to 3.7% of those not using a

modern method. The association between receiving FP messages through the mobile via

text or email and using a modern method is statistically significant; chi-square equals

15.98 with 1 degree of freedom p is less than 0.01. Sixty-seven point three percent

(67.3%) of the postpartum women reported that they were the main decision makers on

FP. Among those using a modern method, 67.6 reported to be the main decision makers

on FP compared to 61.5% of those not using a modern method.

Interpersonal Factors

Among the postpartum women, 65.2% reported that their husbands/partners

approved the use of FP. Among those using a modern method of FP 88.7% reported that

their partners/husbands approved FP use compared to 50.6 of those not using FP. The

association between the use of a modern methods and the husband/partners’ approval is

statistically significant; chi-square equals 263.73 with 1 degree of freedom, p is less than

0.01.

Seventy-one point none of the postpartum women reported never to have talked to

their husbands/partners on FP. Of those using a modern method, 91.1% reported to have

talked to their husbands/partners on FP compared to 58.7% of those not using a modern

method. The association between the using a modern method and having talked to the

husband/partners on FP is statistically significant; chi-square equals 210.89 with 1 degree

of freedom, p is less than 0.01.

Among the postpartum women 69.4% reported to have talked to their husband

/partner on FP and 90% reported that their husband/partner knew they were using FP. Of
 108
those using FP 91.1% reported to have ever talked to their husbands/partners on FP

compared to 58.7% who were not using a modern method. The association between

having talked to the husband/partner on FP and the use of a modern method is

statistically significant; chi-square equals 210.89 with 1 degree of freedom p is less than

0.01. Of those using a modern method 89.8% reported that their husband/partner knew

they were using FP methods compared to 92.9% of those who were not using. The

association between the husband/partner knowing that the respondent was using FP and

the use of modern method is not statistically significant; chi-square equals 0.27 with 1

degree of freedom p equals 0.60.

Community Factors

Slight over a quarter (28.3%) of the postpartum women had heard FP messages in

public forums and of those using modern FP, 36.7% had heard FP messages in a public

forum compared to 28.5% who were not using a modern method. The association

between having herd FP messages in a public forum and using a modern method is

statistically significant; chi-square equals 40.72 with 1 degree of freedom p is less than

0.01

Slightly over a quarter (26.4%) of the postpartum women heard political,

religious, or community leaders talk favorably at about FP. Of those using a modern

method 35.9% had heard political, religious, or community leaders talk favorably at

about FP compared to 21% of those not using a modern method. The association between

having heard political, religious, or community leaders talk favorably at about FP and
 109
using a modern method is statistically significant; chi-square equals 55.15 with 1 degree

of freedom p is less than 0.01.

Organizational Factors

Over the last 12 months preceding the survey, 9.8% of the postpartum women

were visited by a FP worker, while 78.2% had visited a health facility and 38.2% had

been told at the health facility. Forty-three point eight percent (43.8%) of the postpartum

women were asked of their FP needs during their last delivery. Of those using a modern

method 56.8% were asked of their FP needs compared to 35.9% of those not using a

modern method. The association between having been asked of their FP needs during

their last delivery and using a modern method is statistically significant; chi-square

equals 81.64 with 1 degree of freedom, p equals 0.01.

Only 8.2% of the postpartum women were visited by a health worker to discuss

FP. Of those using modern FP 9.6% were visited by a health worker to discuss FP

compared to 5.7% of those not using a modern method. The association between being

visited by a health worker to discuss FP and using a modern method is statistically

significant; chi-square equals 14.42 with 1 degree of freedom p is less than 0.01.

Of the postpartum women using a modern method of FP, 11.0% been visited by

FP worker within the past 12 months of the survey, compared to 9.1% of those not using

a modern method. The association between having been visited by a health worker and

using a modern method is not statistically significant; chi-square equals 1.86 with 1

degree of freedom p equals 0.17. Among the postpartum women using a modern method

86.2% visited the health facility within the past year preceding the survey, compared to
 110
73.7% of those not using a modern method. There is a statistically significant association

between having visited a health facility and using a modern method; chi-square equals

44.48 with 1 p is less than 0.01. Of the postpartum women using a modern method of FP

43.9% were told about FP at the health facility compared to 34.4% of those not using.

The association between use of modern FP and having been told about FP in a health

facility is statistically significant; chi-square equals 15.13 with 1 degree of freedom p is

less than 0.01.

Utilization of Services (Covariates)

Among the postpartum women 93.4% attended ANC at least once during the last

pregnancy prior to the survey. Of those using a modern method, 97.8% had attended

ANC compared to 90.9% who were not using a modern method. The association between

having attended ANC and using a modern method is statistically significant; chi-square

equals 76.39 with 1 degree of freedom p is less than 0.01. Among the postpartum women

60.3% had delivered in a facility. Of those using a modern method, 76.1% had a facility

delivery compared to 51.2% among those not using a modern method. The association

between having delivered in a facility and using a modern method is statistically

significant; chi-square equals 256 with 1 degree of freedom p is less than 0.01. Seven

point seven percent of the postpartum women had delivered by caesarean section (C/S).

Among those using modern methods 10.3% had delivered by C/S compared to 6.3%

among those not using a modern method. There was a statistically significant relationship

between having delivered by C/S and the use of a modern method; chi-square equals

22.41 with 1 degree of freedom p is less than 0.01. Eighty-three point three percent
 111
(83.3%) of the postpartum women reported to have taken their infants for DPT at first

month after the last delivery. Of those using a modern method 98.4% had taken their

infants for DPT 1 compared to 75.2% of those not using. The association between having

taken the infant for DPT 1 and using a modern method is statistically significant; chi-

square equals 346.23 with 1 degree of freedom p is less than 0.01. Fifty-nine percent

(59%) of the postpartum women took their infants for PNC two months post-delivery. Of

those using a modern method 71% had taken their infants for PNC compared to 52.3% of

those note using a modern method. The association between having taken the infant for

PNC and using a modern method is statistically significant; chi-square equals 69.57 with

1 degree of freedom p is less than 0.01.

The association between having taken the infant for PNC and using a modern

method is not statistically significant; chi-square equals 0.206 with 1 degree of freedom p

equals 0.65. Considering these results, the Ho 2 that intrapersonal, interpersonal,

community, and organizational factors are not associated with the uptake of FP among

postpartum women in Kenya was rejected.

Extent of Association of Family Planning Use with Various Characteristics Among

the Postpartum Women

To determine the extent of association of various variables and the uptake of FP

among postpartum women, a binary logistic regression was performed. The variables

were grouped according to those that are related to seeking care and those that are not.

For the variables related to seeking care, the logistic regression model was statistically

significant, χ2(7) = 225.38, p < .001. The model explained 23.8% (Nagelkerke R2) of the
 112
variance in the use of FP and correctly classified 67.5% of cases. The sensitivity was

72.2%, and the specificity was 63.3%. The following predictor (independent) variables

were statistically significant: the place of delivery (p <.001), having taken the baby for

PNC (p = .006), Having taken the baby for immunizations at six weeks (p < .001), and

having been asked of one’s FP needs after p = .002.

The post–partum women who delivered at a facility had increased odds of using

FP compared to those who did not, OR = 3.05; 95% CI [2.304,4.040]. Those who had

taken the baby for immunizations at six weeks had higher odds of using FP than those

who had not, OR = 1.93; 95% CI [6.64, 22.70]. Having seen FP messages on T.V

increased the odds of using FP, OR = 1.89; 95% CI [1.12, 3.21], while having been asked

of one’s FP needs after delivery increased the odds of using FP, OR = 1.54, 95% CI

[1.169, 2.015].

For the other variables, the logistic regression model was statistically significant,

χ2(7) = 330.77, p < .001. This model explained 33.4% (Nagelkerke R2) of the variance in

the use of FP and correctly classified 71.5% of cases. The sensitivity was 78.1%, and the

specificity was 65.4%. The following predictor (independent) variables were statistically

significant: having seen FP messages in the T.V (p <.001), having husband’s approval (p

< .001), and having talked to the husband on FP (p =.030). Having seen FP messages on

T.V increased the odds of postpartum women using FP, OR = 1.47, 95% CI [1.05, 2.05].

Having the husband’s approval to use FP increased the odds of using, OR = 3.57, 95% CI

[2.40, 5.3] while having talked to the husband about FP increased the odds of using OR =

1.67, CI [1.051, 2.659]. According to these results among the postpartum women in
 113
Kenya, factors at the intrapersonal and inter-personal level of the SEM contribute most

significantly to the uptake of FP.

Distribution of Factors Associated with Family Planning Uptake Among

Adolescents Across the Counties in Kenya

To answer research question three, the extent to which intrapersonal,

interpersonal, community, and organizational factors associated with FP uptake among

postpartum women and adolescents differ by counties in Kenya, the distribution of the

factors that were found to be significant in question one and two was analyzed. Using the

chi-square test, the difference in this distribution was tested to determine if it was

statistically significant. The survey had grouped counties into eight regions following the

administrative demarcation that existed in the earlier constitution. This categorization

was used.

Demographic Characteristics of Adolescents Across the Regions in Kenya

Analysis was done to determine the distribution of various characteristics of

adolescents across the counties in Kenya. North eastern (N/E) is leading with the highest

proportion of adolescents being poor (60.9%), followed by the Coast region (56.9%).

Nairobi has the lowest percentage of adolescents in the poor bracket (0.8%), followed by

Central at 19.3% and Western at 39.8%. The N/E region leads with the highest proportion

of residents who cannot read (41.5%) followed by Rift Valley (8.9%). N/E has the lowest

education attainment with 77.7% of the adolescents having education level of primary

and below, followed by Coast with 70.0%. Central the lowest number of adolescents with

education attainment of primary school and below (27.8%), followed by at Nairobi 28%
 114
and Nyanza at 54.6%. N/E has the highest proportion of adolescents who cannot read

(41.5%) followed by Rift Valley at 8.9%. Central has the lowest proportion of

adolescents who cannot read (1.3%) followed by Nyanza at 2.3% and Nairobi at 2.4%.

N/E region has the lowest proportion of adolescents using modern FP methods (0.6%)

followed by Central at 6.4% and then Eastern at 7.2%. Nairobi leads with 15.2% of

adolescents using modern FP methods followed by Nyanza at 12% and Coast at 8.1%.

Table 16 summarizes the proportion of selected characteristics across the regions.

