‫كتيب مختصر متضمن لعناصر دورة‬

‫أساسيات البحث الطبي‬

‫‪Booklet of the course‬‬
‫‪Basics of Medical Research‬‬
‫البروفيسور إياد قنيبي و البروفيسور طالل إبراهيم‬
 ‫أهداف الدورة‬
‫‪ .1‬إعطاء فكرة عن البحث العلمي وأهميته‬

‫‪ .2‬بناء عقلية بحثية بمنهجية ناقدة‬

‫‪" .3‬كسر الرهبة" وتشجيعك على البحث العلمي‬

‫‪ .4‬مساعدتك على اختيار موضوع بحثي نافع‬
Outline of the Course
 1. Why is Medical
Research Important?
Personal Benefits
Broader Benefits
 2. Building a
Researcher's Mindset
 Cultivating a Researcher's Mindset
• This section focuses on building a systematic and
analytical thought process characteristic of
successful researchers.
• We'll explore how to approach problems
methodically within the scientific framework.
 Steps for Performing a Medical Research?
1. Identify the area of interest.
2. Identify a problem/gap of knowledge within this area.
3. Formulate a hypothesis or research question.
4. Formulate a specific aim of the research.
5. Prepare a mitigation plan.
6. Define and apply the best methodology to test the
hypothesis/answer the research question.
7. Correctly interpret the results.
8. Write & publish your research.
 Steps for Performing a Medical Research?
Each step builds on the previous one,
creating a structured approach to conducting
research.

• Exercise on research related to treatment

of Alzheimer’s Disease .
BUT !!

Not All research following

these steps is considered a
Good Research.
 FINER Criteria for Good Research
1. Feasible: can be reasonably conducted with the available time,
resources, & expertise.
2. Interesting: the research question captivates the attention of the
scientific community & possibly the public.
3. Novel: provides new insights or approaches, contributing to the
advancement of knowledge.
4. Ethical: adheres to ethical standards, respecting the dignity and
safety of participants. Even animal research.
5. Relevant: addresses an important issue, offering practical benefits
and contributing to theory & practice.
 Quizzes 1-4
For each of the following researches, mention which criterion is
missing:
1. A study in a Gulf country aimed at local health improvement,
assessing a new medication for leprosy:
A) Feasible
B) Interesting & relevant
C) Novel
D) Ethical
2. A research team in one of the universities intends to evaluate
the long-term effects of a new cardiac drug by following
patients for a period of 20 years:
A) Feasible
B) Interesting
C) Novel
D) Ethical
E) Relevant
3. A research replicating the exact aim of a previous robust
large-scale study on the effect of aspirin in reducing
cardiovascular events, without introducing new variables,
outcomes, or a distinct population group:
A) Feasible
B) Novel
C) Ethical
D) Relevant
4. A clinical trial investigating a new drug for the treatment of
liver cirrhosis without performing the preclinical toxicity
testing:
A) Feasible
B) Interesting
C) Novel
D) Ethical
E) Relevant
3. Types of Medical
Research
Types of Medical Research
 Laboratory (Basic) Research
Basic research in biomedical sciences can often be categorized into these
different experimental approaches:
1. In Silico: Research using computer simulations or computational
models to study biological processes, define targets, and suggest “hits”.
2. In Vitro: Studies conducted with cells or biological molecules outside
their normal biological context, typically in a lab dish.
3. Ex Vivo: Experiments performed on tissues or organs taken from an
organism in an external environment.
4. In Vivo "Animal": Studies conducted in living organisms, often animal
models, to observe biological processes in a natural context.
Levels of Basic Research
 1. In silico Research
- Experiments conducted via computer simulation and modeling.
- Starting with in silico studies to identify targets and hits and
proceeding to in vitro testing can be a cost-effective approach.
Examples:
 2. Research on Cell Components
Example:
- Cell membrane fragments including receptors and their signaling
pathway.
 3. Research on Whole Cells
Example:
- Studying the efficacy of chemotherapy drugs on cancer cells.
4. Research on Tissues
5. Research on Organs
6. In vivo “Animal” Research
Translational Research
Translational Research

https://tri.uams.edu/about-tri/what-is-translational-research/
 Translational Research
• Translational research refers to the process of applying findings
from basic science to enhance human health and well-being.
• It often involves translating laboratory discoveries or
observations into new therapies, diagnostics, or procedures that
can be used in clinical practice.
• Essentially, it bridges the gap between laboratory research and
patient care.
Types of Medical Research
 Epidemiological Research
• Studies the patterns, causes, and effects of diseases or
health-related events or behavior in specific populations.
 Health Services Research
• Focuses on how people access healthcare, healthcare costs, and outcomes
of healthcare services.
• Aims to identify the barriers and the most effective ways to deliver high-
quality care.
• This includes assessing the cost-effectiveness of medical treatments &
procedures.
• Aims to inform decisions that improve health care quality, accessibility, &
affordability.
 Questionnaire-Based Research
• Many epidemiological, health services research,
and clinical research involves questionnaires.
• Formulating an accurate validated questionnaire
can be the topic of a separate workshop/course.
 Genetic and Genomic Research