 115
Table 16

Distribution of Various Characteristics Among the Adolescent Across Regions

Characteristics Region
Co N/ E E Cent R/ V W Nya Nai Total
Wealth index
poor Frequency 459 218 523 93 866 269 403 1 2832
% 56.9 60.9 50.7 19.3 51.0 39.8 44.8 0.8 46.6
Not poor Frequency 348 140 509 389 832 407 497 124 3246
% 43.1 39.1 49.3 80.7 49.0 60.2 55.2 99.2 53.4
Marital status
Never been Frequency 650 286 922 446 1430 599 775 106 5214
in union
% 80.5 79.9 89. 92.5 84.2 88.6 86.1 84.8 85.8
In/have Frequency 157 72 110 36 268 77 125 19 864
been in
union
% 19.5 20.1 10.7 7.5 15.8 11.4 13.9 15.2 14.2
Place of residence
Urban Frequency 278 186 313 173 439 152 281 125 1947
% 34.4 52. 30.3 35.9 25.9 22.5 31.2 100. 32.0
0 0
Rural Frequency 529 172 719 309 1259 524 619 0 4131
% 65.6 48. 69.7 64.1 74.1 77.5 68.8 0.0 68.0
Literacy
Cannot Frequency 70 148 63 6 151 16 21 3 478
read
% 8.7 41. 6.1 1.3 8.9 2.4 2.3 2.4 7.9
5
Can read Frequency 732 209 966 473 1540 656 874 121 5571
% 91.3 58.5 93.9 98.7 91.1 97.6 97.7 97.6 92.1
Education level
Primary Frequency 565 279 564 134 1048 386 491 35 3502
and below
% 70.0 77.9 54.7 27.8 61.7 57.1 54.6 28.0 57.6
Secondary Frequency 242 79 468 348 650 290 409 90 2576
and above
% 30.0 22.1 45.3 72.2 38.3 42.9 45.4 72.0 42.4
Use of modern method
Not Using Frequency 742 356 958 451 1568 631 792 106 5604
% 91.9 99.4 92.8 93.6 92.3 93.3 88.0 84.8 92.2
Using Frequency 65 2 74 31 130 45 108 19 474
% 8.1 0.6 7.2 6.4 7.7 6.7 12.0 15.2 7.8
 116
The demographic characteristics and the characteristics that were significantly

associated with the use of FP were considered. The socio-demographic characteristics

tested included wealth index, level of education, marital status, and literacy. In all these

characteristics, the difference across the different regions was statistically significant as

shown in Table 17. The use of modern methods of FP across the regions was statistically

significant, X2 (7, N = 6078) = 60.88, p <.01.

Table 17

Distribution of Adolescents’ Socio-demographic Characteristics across regions Chi-

square Test Results

The Socio Demographics Characteristic Chi-square Results

Wealth index X2 (7, N = 6078) =347.34, p <.01
Marital status X2 (7, N = 6078) =65.09, p <.01
Type of residence X2 (7, N = 6078) =395.67, p <.01
Literacy X2 (7, N = 6049) =660.31, p <.01
Education level X2 (7, N = 6078) =350.60, p <.01
Use of modern method X2 (7, N = 6078) =60.88, p <.01

Various characteristics whose association with the use of modern methods of FP

that had been shown to be statistically significant at the bivariate level of analysis were

tested. The majority of the characteristics tested were shown to be statistically

significantly different across the different regions as shown in Table 18. There was

statistically significant difference in the distribution of most of the factors that are

significantly associated with uptake of FP among the adolescents.

 117
Table 18

Distribution of Adolescents’ Various Characteristics Across Regions Chi-Square Test

Results

The Characteristic Chi-square Results

heard political, religious, or community leaders talk
favorably at bout FP X2 (7, N = 2089) =91.90, p <.001

hearing FP messages in a public forum X2 (7, N = 2090) =84.36, p <.001

Seen FP informational materials X2 (7, N = 2088) =163.86, p <.001

Organizational level
having been asked of the FP needs after delivery X2 (14, N = 1951) =124.74, p <.001

ANC Attendance X2 (7, N = 4334) =188.81, =.001

Having delivered in a facility X2 (1, N = 880) =46.62, p <.001

visited by a health worker to discuss FP X2 (7, N = 2086) =56.59, p =.020

received FP messages through the mobile via X2 (7, N = 2090) =17.83, p =013
text/email
received FP messages through social media X2 (7, N = 2090) =39.06, p <.01

Seen FP informational materials X2 (7, N = 2862) =114.62, p <.01

visited the health facility X2 (7, N = 2090) =154.81, p =.01

At health facility told about FP X2 (7, N = 1643) = 46.17, p =.01

visited by a FP worker X2 (7, N = 2090) =50.18, p <.01

Heard FP messages through radio X2 (7, N = 2090) =351.97, p <.01

heard family FP messages on TV X2 (1, N = 2090) = 181.86 p <.01

Seen family FP messages on newspaper/magazine X2 (1, N = 2089) = 87.06, p <.01

 118
Distribution of Factors Associated with Family Planning Uptake Among

Postpartum Women Across the Regions in Kenya

First the distribution of various characteristics of the postpartum women was

analyzed. North eastern (N/E) is leading with the highest proportion of postpartum

women being poor (76.7%), followed by the Coast region (65.3%). Nairobi has the

lowest percentage of postpartum women in the poor bracket (1.9%), followed by Central

at 15.8% and Western at 47.7%. All the other regions have more than 50% of the

postpartum women in the poor bracket. The N/E region leads with the highest proportion

of residents who cannot read (35.3%) followed by Coast (88.0%). N/E has the lowest

education attainment with 95.9% of the postpartum having education level of primary

and below, followed by Coast with 84.2%. Nairobi the lowest number of postpartum

women with education attainment of primary school and below (36.2%), followed by

Central at 46.6% and Nyanza at 66%. N/E has the highest proportion of postpartum

women who nan not read (35.3%) followed by Coast at 88.0%. Nairobi has the lowest

proportion of postpartum women who cannot read (2.6%) followed by Central at 2.9%

and Nyanza at 8.5%. N/E region has the lowest proportion of women using modern FP

methods (3.5%) followed by R/V at 29.6% and then Coast at 32.1%. Central leads with

64.1% of postpartum women using modern FP methods followed by Western at 50% and

Nairobi at 49.5%.
 119
Table 19

Distribution of Selected Characteristics of Postpartum Women Across the Regions in

Kenya
Characteristics Region
Co N/ E E Cent R/ V W Nya Nai Total
Wealth index
poor Frequency 376 243 379 42 873 189 307 2 2411
% 65.3 76.7 59.5 15.8 59.5 47.7 53.5 1.9 55.6
Not poor Frequency 200 74 258 224 594 207 267 103 1927
% 34.7 23.3 40.5 84.2 40.5 52.3 46.5 98.1 44.4
Marital status
Never been Frequency 33 0 53 21 165 43 67 10 392
in union
% 5.7 0.0 8.3 7.9 11.2 10.9 11.7 9.5 9.0
In/have Frequency 543 317 584 245 1302 353 507 95 3946
been in
union
% 94.3 100.0 91.7 92. 88.8 89.1 88.3 90.5 91.0
1
Place of residence
Urban Frequency 205 118 186 126 418 95 183 105 1436
% 35.6 37.2 29.2 47. 28.5 24.0 31.9 100. 33.1
4 0
Rural Frequency 371 199 451 140 1049 301 391 0 2902
% 64.4 62.8 70.8 52.6 71.5 76.0 68.1 0.0 66.9
Literacy
Cannot Frequency 202 279 181 7 428 40 48 3 1188
read
% 35.3 88.0 28.4 2.6 29.2 10.2 8.5 2.9 27.5
Can read Frequency 370 38 456 258 1037 354 520 102 3135
% 64.7 12.0 71.6 97.4 70.8 89.8 91.5 97.1 72.5
Education level
Primary Frequency 485 304 482 124 1073 266 379 38 3151
and below
% 84.2 95.9 75.7 46.6 73.1 67.2 66.0 36.2 72.6
Secondary Frequency 91 13 155 142 394 130 195 67 1187
and above
% 15.8 4.1 24.3 53.4 26.9 32.8 34.0 63.8 27.4
Use of modern method
Not Using Frequency 391 306 363 94 1033 198 327 53 2765
% 67.9 96.5 57.0 35.3 70.4 50.0 57.0 50.5 63.7
Using Frequency 185 11 274 172 434 198 247 52 1573
% 32.1 3.5 43.0 64. 29.6 50.0 43.0 49.5 36.3
7
 120
Then, to determine if the distribution of the characteristics across the counties in

Kenya was statistically significant, bivariate analysis was done using the Chi-square test.

The demographic characteristics and the characteristics that were significantly associated

with the use of FP were considered. The socio-demographic characteristics tested

included wealth index, level of education, marital status and literacy. In all these

characteristics, the difference across the different regions was statistically significant as

shown in Table 20. The use of modern methods of FP across the regions was statistically

significant, X2 (7, N = 4338) =337.17, p <.01.

Table 20

Distribution of Postpartum Women’s Sociodemographic Characteristics Across Regions

Chi-Square Test Results

The Socio Demographics Characteristic Chi-square Results

Wealth index X2 (7, N = 4338) =396.14, p <.01
Marital status X2 (7, N = 4338) =55.18, p <.01
Type of residence X2 (7, N = 4338) =274.37, p <.01
Literacy X2 (7, N = 4323) =879.51, p <.01
Education level X2 (7, N = 4338) =307.55, p <.01
Use of modern method X2 (7, N = 4338) =337.17, p <.01

Various characteristics whose association with the use of modern methods of FP

that had been shown to be statistically significant at the bivariate level of analysis were

tested. The majority of the characteristics tested were shown to be statistically

significantly different across the different regions as shown in Table 21.

 121
Table 21

Distribution of Postpartum Women’s Various Characteristics Across Regions Chi-Square

Test Results

The Characteristic Chi-square Results

Intrapersonal level
received FP messages through the mobile via text/email X2 (7, N = 2090) =17.83, p =013
received FP messages through social media X2 (7, N = 2090) =39.06, p <.01
Seen FP informational materials X2 (7, N = 2088) =163.86, p <.001
Heard FP messages through radio X2 (7, N = 2090) =351.97, p <.001
Heard FP messages on TV X2 (1, N = 2090) = 181.86 p <.01
Seen family FP messages on newspaper/magazine X2 (1, N = 2089) = 87.06, p <.01
Interpersonal level
Talked to husband on FP X2 (7, N = 1735) =321.30, p <.001
Having husbands approval to use FP X2 (7, N = 1557) =144.68, p <.001
Community level
heard political, religious, or community leaders talk X2 (7, N = 2089) =91.90, p <.001
favorably at bout FP
hearing FP messages in a public forum X2 (7, N = 2090) =84.36, p <.001
Organizational level
having been asked of the FP needs after delivery X2 (14, N = 1951) =124.74, p <.001
visited by a health worker to discuss FP X2 (7, N = 2086) =56.59, p =.020
At health facility told about FP X2 (7, N = 1634) = 46.17, p =.01
visited by a FP worker X2 (7, N = 2089) =50.18, p <.01
(utilization of services) Covariates
Having taken the infant for DPT X2 (1, N = 4210) = 75.71, p <.01
Having taken the infant for PNC X2 (1, N = 2084) = 187.12, p <.01
ANC Attendance X2 (7, N = 4334) =188.81, =.001
Having delivered in a facility X2 (1, N = 4329) =341.08, p <.001
visited the health facility X2 (7, N = 2090) =154.81, p =.01

There was statistically significant difference in the distribution of most of the

factors that are significantly associated with uptake of FP among the postpartum women.

Thus the Ho 3 that factors associated with FP uptake among postpartum women and

adolescents in Kenya do not differ by counties was rejected.

Summary

Analysis was carried out to answer the three research questions. The results show

that there is statistically significant association between the use of FP among adolescents
 122
in Kenya and intrapersonal, interpersonal, community, and organizational factors. Factors

at the intrapersonal and organizational level of the SEM contribute most significant to the

uptake of F. The adolescents who had seen FP informational materials had higher odds of

using FP than those who had not seen, OR = 1.93; 95% CI [1.16, 3.20]. Having seen FP

messages on T.V increased the odds of using FP, OR = 1.89; 95% CI [1.12, 3.21], The

adolescents who delivered at a facility had increased odds of using FP compared to those

who did not, OR = 1.78; 95% CI [1.06, 2.98]. while having been asked of one’s FP needs

after delivery increased the odds of using FP, OR = 2.04, 95% CI [1.27, 3.28].

For the postpartum women in Kenya, the results showed that there was

statistically significant association in the use of FP with intrapersonal, interpersonal,

community, and organizational factors. Factors at the intrapersonal and organizational

level of the SEM contribute most significantly to the uptake of FP. The most significant

factors in the use of FP among the postpartum women were the intrapersonal and

interpersonal factors. Having seen FP messages on T.V increased the odds of postpartum

women using FP, OR = 1.47, 95% CI [1.05, 2.05]. Having the husband’s approval to use

FP increased the odds of using, OR = 3.57, 95% CI [2.40, 5.3] while having talked to the

husband about FP increased the odds of using OR = 1.67, CI [1.051, 2.659]. The results

also revealed that there is statistically significant difference in the regional distribution of

various characteristics among the adolescents and postpartum women. This study has

provided critical knowledge that will add to the knowledge base on FP particularly in

Kenya and the region. This information has the potential to improve the FP programming
 123
among the adolescents and postpartum women. It will also be useful in enhancing equity

across the regions in Kenya.

 124
Section 4: Application to Professional Practice and Implications for Social Change

Introduction

With a view to contributing to effective FP programming, this study was

conducted to examine the factors that are associated with FP uptake in Kenya among

adolescents and postpartum women and to determine the differences in these factors

across the regions. FP uptake was the main dependent factor, and various independent

factors organized according to the SEM were studied. The 2014 Kenya DHS data were

used. This study provides feedback to inform subsequent KDHS questions in FP. It forms

a baseline for comparison with future surveys on the status of factors influencing FP

uptake across the counties for postpartum women and adolescents.

Key Findings

The use of FP by adolescents is low despite 36.7% of them reporting some sexual

activity. There was statistically significant difference in the sociodemographic

characteristics between the adolescents and the postpartum women, with adolescents

having poorer parameters. The adolescents had a skewed method mix with more than half

of them relying on the injectables and only 13.1% using the long acting methods. All the

sociodemographic characteristics analyzed were significantly associated with the use of

modern methods of FP. There were statistically significant factors that were associated

with FP use in all the levels of the SEM on bivariate analysis. Further analysis using the

logistic regression model factors at the intrapersonal and organizational level of the SEM

contribute most significantly to the uptake of FP.

 125
The postpartum women accounted for 14% of all the respondents, and most

(29%) were 20 to 24 years of age. Thirty-six-point three percent were using modern FP

methods with a very skewed method mix, with 62.1% of those using modern methods

depending on injectables followed by implants at 17%. A higher proportion of the other

women were using FP methods compared to the postpartum women, and the difference in

use between the two groups was statistically significant; chi-square equals 59.75 with 1

degree of freedom, p is less than 0.01. All the sociodemographic characteristics tested

were significantly associated with use of FP. There were statistically significant factors

associated with FP use at all the levels of the SEM on bivariate analysis. However, on

further analysis using the logistic regression model factors at the intrapersonal and

interpersonal level of the SEM contributed most significantly to the uptake of FP.

There was statistically significant difference between sociodemographic

characteristics as well as the factors that were associated with use of FP among the

adolescents as well as among the postpartum women. The Northeastern region had the

lowest use of FP by both adolescents and postpartum women and had the respondents in

this region had the worst sociodemographic parameters and other factors. The Coast

region had the second worst indicators in almost all aspects. The Central and Nairobi

regions, on the other hand, had the best indicators in almost all categories.

Interpretation of the Findings

This study provides a comparison with other studies to either confirm or challenge

previous findings. The distribution of the respondents, both adolescents and postpartum

women) follows the general population distribution. The larger regions such as the Rift
 126
valley had the most respondents and the regions known to have low population such as

the Northeastern had the least numbers, and most of the respondents were from the rural

areas. This is due to the sampling being done to be representative nationally.

Adolescents

Sociodemographic Characteristics of Adolescents

There was a statistically significant difference between adolescents and older

women in all the demographic characteristics. Though there were more adolescents

whose highest level of education was secondary compared to the older women, only

1.6% went beyond secondary school compared to 10.4% of the older women. It may be

because the adolescents are still young, and some of them may still be continuing with

education compared to older women. There were more adolescents in the poor wealth

status than older women. Most adolescents are not yet employed and as noted, they might

still be in school. The finding that the adolescents have lower academic achievement and

lower wealth status is similar to what Vogel et al. (2015) reported, noting the association

between the low socioeconomic status of adolescents and intended pregnancies.

There was also a significant association between education attainment, wealth

status, area of residence, marital status, and literacy levels. These findings are similar to

what other studies have found. Ochako et al. (2017), Johnson (2017), and Jalang’o et al.

(2017) reported an association between the use of FP with being married, higher

education level, area of residence, and being employed. However, though Jalang’o et al.

reported younger age being associated with FP use, in this study, the older women were

more likely to use FP than the adolescents, which could be due to the poorer
 127
socioeconomic characteristics associated with the adolescents. Thus, the findings in this

study support the findings that adolescents may be disadvantaged in utilizing FP services,

and older women are more likely to use FP (Mutumba et al., 2018). It is therefore critical

to empower adolescents in their education and provide opportunities for income

generation. In doing so, their capacity to access and use modern methods of FP is

enhanced, thus reducing the occurrence of unintended pregnancies and the associated ill

effects.

Use of Family Planning Among Adolescents

Only 8.6% report of adolescents were using any form of FP, and 7.8% were using

modern methods in this study, which is low compared to 46% for the older women for

any method and 42.2 % for modern methods. The low use among adolescents can be

partly be attributed to the fact that many adolescents were not sexually active. However,

864 reported being in union, and over 2,000 (36.7%) had had sexual encounters, with a

mean age at first sex at 14.5 years. That means that there were still many sexually active

adolescents who were not using FP, yet it is unlikely that they were planning to start

childbearing at that age. Some adolescents reported being pregnant during the survey

with pregnancy that was not planned for. It is notable that the average age at first birth of

all respondents was 19.42 years, indicating very high levels of teenage pregnancies.

The most popular method among the adolescents were the injectable methods

followed by pills, which was also true for the older women. These results are similar to

what other studies have found that the injectables were the most popular among youths

and adolescents (Dennis et al., 2017). There has been a concern that this trend may be
 128
influenced by health workers who find it easier to administer the method compared to the

other methods (Bertrand et al., 2014). It is known that the longer acting methods are more

cost-effective and are less likely to be discontinued (Benson et al., 2017; Keesara et al.,

2018; Ochako et al., 2015). The implants and IUDs, though they require a health worker

for insertion and removal, are not user dependent, so they are effective at 99% and do not

need action around sexual intercourse making them convenient to use (WHO, 2018). In

this study, the most discontinued methods by adolescents were short-term methods

(injectables and condoms). There is need to understand why the implants and IUDs

despite their advantages are not popular among the adolescents. Besides the actual use of

FP, the method mix is an important indicator of the quality of care and is related to

informed consent which is a right of every woman.

Various factors were given for not using FP such as fear of side effects and health

concerns. Fear of side effects and health concerns was cited by both adolescents and

older women as reasons for not using FP or and for discontinuation. This is similar to

what was reported by Ajong et al. (2016), Nanvubya et al. (2015), and Woog et al.

(2015). Besides being common reasons for women and girls not using FP, these reasons

are also mentioned as factors in discontinuation. In this study, nearly a quarter of

discontinuation was due to method related reasons (side effects, desires for a more

effective method and inconvenience use), with side effects leading. This indicates a

potentially missed opportunity at the point of offering the methods by the health workers.

If appropriate counseling is done, then clients would be aware of the side effects to

expect and how to deal with them thus reducing anxiety if side effects occur. In this study
 129
less than half of the adolescents who used modern FP methods reported having been told

about side effects. This points to poor counseling and may have contributed to

discontinuation. It is critical for health workers to give information in a way that it will be

understood by the clients (Woog et al., 2015). Information should also include how other

methods work, their effectiveness and their side effects. If this information is

communicated in a way that clients understand, there would be little need for clients to

switch methods looking for a more effective method.

Adolescents also cited that not being married was a reason for not using FP.

These findings are similar to Sieverding et al. (2018) finding that unmarried adolescents

did not feel comfortable using FP; they faced social stigma due to the community’s view

of sexuality. However, the results indicated that many unmarried adolescents are sexually

active and are therefore at the risk of unplanned pregnancies. Teenage pregnancies are

associated with poor medical and socioeconomic outcomes, hence the need to address the

low use of FP methods among adolescents. Given that the adolescents in this study have

lower education attainment and lower wealth status compared to the older women, it is

critical that unplanned pregnancies do not interfere with the opportunity of the

adolescents realizing their potential.

It is also notable that more than 63.2% of adolescents intended to use FP in the

future, and 12.5% were not sure whether they would want to use or not and nearly a

quarter do not intend to use. This positive future intention to use FP as well as those not

sure is an opportunity to provide support and increase uptake of FP as these adolescents

mature. It is therefore critical to understand the barriers to FP uptake in order to address

 130
them and enable these willing future users to use FP optimally while encouraging the

undecided to make positive informed decisions. Understanding and addressing the

reasons why some do not intend to use FP will facilitate those who need FP in this group

to change their attitude to be willing to use.

Factors Associated with Use of Family Planning Among Adolescents

The various sociodemographic characteristics tested in this study were

statistically significantly associated with the use of modern methods of FP. These

characteristics included literacy, education attainment, marital status, area of residence,

wealth status, and the mean age at first birth.

Intrapersonal Characteristics

Knowledge is essential in making informed choice on FP. Women need to know

of the different methods that exist, availability, and other details that support use. Some

studies have reported that there was no association between having knowledge of FP and

the use; despite high knowledge the use was low (Durowade et al., 2017; Gbagbo &

Nkrumah, 2019). In this study, however, knowledge of a modern method was associated

with use of modern method. In the study by Gbagbo and Nkrumah (2019), the emergence

contraception was the most common method used unlike in this study where the

emergency pill is hardly used. The difference may be due to the different study

populations; this study consists of a nationally representative survey unlike the Gbagbo

and Nkrumah study that focused on college students who are more likely to have

knowledge of and access to the emergency contraception. As it would be expected, none

of those who reported not to know of a modern method was using any. These findings are
 131
similar to Shahabuddin et al. (2019), who reported association between knowledge of FP

services and use of modern methods. Not knowing about modern methods or of a source

was cited as a reason for not using by some adolescents.

The radio was a common method of respondents receiving FP messages followed

by the TV at 32% while more than half of the adolescents had heard FP messages though

the radio, the newspaper/magazines were less common as sources of FP information. All

the three methods were significantly associated with the use of modern FP methods.

Having seen informational materials on FP and having seen FP messages on TV

remained significant factors in the use of FP even after controlling for other factors. This

emphasizes the need for providing FP information and particularly in visual materials.

Other important sources of FP information were FP informational materials, the

mobile phone, and their association with using modern methods was statistically

significant whereas that of social media was not. It was notable that more of older women

received FP messages through the social media and mobile phones than adolescents. This

finding is contrary to the common belief that young people are easily reached through

technology. It is probable that not many of the adolescent own phones and have access to

internet, which could be due to the low socioeconomic status and low academic

achievement as noted in this study. This may point to inequity that contributes to poor use

of FP methods among this critical group.

Finally, being able to make the decision on whether to use the FP is expected to

significantly influence the use as reported in various studies (Shahabuddin et al., 2019;

Wegs et al., 2016). However, in this study, making the decision on FP use was not
 132
significantly associated with use of FP among the adolescents. Other factors seemed to

have had more importance.

Interpersonal Level

Most of the adolescents using modern methods of FP had talked to their partners

on FP (90.1%) and had their partners approve the use (89%). Talking to the partner on FP

and having them approve were both associated significantly with use of FP. Husband’s

opposition and opposition from others including religious prohibition was reported as a

reason for discontinuation or for not using a modern method. This finding is similar to

what Jalu et al. (2016), Nanvubya et al. (2015), and Ochako et al. (2015) reported, which

is that many women relied on their partner to make decisions on FP and may highlight

the patriarchal nature of the communities. It also highlights the need to involve the male

partners in FP counseling. But FP counseling has largely targeted women, and men rarely

get correct and suitable information (Capurchande et al., 2017).

Community Factors

The evaluated community factors in this study included hearing FP messages at

public forum and hearing political, religious, or community leaders talk favorably about

FP, both of which were associated with use of modern methods.

Organizational Factors

Many organizational factors were found to be significantly associated with use of

modern methods. They included being visited by a health worker to discuss FP, being

visited by a health worker, being told about a method at the health facility, visiting a

health facility, and being asked of their FP needs after the delivery of their last baby.
 133
Even after controlling for other factors, being visited by health worker to discuss FP and

being asked about FP needs after delivery remained significant factors in the use of FP

methods, thus showing the central role played by health workers involvement in the

provision of FP information.

The involvement of health workers in providing information on FP is critical in

the provision of informed choice for a modern method. The client needs to be aware of

the different methods, how they work, and the side effects. Additionally, any questions

they have needs to be answered satisfactorily, which contributes to dealing with the

myths and misconceptions. In this study, there were respondents who cited fear of side

effects, health concerns including interfering with body processes, and finding methods

inconvenient to use as reasons for discontinuation or for not using FP. These fears and

misconceptions can be dealt with quality counseling.

Though many of these factors were statistically significant in their association

with FP uptake, it is notable that there is need for improvement on the side of the health

system. For example, very few respondents were visited by FP health workers in the

preceding 12 months and being told about FP at the facility. The proportion was even

lower among adolescents. Though the need for spacing is clearly known, less than half of

all the respondents were asked of their FP needs after their last delivery and the

proportion was less among the adolescents. These are missed opportunities that the health

system should focus on for efficient and cost-effective dissemination of correct

information o FP.
 134
A small percentage of respondents also cited lack of access or the services being

too far as a reason for not using FP methods. While access remains a challenge in some

settings, perceptions on the effects of modern methods seem to be a greater concern for

most clients.

Utilization of Services (Covariates)

Utilization of related maternal and child health services characteristics was taken

as co-variates. The evaluated factors included attending ANC, facility deliveries, delivery

by cesarean section, taking infants for PNC, and the first DPT. These factors were

significantly associated with using modern FP methods except having delivered by C/S

and having taken the baby for PNC. Utilization of these maternal and child health

services may be related to the SEM intra-personal level factors, such as knowledge and

attitude towards the services. They could also be related to organizational factors such as

availability and quality of services, including health worker attitudes. It is also critical to

note that these maternal and child health services are usually offered at the same public

primary health facilities at no cost to the user in Kenya. In these same facilities, FP

services are also offered. That means where mothers were able to access the maternal and

child health services; it is most probable that FP services were available. Other factors

such as the quality of services, personal and community factors may come into play both

for FP utilization and other maternal and child health services.

These findings, therefore, led to the rejection of the hypothesis that intrapersonal,

interpersonal, community, and organizational factors are not associated with the uptake
 135
of FP among adolescents in Kenya. They have collaborated on other findings, providing

new insights into the factors associated with FP uptake among adolescents.

Postpartum Women

Baseline Descriptive and Demographic Characteristics of the Postpartum Women

The majority of the postpartum women were young, being in the age group 20

to24 years of age. Compared to the other women, the postpartum women had less

educational attainment, less wealth, and more of them could read, and the difference in

these characteristics was significant statistically. These findings may be related to the

younger age of postpartum women compared to older women.

There is a statistically significant association between educational attainment,

wealth status, marital status, area of residence and literacy levels, and modern methods of

FP. These results are similar to what Pasha et al., 2015 found that socio-demographic

characteristics such as education were associated with FP use. The findings are also

similar to those reported by Belay et al., 2016 that women with income and those with

higher education status are more likely to have higher decision-making power concerning

their fertility and are more likely to use FP. Ochako et al. (2015) also found that the area

of residence, marital status, and wealth status are associated with uptake of FP. The

residence area may affect the access to the FP services, including information, with those

living in the rural setting having less access due to various factors (Jalu et al., 2019). The

results of this study show that those in urban areas are more likely to use FP. The issues

of equity in FP access need to be addressed to ensure all women realize their right to

informed choice.
 136
Uptake of Family Planning Among Postpartum Women

More than half of the postpartum women were not using any FP method,

indicating a high risk of close births and possibly unmet need for spacing during this

critical period during which pregnancy leads to too close spacing (Rossier et al., 2015).

As many as 62.8% reported intention of using FP later in this study, indicating their

desire to space. In contrast, Pasha et al. (2015) reported that only 5% wanted to have a

pregnancy within the first year after delivery. Thus, though many postpartum women are

not ready for another pregnancy, they do not use FP to prevent. Some studies have

reported more than half of repeat births within an interval that was too short (Keogh et

al., 2015; Moore et al., 2015). Some studies have reported that the concept of unwanted

pregnancy is not appreciated among married women as it is seen as applying more to

unmarried adolescents and young women (Capurchande et al., 2017). Such an attitude

could contribute to the low enthusiasm for FP despite the women not desiring a

pregnancy. Therefore, it is critical to understand why such women are postponing the use

of FP and address any existing barriers.

Those using modern methods mainly use short-acting methods, which are less

cost-effective and are more likely to be discontinued. Like other studies have found the

injectables were the most commonly used (Dennis et al., 2017). Different reasons, such

as limitations in supplies and health workers' skills, provider bias, or community bias,

contribute to skewed method mix (Bertrand et al., 2014). Kriel et al. (2019) found that

though health workers reported that the best method is the one that fitted the client's

lifestyle and medical needs, they still felt that injectables were suitable for all women,
 137
thus indicating bias. Hardee et al. (2014) noted that every woman has a right to make a

voluntary informed choice on FP. Having the right method mix is a critical component of

quality FP services that support an informed choice of method and enhance FP uptake, as

reported in some studies (Amo-Adjei et al., 2017; Silumbwe et al., 2018). Though the

LAM has been recognized as an effective modern method when used by postpartum

women who are breastfeeding exclusively, are amenorrheic, and within six months of

delivery (International Household Survey Network, 2019; Singh et al., 2014). Thus, it is

available at no financial cost. Despite the advantages of LAM, only 0.4% of postpartum

women were using it in this study. It is critical to understand the barriers that lead to low

utilization of LAM. Women's challenge relying on LAM even when they are not

protected (have not met the criteria) has been reported in other studies (Rossier et al.,

2015). In this study, a few women reported using LAM even though they were not in the

postpartum period, highlighting that more within the postpartum period could be using it

when they have not met the criteria. Hence the need to ensure postpartum women fully

understand when the method protects them. It is a suitable method for women who fear

the use of FP for health reasons. It is also suitable for use as the mothers prepare for a

more effective method once they no longer meet the LAM criteria (after six months since

delivery, if periods return, or stop exclusive breastfeeding).

The majority of those who had discontinued FP's use within the preceding five

years had discontinued a short method. This finding is similar to what other studies have

found (Benson et al., 2017; Keesara et al., 2018; Ochako et al., 2015). The

discontinuation was mainly related to pregnancy (desire to get pregnant or got pregnant),
 138
but nearly a quarter of discontinuations had to do with method related reasons. The

method related issues included a desire for a more effective method and inconvenience in

use, both of which tend to be associated with short term methods. The longer-acting

methods (implants and IUDs) are less likely to be discontinued. Discontinuation is

associated with user characteristics and the need for a health worker to remove them

(Hubacher et al., 2017). Deliberate efforts to promote longer-acting methods while

respecting the choice principle have increased the uptake (Chakraborty et al., 2016;

Benson et al., 2017). It is critical to directly target effort in educating women and girls on

the advantages of the longer-term methods while providing a full method mix. Targeting

adolescents who are at the beginning of their reproductive life is essential. Adolescents

are likely to use these methods for the rest of their lives once they adopt them early in

life, thus achieving benefits for the users and the health system.

Different Characteristics with Family Planning Among Postpartum Women

Intrapersonal Factors

Among the intrapersonal factors, the knowledge of a modern method was

widespread. A good starting point since one has to know a method in order to consider

using one. However, knowledge alone does not translate to use, as seen in this study,

where knowledge of a method is almost universal, but the use is limited. Like among the

adolescents, the radio was the most familiar source of hearing FP messages, followed by

the TV. Among the postpartum women, all the sources of FP messages evaluated had a

statistically significant association with FP use. These sources included; the radio, TV,

newspaper/magazine, informational materials, socio-media, and the mobile phone. These

 139
results are similar to what Prata et al. (2016) found that being exposed to FP information

from various sources such as media improved FP use self-efficacy. However, it is notable

that the association between having received FP messages through social media and the

mobile phone was not significantly associated with FP use among adolescents. The

difference between postpartum women and adolescents could be due to limited access by

adolescents to technology. Hence, making technology related means not significant ways

of receiving FP information.

Fear of side effects and health concerns were cited as reasons for not using FP.

These reasons may indicate misconceptions, negative perceptions, and attitudes towards

FP, probably due to inadequate information. Ajong et al. (2016) and Nanvubya et al.

(2015) reported fear of side effects as a barrier to use of FP Keesara et al. (2018) and

Ochako et al. (2015) cited various misconceptions such as contraceptives leading to

malformed babies, IUDs penetrate the body, infertility, paralysis, and need for

hysterectomy among women. With such misconceptions that associate FP use with

serious poor health outcomes, it is not surprising that many women who do not wish to

get pregnant still do not use FP. Tailored interventions to understand the specific fears

and misconceptions in a community are critical. The resulting information should then be

used to inform targeted interventions to deal with the identified barriers.

Interpersonal Factors

When it comes to SEM's interpersonal level, there is a statistically significant

association between FP's use and the husband/partner; having talked to him, having his

approval, and knowing that the respondents are using FP methods. These findings agree
 140
with what was reported by Ajong et al.,2(016) Durowade et al. (2017) and Prata et al.

(2016) that discussion of FP within the couple increases the likelihood of uptake of FP.

Bwazi et al. (2014) reported a significant and positive association of the utilization of

postpartum FP services when there was the husband's approval of contraception. Having

the husband's/partner's approval and talking to them was significant even after controlling

for other factors in the regression model. The importance of partners' involvement in FP

discussions is critical. Other studies such as the one by Shahabuddin et al. (2019) found

that dependency on partners and influence by the partner on decision-making was

associated with the FP use. Jalu et al. (2019) described that some women considered it

their religious requirement to follow the husband's decision on whether to use FP. Other

studies have shown that some partners do not consider it their responsibility to engage in

FP discussions yet, they make decisions on whether their wives should use FP or not

(Kriel et al., 2019; Withers et al., 2015). These decisions may be based on fears and

misconceptions (Adanikin et al., 2017). There is, therefore, a need to continue engaging

partners in promoting the use of FP. It is also critical to empower women to engage their

partners on FP rather than the partner deciding. The women should tell the husband they

desire to use FP, and if it calls for use without involving the partner, they should do that.

It has been documented in other studies that women who can engage with their partners

are more likely to use FP (Wegs et al., 2016). Other studies have pointed out that

empowered women to decisions concerning FP use are more likely to utilize

contraceptives (Belay et al., 2016). Given the influence of partners on women's use of FP,
 141
it is critical to empower women to take the right action as concerns the partner and FP

use.

Community Factors

Hearing FP messages in a public forum and hearing a leader (political, religious,

or other community leaders) talk favorably on FP indicates community acceptance of FP

use. SEM's community level is particularly critical in the African setting; women care

and consider the community's attitude and perceptions on FP. Other studies have reported

similar findings. Shahabuddin et al. 2019 reported in their study found that a supportive

community environment was associated with FP use. Wegs et al. (2016) reported that

disapproval of FP use by the community members discourages women from using

contraception. The findings also support what David & Allan, 2018 found that involving

community gatekeepers in promoting FP in a community enhances FP use.

Organizational Level Factors

Organizational level factors are critical in the utilization of FP. Many studies such

as those by David and Allan, 2018 and Silumbwe et al. (2018) have looked at physical

financial accessibility, which they found to be significant barriers to uptake of FP.

However, in this study, few women indicated that they failed to use FP due to cost or lack

of access. The organizational level factors that were statically significant in FP's use

among the postpartum women in this study were mainly related to interaction with health

care providers. These results are similar to what was reported by Kumar et al. (2020) that

appropriate interaction between health workers and community members increased not

only knowledge of FP but also uptake. Health workers face FP provision challenges and
 142
do not feel fully supported to offer optimal FP services (Lince-Deroche et al., 2020). One

of the main challenges that the health system faces is to have the clients get to the

facilities. Health workers reaching out to the community is also limited; in this study, less

than ten percent of the respondents reported being visited by health workers. Hence when

women get to the facility, full advantage should be taken to promote their health. In this

study, 78.2% of the postpartum women and 61.6% of other women had visited a health

facility in the year preceding the survey. The proportion of the postpartum women who

visited a facility is high, but it is not surprising given that majority had facility deliveries

and attended ANC within that period. The pregnancy and postpartum period is a suitable

time to engage women on the need for contraception. Being told about FP at the facility

and being asked about FP needs after delivery were significantly associated with FP use.

These findings are in line with what Ochako et al. (2015) and Woog et al. (2015) found

that interaction with health workers influences FP uptake. The immediate postpartum

period is an appropriate time to counsel the women on FP and even offer the appropriate

methods. Given that some of the women will face different challenges in returning for FP

services, full advantage should be taken to reach the mothers who deliver in the facility

and support them in making informed FP choices. It is best to start counseling during the

pregnancy period, and the ANC provides the opportunity. Engagement during ANC gives

the client enough time to make informed decisions. Warren et al. (2010) and Bwazi et al.

(2014) reported improved FP uptake among postpartum women with close follow-up of

women and targeted provision of clear messages. Clients have reported dissatisfaction

with the interaction with service provers; they spend much time to get to see the provider,
 143
but very little time is spent in the consultation, limiting the possibility of effective

communication of FP (Ajong et al., 2016; Prata et al., 2016). Maximizing benefits for

postpartum mothers to enhance their understanding of FP during the interaction with

health care providers should be encouraged and promoted.

Another strategy employed to reach women with FP information is visiting them

at the community level. Though the visits by health workers or FP workers were low in

this study, they were significantly associated with modern methods of FP. The strategy

can embrace the provision of methods and information during the health /FP workers'

visits. The community-based distribution program has been used but to a limited scale.

These results suggest that if this strategy is scaled up, there is a possibility that uptake of

FP could improve.

Given that the areas where FP uptake is low are known, a targeted scale is

suggested. The advantage of community-based FP services is that it has the capacity to

over several barriers concurrently. Provision of correct and consistent information by

community members who understand the context enhances the dissemination of correct

knowledge and deals with myths and misconceptions. Where methods are provided,

access-related barriers are addressed as well.

Covariates (Utilization of Services)

There was a statistically significant association between the use of most maternal

and child health services evaluated and FP use. These services included ANC, facility

delivery, delivery by C/S, taking the infant for the first DPT, and taking their infants for

PNC. These factors, except delivery by C/S, were significant even after controlling for
 144
other factors through regression analysis. Other studies have found an association

between the use of various maternal and child health services and FP use. Pasha et al.,

2015 reported that facility deliveries were associated with FP use. In Ethiopia,

postpartum women who had attended ANC and delivered in a health facility were more

likely to use FP (Dagnew et al., 2020). As noted in the section on adolescents, the

maternal and child health services are offered in the same primary care facilities or within

the same department in major hospitals in Kenya and at no cost to the client. There is,

therefore, an excellent opportunity to support mothers utilizing these services to use FP.

Many mothers still do not use FP for different reasons, even when accessing these other

services. For maximum benefit of the mothers, integration should be enhanced.

Distribution of Factors Associated with Family Planning Uptake Across the Regions

in Kenya

This study shows that there is a statistically significant difference across the

different regions in the socio-demographic characteristics of the adolescents; these

characteristics include wealth index, level of education, marital status, and literacy. The

northeastern region has the worst indicators among the adolescents, followed by the coast

region with the highest number of the poor, lowest education attainment, and lowest

literacy levels. The Central province and Nairobi have the best indicators in these

characteristics. On the use of modern methods, Nairobi has the highest proportion of

adolescents using, followed by Nyanza and coat regions. The Northeastern region has the

lowest use of modern methods by adolescents at 0.6%, while all the other regions are

above 6%.
 145
A similar pattern follows the postpartum women, with the Northeastern and Coast

regions having the worst socio-demographic parameters while the Central and Nairobi

regions have the best. However, in modern methods, Central leads 64.7%, followed by

the Western region with 50%, and Nairobi at 49.5%. The northeastern region has the

lowest level of modern methods use at 3.5%, with all the other regions being above 30%.

There was a significant difference across the regions in the distribution of all the

factors associated with FP use. The differences in the various regions represent inequities

that could contribute to FP use. It is notable that the regions with low indicators, such as

educational attainment and literacy, also have low FP use. These findings are similar to

what was documented by Jalu et al. (2019) and Mutombo et al. (2014) that some areas

had low levels of correct FP knowledge with high levels of myths and misconceptions.

These findings of disparities across the regions call for more investments in lagging areas

to enhance equity.

FP's use also differed significantly across the regions, and so did the majority of

the characteristics that are significantly associated with FP use. These differences across

the regions may point to underlying inequalities that, in turn, influence the uptake of FP.

Socio-demographic characteristics such as wealth, education level, and literacy levels

significantly contribute to the utilization of FP services.

Limitations of the Study

One of the study's limitations is that it was a self-reported survey, leading to bias

in the responses. The interviewers were trained to explain to the respondents the need to
 146
be as accurate as possible since there was no victimization on the answers' bases, thus

reducing the bias. Recall bias is another limitation.

Being a secondary data analysis, it was not possible to tailor questions to capture

some aspects that the investigator would have liked to explore. There was a challenge,

particularly in capturing some aspects of intrapersonal factors. For example, while there

were several questions on the ways respondents received FP messages, there were limited

questions to capture the attitudes, perceptions, and beliefs on FP, values, and personality

traits. It was also challenging to capture the community's perceptions and attitudes in

depth though some proxy indicators were analyzed. Reasons given for discontinuation or

for not using FP were used to gain insights into the respondents' perceptions and beliefs.

Another limitation of the study is that though it is the latest DHS, it was done six

years ago. A more recent KDHS data would have been desirable to reflect Kenya's

current situation better. However, there was none. This secondary analysis is still relevant

since the factors contributing to FP use do not change very rapidly. Hence the results are

still applicable. This study's findings will be useful as a comparison of the changes that

will have taken place once the next DHS is done.

Recommendations

The KDHS provides nationally representative information; however, being a

survey, it does not provide an opportunity to test specific interventions. In this regard,

based on this study's findings and others, the following recommendations are made.

There is a need for implementation research to determine the best way to empower

adolescents to address the inequities that disadvantage them to support them make more
 147
informed choices on FP. This study has supported the findings of previous studies that

older women have better socioeconomic parameters than adolescents. Better

socioeconomic status is associated with FP's better uptake, as documented in this and

other cited studies. This study also documented that many sexually active adolescents are

not using FP and are, therefore, at high risk of unplanned pregnancies and the related

consequences. Significant investment should be made to reach these sexually active

unmarried adolescents to empower them to avoid unplanned pregnancies. The disparity

regarding technology has been identified and is a concern given the growing role of

digital health; therefore, further research among the adolescents is needed to position the

adolescents to benefit from technology as digital health interventions increase.

There is a need for further research to understand why implants and IUDs, despite

their advantages, are not popular among adolescents and postpartum women. Skewed

method mix is an essential indicator of the low quality of care and is related to informed

choice, which is every woman's right. This study has shown that the various SEM levels

are critical for FP uptake among adolescents and postpartum women. Further

implementation research is needed to shed more light on how best to target the different

levels concurrently. Research to understand why many postpartum women, though they

want to delay pregnancy, are not currently on FP and report the desire to use it later. The

existing barriers should be identified and addressed any existing barriers.

 148
Implications for Professional Practice and Social Change

Recommendations for Professional Practice

Health workers play an essential role in influencing FP uptake through their

interaction with clients. This study's findings point to the importance of positive

interaction of health workers with the clients, including provider-initiated health

information. There is an urgent need to urgently address the missed opportunities that

continue to occur in our facilities; potential beneficiaries of FP do not get information at

every contact with health workers. Some contacts such as ANC and delivery are

particularly suitable for FP information sharing and should be utilized fully. Health

workers need to target specific FP uptake barriers such as unfounded fears of side effects

and health effects of FP.

Many respondents, particularly the postpartum women, desire to use FP later,

despite their current need; there is a need for deliberate effort to reach them. There are

still missed opportunities that should be utilized to reach more clients who contact health

workers.

There is a need to continue reaching out to communities with the right

information since the community members, particularly the male partners and opinion

leaders (community leaders), are an important influence to both the adolescents and

postpartum women in FP uptake.

In line with the findings of this study (and other studies) on the contribution of

different SEM levels to FP uptake, FP program implementers and policymakers need to

target more than one level of the SEM when implementing FP programs.
 149
Kenya Demographic and Health Survey Methodological Implications

The DHS studies should consider adding more questions to capture women's

perceptions and attitudes and those of the community in general concerning FP.

Theoretical and Empirical Implications

In line with the findings of this study (and other studies) on the contribution of

different SEM levels to FP uptake, FP program implementers and policymakers need to

target more than one level of the SEM when implementing FP programs.

This study adds to the body of evidence that supports the proposition that the

interaction with the environment influences behavior, and the influence is at the different

levels described by the SEM. Interventions to improve FP uptake should take into

consideration the different levels of influence.

Positive Social Change

This study gives recommendations on areas of research that could potentially

improve the uptake of FP by adolescents. With the implementation of the

recommendations, modern FP methods are expected to increase among adolescents, thus

reducing unintended pregnancies and the associated ill effects. When adolescents avoid

unintended pregnancies, their chance of continuing with education is improved.

Education empowers adolescents to have better economic opportunities and make better

decisions in many aspects of life. Empowered adolescents have a better quality of life,

which means positive social change for adolescents and the community at large. Reduced

unintended pregnancies, particularly among adolescents, lead to improving health

outcomes for the adolescents, thus enhancing positive outcomes for the adolescents and
 150
the society at large and, hence, positive social change. Reduced unintended pregnancies

lead to economic savings, thus availing funds to be invested in the economy resulting in

positive social change. Implementation of these recommendations will enhance universal

access to FP and equity and thus contribute to achieving sustainable development goals

three (Ensure healthy lives and promote well-being for all ages) and five (Achieve gender

equality and empower all women and girls).

Conclusion

This study assessed the factors associated with the uptake of FP among

adolescents and postpartum women in Kenya. This study shows that all the SEM model

levels have a contribution to the uptake of FP. However, the contribution in each level

and across the levels varies. There is significant regional variation in FP uptake and many

of the factors associated with the uptake of FP. Therefore, it is critical to have tailored

research and program implementation interventions targeting different SEM levels to

support informed FP method choice. Special effort should be put into investing in

adolescents and the regions that are lagging in FP uptake and associated factors to ensure

equity, which is the right of every Kenyan.

 151
References

Adanikin, A. I., McGrath, N., & Padmadas, S. S. (2017). Impact of men’s perception on

family planning demand and uptake in Nigeria. Sexual & Reproductive

Healthcare, 14, 55–63. https://doi.org/10.1016/j.srhc.2017.10.002

Ahmed, S., Li, Q., Liu, L., & Tsui, A. O. (2012). Maternal deaths averted by

contraceptive use: An analysis of 172 countries. The Lancet, 380(9837), 111–125.

https://doi.org/10.1016/S0140-6736(12)60478-4

Ajong, A. B., Njotang, P. N., Yakum, M. N., Essi, M. J., Essiben, F., Eko, F. E., Kenfack,

B., & Mbu, E. R. (2016). Determinants of unmet need for family planning among

women in Urban Cameroon: A cross sectional survey in the Biyem-Assi Health

District, Yaounde. BMC Women’s Health; London, 16.

https://doi.org/10.1186/s12905-016-0283-9

Amo-Adjei, J., Mutua, M., Athero, S., Izugbara, C., & Ezeh, A. (2017). Improving family

planning services delivery and uptake: Experiences from the “Reversing the Stall

in Fertility Decline in Western Kenya Project.” BMC Research Notes, 10(1), 498.

https://doi.org/10.1186/s13104-017-2821-4

Benson, A., Calhoun, L. M., Corroon, M., Lance, P., O’Hara, R., Otsola, J., Speizer, I. S.,

& Winston, J. (2017). Longitudinal evaluation of the Tupange Urban Family

Planning Program in Kenya. International Perspectives on Sexual and

Reproductive Health; New York, 43(2), 75–87. https://doi.org/10.1363/43e4117

Bertrand, J. T., Sullivan, T. M., Knowles, E. A., Zeeshan, M. F., & Shelton, J. D. (2014).

Contraceptive method skew and shifts in method mix in low- and middle-income
 152
countries. International Perspectives on Sexual and Reproductive Health, 40(03),

144–153. https://doi.org/10.1363/4014414

Bishwajit, G., Tang, S., Yaya, S., & Feng, Z. (2017). Unmet need for contraception and

its association with unintended pregnancy in Bangladesh. BMC Pregnancy and

Childbirth, 17(1), 186. https://doi.org/10.1186/s12884-017-1379-4

Boo, S., & Froelicher, E. S. (2013). Secondary analysis of national survey datasets:

Secondary analysis of survey data. Japan Journal of Nursing Science, 10(1), 130–

135. https://doi.org/10.1111/j.1742-7924.2012.00213.x

Bronfenbrenner, U. (1977). Toward an Experimental Ecology.pdf. American

Psychologist. 32:513-31.

http://garfield.library.upenn.edu/classics1983/A1983QW37000001.pdf

Brown, W., Ahmed, S., Roche, N., Sonneveldt, E., & Darmstadt, G. L. (2015). Impact of

family planning programs in reducing high-risk births due to younger and older

maternal age, short birth intervals, and high parity. Seminars in Perinatology,

39(5), 338–344. https://doi.org/10.1053/j.semperi.2015.06.006

Bwazi, C., Maluwa, A., Chimwaza, A., & Pindani, M. (2014). Utilization of postpartum

family planning services between six and twelve months of delivery at Ntchisi

District Hospital, Malawi. Health, 06(14), 1724–1737.

https://doi.org/10.4236/health.2014.614205

Cahill, N., Sonneveldt, E., Stover, J., Weinberger, M., Williamson, J., Wei, C., Brown,

W., & Alkema, L. (2018). Modern contraceptive use, unmet need, and demand

satisfied among women of reproductive age who are married or in a union in the
 153
focus countries of the Family Planning 2020 initiative: A systematic analysis

using the family planning estimation tool. The Lancet, 391(10123), 870–882.

https://doi.org/10.1016/S0140-6736(17)33104-5

Canning, D., & Schultz, T. P. (2012). The economic consequences of reproductive health

and family planning. The Lancet, 380(9837), 165–171.

https://doi.org/10.1016/S0140-6736(12)60827-7

Capurchande, R., Coene, G., Roelens, K., & Meulemans, H. (2017). “If I have only two

children and they die who will take care of me?” -a qualitative study exploring

knowledge, attitudes and practices about family planning among Mozambican

female and male adults. BMC Women’s Health; London, 17.

https://doi.org/10.1186/s12905-017-0419-6

Chandra-Mouli, V., McCarraher, D. R., Phillips, S. J., Williamson, N. E., & Hainsworth,

G. (2014). Contraception for adolescents in low and middle income countries:

Needs, barriers, and access. Reproductive Health, 11(1), 1.

https://doi.org/10.1186/1742-4755-11-1

Choi, Y., Short Fabic, M., Hounton, S., & Koroma, D. (2015). Meeting demand for

family planning within a generation. Global Health Action, 8(1), 29734.

https://doi.org/10.3402/gha.v8.29734

CIA World Factbook. (2019). Kenya demographics profile 2019.

https://www.indexmundi.com/kenya/demographics_profile.html

Cleland, J., Bernstein, S., Faundes, A., Glasier, A., & Innis, J. (2006). Family planning:

The unfinished agenda. The Lancet, 368(9549), 1810–1827.

 154
https://doi.org/10.1016/S0140-6736(06)69480-4

Creswell, J. W., & Creswell, J. (2017). Research design qualitative, quantitative, and

mixed methods approaches: Vol. 5th Edition. SAGE Publications, Inc.

Crichton, J. (2008). Changing fortunes: Analysis of fluctuating policy space for family

planning in Kenya. Health Policy and Planning, 23(5), 339–350.

https://doi.org/10.1093/heapol/czn020

Dagnew, G.W., Asresie, M.B., Fekadu, G.A., & Gelaw. Y.M. (2020). Modern

contraceptive use and factors associated with use among postpartum women in

Ethiopia; Further analysis of the 2016 Ethiopia Demographic and Health Survey

data. BMC Public Health, 20(1), 661. https://doi.org/10.1186/s12889-020-08802-

David, S., & Allan, C. (2018). Midwives influencing community acceptance of family

planning to reduce maternal mortality rates in remote villages of Papua New

Guinea where men are the gatekeepers. Women and Birth, 31, S52.

https://doi.org/10.1016/j.wombi.2018.08.154

Dennis, M. L., Radovich, E., Wong, K. L. M., Owolabi, O., Cavallaro, F. L., Mbizvo, M.

T., Binagwaho, A., Waiswa, P., Lynch, C. A., & Benova, L. (2017). Pathways to

increased coverage: An analysis of time trends in contraceptive need and use

among adolescents and young women in Kenya, Rwanda, Tanzania, and Uganda.

Reproductive Health, 14(1), 130. https://doi.org/10.1186/s12978-017-0393-3

DHS Program. (2019). DHS model questionnaires.

https://dhsprogram.com/publications/publication-DHSQ8-DHS-Questionnaires-
 155
and-Manuals.cfm

Do, M., & Hotchkiss, D. (2013). Relationships between antenatal and postnatal care and

postpartum modern contraceptive use: Evidence from population surveys in

Kenya and Zambia. BMC Health Services Research, 13(1), 6.

https://doi.org/10.1186/1472-6963-13-6

Durowade, K. A., Omokanye, L. O., Elegbede, O. E., Adetokunbo, S., Olomofe, C.O.,

Ajiboye, A. D., Adeniyi, M. A., & Sanni T. A. (2017). Barriers to contraceptive

uptake among women of reproductive age in a semi-urban community of Ekiti

State, Southwest Nigeria. Ethiopian Journal of Health Sciences, 27(2), 121.

https://doi.org/10.4314/ejhs.v27i2.4

Frost, J. J., Sonfield, A., Zolna, M. R., & Finer, L. B. (2014). Return on investment: A

fuller assessment of the benefits and cost savings of the US publicly funded

family planning program. Milbank Quarterly, 92(4), 696–749.

https://doi.org/10.1111/1468-0009.12080

Ganatra, B., & Faundes, A. (2016). Role of birth spacing, family planning services, safe

abortion services and post-abortion care in reducing maternal mortality. Best

Practice & Research Clinical Obstetrics & Gynaecology, 36, 145–155.

https://doi.org/10.1016/j.bpobgyn.2016.07.008

Gbagbo, F. Y., & Nkrumah, J. (2019). Family planning among undergraduate university

students: A CASE study of a public university in Ghana. BMC Women’s Health;

London, 19. http://doi.org/10.1186/s12905-019-0708-3

Glasier, A., Gülmezoglu, A. M., Schmid, G. P., & Moreno, C. G. (2006). Sexual and
 156
reproductive health 1 sexual and reproductive health: A matter of life and death.

The Lancet 368, (14),1595-1607.

https://www.who.int/reproductivehealth/publications/general/lancet_1.pdf.

Gueye, A., Speizer, I. S., Corroon, M., & Okigbo, C. C. (2015). Belief in family planning

myths at the individual and Community levels and modern contraceptive use in

urban Africa. International Perspectives on Sexual and Reproductive Health,

41(04), 191–199. https://doi.org/10.1363/4119115

Hardee, K., Harris, S., Rodriguez, M., Kumar, J., Bakamjian, L., Newman, K., & Brown,

W. (2014). Achieving the goal of the London summit on family planning by

adhering to voluntary, rights-based family planning: What can we learn from past

experiences with coercion? International Perspectives on Sexual and

Reproductive Health, 40(04), 206–214. https://doi.org/10.1363/4020614

Hardee, K., Kumar, J., Newman, K., Bakamjian, L., Harris, S., Rodríguez, M., & Brown,

W. (2014). Voluntary, human rights–based family planning: A conceptual

framework. Studies in Family Planning, 45(1), 1–18.

https://doi.org/10.1111/j.1728-4465.2014.00373.x

Harrington, E. K., Dworkin, S., Withers, M., Onono, M., Kwena, Z., & Newmann, S. J.

(2016). Gendered power dynamics and women’s negotiation of family planning in

a high HIV prevalence setting: A qualitative study of couples in western Kenya.

Culture, Health & Sexuality, 18(4), 453–469.

https://doi.org/10.1080/13691058.2015.1091507

Hubacher, D., Spector, H., Monteith, C., Chen, P.-L., & Hart, C. (2017). Long-acting
 157
reversible contraceptive acceptability and unintended pregnancy among women

presenting for short-acting methods: A randomized patient preference trial.

American Journal of Obstetrics and Gynecology, 216(2), 101–109.

https://doi.org/10.1016/j.ajog.2016.08.033

Hubacher, D., & Trussell, J. (2015). A definition of modern contraceptive methods.

Contraception, 92(5), 420–421.

https://doi.org/10.1016/j.contraception.2015.08.008

ICF International. (n.d.). The Demographic and Health Surveys.

International Household Survey Network. (2019). Demographic and Health Surveys.

IHSN.docx. http://www.measuredhs.com/

Jain, A. K., & Winfrey, W. (2017). Contribution of contraceptive discontinuation to

unintended births in 36 developing countries. Studies in Family Planning, 48 (3)

269 - 278. DOI 10.1111/sifp.12023

Jalang’o, R., Thuita, F., Barasa, S. O., & Njoroge, P. (2017). Determinants of

contraceptive use among postpartum women in a county hospital in rural

KENYA. BMC Public Health, 17(1), 604. https://doi.org/10.1186/s12889-017-

4510-6

Jalu, M. T., Ahmed, A., Hashi, A., & Tekilu, A. (2019). Exploring barriers to

reproductive, maternal, child and neonatal health-seeking behaviors in Somali

region, Ethiopia. PLOS ONE, 14(3), e0212227.

https://doi.org/10.1371/journal.pone.0212227

Johnson, O. E. (2017). Determinants of modern contraceptive uptake among Nigerian

 158
women: Evidence from the national demographic and health survey. African

Journal of Reproductive Health, 21(3), 89–95.

https://doi.org/10.29063/ajrh2017/v21i3.8

Kaliyadan, F., & Kulkarni, V. (2019). Types of variables, descriptive statistics, and

sample size. Indian dermatology online journal, 10(1), 82–86.

https://doi.org/10.4103/idoj.IDOJ_468_18

Keesara, S., Juma, P. A., Harper, C. C., & Newmann, S. J. (2018). Barriers to postpartum

contraception: Differences among women based on parity and future fertility

desires. Culture, Health & Sexuality, 20(3), 247–261.

https://doi.org/10.1080/13691058.2017.1340669

Kennedy, E., Gray, N., Azzopardi, P., & Creati, M. (2011). Adolescent fertility and

family planning in East Asia and the Pacific: A review of DHS reports.

Reproductive Health, 8(1), 11. https://doi.org/10.1186/1742-4755-8-11

Kenya National Bureau of Statistics. (n.d.). Population by Sex and Age Groups—Census

report Volume 1c. https://knbs.or.ke/visualizations/?page_id=3126

Kenya National Bureau of Statistics. (2015). Kenya Demographic and Health Survey

2014.pdf. https://dhsprogram.com/pubs/pdf/fr308/fr308.pdf

Kenya National Bureau of Statistics. (2018). Basic report on well-being in Kenya: Based

on the 2015/16 Kenya Integrated Household Budget Survey (KIHBS).

https://www.google.com/search?q=Basic+Report+on+Well-

Being+in+Kenya&rlz=1C1GCEU_enKE887KE887&oq=Basic+Report+on+Well

Being+in+Kenya&aqs=chrome..69i57.396370937j0j0&sourceid=chrome&ie=UT
 159
F-8

Kenya National Bureau of statistics. (2019). Kenya population and housing census

results. https://www.knbs.or.ke/?p=5621

Keogh, S. C., Urassa, M., Kumogola, Y., Kalongoji, S., Kimaro, D., & Zaba, B. (2015).

Postpartum Contraception in Northern Tanzania: Patterns of use, relationship to

antenatal intentions, and impact of antenatal counseling. Studies in Family

Planning, 46(4), 405–422. https://doi.org/10.1111/j.1728-4465.2015.00040.x

Kissoon, N., Dugani, S., & Bhutta, Z. A. (2015). Maternal and child health: Gains, but a

long journey ahead. Canadian Medical Association Journal, 187(16), E471–

E472. https://doi.org/10.1503/cmaj.150725

Koffi, T. B., Weidert, K., Ouro Bitasse, E., Mensah, M. A. E., Emina, J., Mensah, S.,

Bongiovanni, A., & Prata, N. (2018). Engaging men in family planning:

perspectives from married men in Lomé, Togo. Global Health: Science and

Practice, 6(2), 317–329. https://doi.org/10.9745/GHSP-D-17-00471

Kriel, Y., Milford, C., Cordero, J., Suleman, F., Beksinska, M, Steyn, P., & Smit. J.A.

(2019). “Male partner influence on family planning and contraceptive use:

Perspectives from community members and healthcare providers in KwaZulu-

Natal, South Africa.” Reproductive Health 16(1):89. doi: 10.1186/s12978-019-

0749-y.

Kumar, A., Jain, A.K., Ram, F., Acharya, R., Shukla, A., Mozumdar,A., & Saggurti,N.

(2020). “Health workers’ outreach and intention to use contraceptives among

married women in India.” BMC Public Health 20(1):1041. doi: 10.1186/s12889-

 160
020-09061-1.

Lince-Deroche, N., Hendrickson, C., Moolla, A., Kgowedi, S., & Mulongo. M. (2020).

“Provider perspectives on contraceptive service delivery: Findings from a

qualitative study in Johannesburg, South Africa.” BMC Health Services Research

20(1):128. doi: 10.1186/s12913-020-4900-9.

Li, Q., Louis, T. A., Liu, L., Wang, C., & Tsui, A. O. (2019). Subnational estimation of

modern contraceptive prevalence in five sub-Saharan African countries: A

Bayesian hierarchical approach. BMC Public Health, 19(1), 216.

https://doi.org/10.1186/s12889-019-6545-3

Loaiza, E., & Liang, M. (2013). Adolescent pregnancy_UNFPA.pdf.

https://www.unfpa.org/sites/default/files/pubpdf/ADOLESCENT%20PREGNAN

CY_UNFPA.pdf

McLeroy, K. R., Bibeau, D., Steckler, A., & Glanz, K. (1988). An ecological perspective

on health promotion programs. Health Education Quarterly, 15(4), 351–377.

https://doi.org/10.1177/109019818801500401

Meiksin, R., Meekers, D., Thompson, S., Hagopian, A., & Mercer, M. A. (2015).

Domestic violence, marital control, and family planning, maternal, and birth

outcomes in Timor-Leste. Maternal and Child Health Journal, 19(6), 1338–1347.

https://doi.org/10.1007/s10995-014-1638-1

Moore, Z., Pfitzer, A., Gubin, R., Charurat, E., Elliott, L., & Croft, T. (2015). Missed

opportunities for family planning: An analysis of pregnancy risk and

contraceptive method use among postpartum women in 21 low- and middle-

 161
income countries. Contraception, 92(1), 31–39.

https://doi.org/10.1016/j.contraception.2015.03.007

Mumah, J. N., Machiyama, K., Mutua, M., Kabiru, C. W., & Cleland, J. (2015).

Contraceptive adoption, discontinuation, and switching among postpartum

women in Nairobi’s urban slums. Studies in Family Planning, 46(4), 369–386.

https://doi.org/10.1111/j.1728-4465.2015.00038.x

Mutombo, N., Bakibinga, P., Mukiira, C., & Kamande, E. (2014). Benefits of family

planning: An assessment of women’s knowledge in rural Western Kenya. BMJ

Open, 4(3), e004643. https://doi.org/10.1136/bmjopen-2013-004643

Mutumba, M., Wekesa, E., & Stephenson, R. (2018). Community influences on modern

contraceptive use among young women in low and middle-income countries: A

cross-sectional multi-country analysis. BMC Public Health, 18(1), 430.

https://doi.org/10.1186/s12889-018-5331-y

Mwaikambo, L., Speizer, I. S., Schurmann, A., Morgan, G., & Fikree, F. (2011). What

works in family planning interventions: A systematic review. Studies in Family

Planning, 42(2), 67–82. https://doi.org/10.1111/j.1728-4465.2011.00267.x

Nanvubya, A., Ssempiira, J., Mpendo, J., Ssetaala, A., Nalutaaya, A., Wambuzi, M.,

Kitandwe, P., Bagaya, B. S., Welsh, S., Asiimwe, S., Nielsen, L., Makumbi, F., &

Kiwanuka, N. (2015). Use of modern family planning methods in fishing

communities of Lake Victoria, Uganda. PLOS ONE, 10(10), e0141531.

https://doi.org/10.1371/journal.pone.0141531

National Institute of Health. (2005). Theory at a Glance.pdf. NIH Publication No. 05-
 162
3896; Second Edition.

Ndugwa, R. P., Cleland, J., Madise, N. J., Fotso, J.-C., & Zulu, E. M. (2011). Menstrual

pattern, sexual behaviors, and contraceptive use among postpartum women in

Nairobi urban slums. Journal of Urban Health, 88(S2), 341–355.

https://doi.org/10.1007/s11524-010-9452-6

Neal, J. W., & Neal, Z. P. (2013). Nested or networked? Future directions for ecological

systems theory: Social Development, n/a-n/a. https://doi.org/10.1111/sode.12018

Nove, A., Matthews, Z., Neal, S., & Camacho, A. V. (2014). Maternal mortality in

adolescents compared with women of other ages: Evidence from 144 countries.

The Lancet Global Health, 2(3), e155–e164. https://doi.org/10.1016/S2214-

109X(13)70179-7

Ochako, R., Mbondo, M., Aloo, S., Kaimenyi, S., Thompson, R., Temmerman, M., &

Kays, M. (2015). Barriers to modern contraceptive methods uptake among young

women in Kenya: A qualitative study. BMC Public Health, 15(1), 118.

https://doi.org/10.1186/s12889-015-1483-1

Ochako, R., Temmerman, M., Mbondo, M., & Askew, I. (2017). Determinants of modern

contraceptive use among sexually active men in Kenya. Reproductive Health,

14(1), 56. https://doi.org/10.1186/s12978-017-0316-3

OlaOlorun, F. M., & Hindin, M. J. (2014). Having a say matters: Influence of decision-

making power on contraceptive use among Nigerian women ages 35–49 years.

PLoS ONE, 9(6), e98702. https://doi.org/10.1371/journal.pone.0098702

Pallone, S. R., & Bergus, G. R. (2009). Fertility awareness-based methods: Another

 163
option for family planning. The Journal of the American Board of Family

Medicine, 22(2), 147–157. https://doi.org/10.3122/jabfm.2009.02.080038

Pasha, O., Goudar, S. S., Patel, A., Garces, A., Esamai, F., Chomba, E., Moore, J. L.,

Kodkany, B. S., Saleem, S., Derman, R. J., Liechty, E. A., Hibberd, P. L.,

Hambidge, K. M., Krebs, N. F., Carlo, W. A., McClure, E. M., Koso-Thomas, M.,

& Goldenberg, R. L. (2015). Postpartum contraceptive use and unmet need for

family planning in five low-income countries. Reproductive Health, 12(S2), S11.

https://doi.org/10.1186/1742-4755-12-S2-S11

Patton, G. C., Sawyer, S. M., Santelli, J. S., Ross, D. A., Afifi, R., Allen, N. B., Arora,

M., Azzopardi, P., Baldwin, W., Bonell, C., Kakuma, R., Kennedy, E., Mahon, J.,

McGovern, T., Mokdad, A. H., Patel, V., Petroni, S., Reavley, N., Taiwo, K., &

Viner, R. M. (2016). Our future: A Lancet commission on adolescent health and

wellbeing. The Lancet, 387(10036), 2423–2478. https://doi.org/10.1016/S0140-

6736(16)00579-1

Peterson, H. B., Darmstadt, G. L., & Bongaarts, J. (2013). Meeting the unmet need for

family planning: Now is the time. The Lancet, 381(9879), 1696–1699.

https://doi.org/10.1016/S0140-6736(13)60999-X

PopulationPyramid,net. (n.d.). Population Pyramids of the World from 1950 to 2100.

PopulationPyramid.Net. https://www.populationpyramid.net/kenya/2019/

Prata, N., Bell, S., Weidert, K., Nieto-Andrade, B., Carvahlo, A., & Neves, I. (2016).

Varying family planning strategies across age categories: Differences in factors

associated with current modern contraceptive use among youth and adult women
 164
in Luanda, Angola. Open Access Journal of Contraception; Macclesfield, 7, 1–9.

http://dx.doi.org.ezp.waldenulibrary.org/10.2147/OAJC.S93794

Rossier, C., Bradley, S. E. K., Ross, J., & Winfrey, W. (2015). Reassessing unmet need

for family planning in the postpartum period. Studies in Family Planning, 46(4),

355–367. https://doi.org/10.1111/j.1728-4465.2015.00037.x

Sallis JF, Owen N, Fisher EB. Ecological models of health behavior. In: Glanz K, Rimer

BK, Viswanath K, eds. (2008). Health behavior and health education. 4th ed. San

Francisco: John Wiley & Sons;465–485.

Shahabuddin, A., Delvaux, T., Nöstlinger, C., Sarker, M., Bardají, A., Sharkey, A.,

Adhikari, R., Koirala, S., Rahman, M. A., Mridha, T., Broerse, J. E. W., & De

Brouwere, V. (2019). Maternal health care-seeking behaviour of married

adolescent girls: A prospective qualitative study in Banke District, Nepal. PLOS

ONE, 14(6), e0217968. https://doi.org/10.1371/journal.pone.0217968

Sibanda, A. (2010). Reproductive change in Zimbabwe and Kenya: The role of the

proximate determinants in recent fertility trends. Social Biology 46:1-2, 82-99,

DOI: 10.1080/19485565.1999.9988989

Sieverding, Schatzkin, Shen, & Liu. (2018). Bias in contraceptive provision to young

women among private health care providers in South West Nigeria. International

Perspectives on Sexual and Reproductive Health, 44(1), 19.

https://doi.org/10.1363/44e5418

Silumbwe, A., Nkole, T., Munakampe, M. N., Milford, C., Cordero, J. P., Kriel, Y., Zulu,

J. M., & Steyn, P. S. (2018). Community and health systems barriers and enablers
 165
to family planning and contraceptive services provision and use in Kabwe

District, Zambia. BMC Health Services Research, 18(1), 390.

https://doi.org/10.1186/s12913-018-3136-4

Sindiga, I. (1985). The persistence of high fertility in Kenya. Social Science & Medicine,

20(1):71-84. doi: 10.1016/0277-9536(85)90314-4.

Singh, S., Darroch, J. E., & Ashford, L. S. (2014). Adding It Up: The Costs and Benefits

of Investing in Sexual and Reproductive Health. Guttmacher Institute and United

Nations Population Fund.

https://www.guttmacher.org/sites/default/files/report_pdf/addingitup2014.pdf

Sonalkar, S., Mody, S., Phillips, S., & Gaffield, M. E. (2013). Programmatic aspects of

postpartum family planning in developing countries: A qualitative analysis of key

informant interviews in Kenya and Ethiopia. 4. Africa Journal of Reproductive

Health 17(3): 54-56 https://www.ajol.info/index.php/ajrh/article/view/93747

Speizer, I. S., Corroon, M., Calhoun, L. M., Gueye, A., & Guilkey, D. K. (2018).

Association of men’s exposure to family planning programming and reported

discussion with partner and family planning use: The case of urban Senegal.

PLOS ONE, 13(9), e0204049. https://doi.org/10.1371/journal.pone.0204049

Stiegler, N., & Susuman, A. S. (2016). A comparative analysis of contraceptive use in

africa: Evidence from dhs. Journal of Asian and African Studies, 51(4), 416–432.

https://doi.org/10.1177/0021909614547462

Suresh, K., & Chandrashekara, S. (2012). Sample size estimation and power analysis for

clinical research studies. Journal of Human Reproductive Sciences, 5(1), 7.

 166
https://doi.org/10.4103/0974-1208.97779

United Nations. (2017). World Family Planning 2017_Highlights.

un.org/en/development/desa/population/publications/pdf/family/WFP2017_Highli

ghts.pdf

United Nations, (2013). World contraceptive patterns 2013. Department of economic and

social affairs, &population division.

https://www.un.org/en/development/desa/population/publications/family/contrace

ptive-wallchart-2013.asp

Vogel, J. P., Pileggi-Castro, C., Chandra-Mouli, V., Pileggi, V. N., Souza, J. P., Chou,

D., & Say, L. (2015). Millennium Development Goal 5 and adolescents: Looking

back, moving forward. Archives of Disease in Childhood, 100(1), S43–S47.

https://doi.org/10.1136/archdischild-2013-305514

Warren, C., Mwangi, A., Oweya, E., Kamunya, R., & Koskei, N. (2010). Safeguarding

maternal and newborn health: Improving the quality of postnatal care in Kenya.

International Journal for Quality in Health Care, 22(1), 24–30.

https://doi.org/10.1093/intqhc/mzp050

Wegs, C., Creanga, A. A., Galavotti, C., & Wamalwa, E. (2016). Community dialogue to

shift social norms and enable family planning: An evaluation of the family

planning results initiative in Kenya. PLOS ONE, 11(4), e0153907.

https://doi.org/10.1371/journal.pone.0153907

Westoff, C. F., & Cross, A. R. (2006). The stall in the fertility transition in Kenya. DHS

analytical studies 9. https://dhsprogram.com/pubs/pdf/AS9/AS9.pdf

 167
Withers, M., Dworkin, S. L., Zakaras, J. M., Onono, M., Oyier, B., Cohen, C. R., Bukusi,

E. A., Grossman, D., & Newmann, S. J. (2015). ‘Women now wear trousers’:

Men’s perceptions of family planning in the context of changing gender relations

in western Kenya. Culture, Health & Sexuality, 17(9), 1132–1146.

https://doi.org/10.1080/13691058.2015.1043144

Woog, V., Singh, S., Browne, A., & Philbin, J. (2015). Adolescent women’s need for and

use of sexual and reproductive health services in developing countries. 63.

Guttmacher Institute.

https://www.researchgate.net/profile/Susheela_Singh/publication/283212654_Ad

olescent_Women's_Need_for_and_Use_of_Sexual_and_Reproductive_Health_Se

rvices_in_Developing_Countries/links/562e2a7f08ae04c2aeb59100.pdf

World Health Organization. (2010). WHO Technical consultation on postpartum and

postnatal care. https://www.ncbi.nlm.nih.gov/books/NBK310595/

World Health Organization. (2013). Programming Strategies for postpartum family

planning.pdf.

https://apps.who.int/iris/bitstream/handle/10665/93680/9789241506496_eng.pdf?

sequence=1

World Health Organization. (2018). Family planning/Contraception.

https://www.who.int/news-room/fact-sheets/detail/family-planning-contraception

Zakiyah, N., Asselt, A. D. I. van, Roijmans, F., & Postma, M. J. (2016). Economic

evaluation of family planning interventions in low and middle income countries;

A systematic review. PLoS One; San Francisco, 11(12), e0168447.

 168
http://dx.doi.org.ezp.waldenulibrary.org/10.1371/journal.pone.0168447