• Genome-Wide Association Study (GWAS): a

research approach used to identify genetic variations
associated with specific diseases or traits. It involves
scanning the genomes of many individuals to find
genetic markers that correlate with these traits.
Example:
 Genetic and Genomic Research
• Pharmacogenomics: the study of how genetic variation affects
drug effectiveness, toxicity, & pharmacokinetics.
Example:
• Gene Therapy: one example: a therapeutic gene is introduced
into the patient's body to compensate for the defective or missing
gene causing the disorder.
Example:
 Clinical Research
• We will discuss this type in more detail.
 Quiz 5
What is the type of research that involves whole bacterial cells?:
A) In vitro
B) In vivo
C) In silico
 Quiz 6
What do we call the research that investigates the
similarity between db/db mouse and type 2 diabetes in
humans to evaluate its suitability to investigate potential
T2DM drugs?
A. Genetic Research
B. Behavioral research
C. Pharmacologic Research
D. Disease Model Research
‫• بعد اطالعنا على أنواع البحث العلمي بإمكاننا أن نأخذ أمثلة‬
‫ً‬
‫على الفرضيات البحثية‪ ،‬والتي سيستخدم كل منها نوعا‬
‫ً‬
‫مختلفا من أنواع البحث العلمي‪.‬‬
‫‪• Examples of hypotheses requiring different types of‬‬
‫‪research.‬‬
‫• هدفان من الكويزات التالية‪:‬‬
‫‪ .1‬أخذ أمثلة على الفرضيات البحثية‪.‬‬
‫‪ .2‬التدريب على اختيارالنوع املناسب من األبحاث لفحص‬
‫الفرضية‪.‬‬
 Quiz 7
• Hypothesis: "Using computational modeling, the drug candidate
XYZ123 will demonstrate a higher selectivity for the dopamine D2
receptor over the D1 receptor, indicating its potential as an
antipsychotic with fewer side effects."
• What type of research do we need to test this hypothesis?:
A. In Silico
B. In vitro
C. Animal in vivo
D. Clinical
 Quiz 8
• Hypothesis: "The drug Remdesivir exhibits higher efficacy in
inhibiting the replication of SARS-CoV-2 in cultured human
lung cells compared to Hydroxychloroquine.“
• What type of research do we need to test this hypothesis?:
A. In Silico
B. In vitro
C. Animal in vivo
D. Clinical
 Quiz 9
• Hypothesis: "Continuous low-dose exposure to Bisphenol A (BPA)
in drinking water will lead to significant cognitive impairment in
adult rats, suggesting potential neurotoxic effects in humans at
environmentally relevant exposure levels.
• What type of research do we need to test this hypothesis?:
A. In vitro
B. Ex vivo
C. Animal-disease model
D. Animal- Toxicology
 Quiz 10
• Hypothesis: "Patients with chronic migraine receiving Botox
injections will report greater pain relief compared to those
receiving standard saline placebo injections.“
• What type of research do we need to test this hypothesis?:
A. Epidemiological
B. Genetic
C. Public health
D. Clinical
 Quiz 10
• Hypothesis: "Patients with chronic migraine receiving Botox
injections will report greater pain relief compared to those
receiving standard saline placebo injections.“
• What type of research do we need to test this hypothesis?:
A. Epidemiological
B. Genetic
C. Public health
D. Clinical
Outline of the Course
4. Research Hypothesis,
Aim, Objectives, &
Outcome Measures
 Hypothesis, Aim, Objectives, & Outcome Measures
- A research hypothesis is a testable prediction about the
relationship between variables. It forms the basis for your
research to either prove or disprove.
- The aim is the main goal directed towards testing the
hypothesis.
- The objectives are specific steps to achieve this goal.
- Outcome measures are specific criteria used to objectively
assess the results of the study in order to prove or disprove the
hypothesis. They are derived from the objectives.
- There are Primary & Secondary Outcome Measures.
‫مقاييس النتائج‬
 Example:
Hypothesis: Drug X is better than standard treatment for diabetic
patients.
Aim: To compare drug X with standard treatment in diabetic
patients.
Objectives:
1. Compare blood glucose levels (measured by HbA1c levels) for
drug X and standard treatment groups. Primary
2. Compare side effects for both groups.
Secondary
3. Compare patient adherence for both groups.
 Identifying Primary & Secondary Outcomes
Example: a study on arthritis treatment,
 Quiz 11
• In a study investigating the effects of a new drug on cancer cell
growth in vitro, which of the following is considered an
objective for this study?
A) Determine the drug's efficacy in reducing cancer cell
proliferation via cell viability assays.
B) Evaluate the potential of the drug for clinical trials.
C) Assess the drug's toxicity profile in human cells.
D) Identify the molecular mechanisms by which the drug acts.
 ‫كل ما ذكرناه ينبغي توظيفه لنفع‬
‫اإلنسان‪ ،‬وهذا يأتي بنا إلى املوضوع‬
‫التالي‪:‬‬
‫‪Evidence-Based Medicine‬‬
‫حيث أن اختيارنا لعالج معين يجب أن‬
‫ً‬
‫يكون مبنيا على الدليل العلمي‪ ،‬وهذا‬
‫ً‬
‫الدليل نأخذه عموما من البحث العلمي‪.‬‬
5. Evidence-Based
Medicine
What is Evidence-Based Medicine?
EBM Model
 How to formulate a clinical research question using
the PICOT framework
• The PICOT format helps you formulate a structured
focused answerable question.
• Has 5 essential components:
P: Patient, population, problem
I: Intervention, exposure, test
C: Comparison
O: Clinical outcome
T: Time
 How to formulate a clinical research question using
the PICOT framework

Patient or
Intervention Comparison Outcome Time
Problem
Description of Could include: Relevant most often Clinical outcome The duration of the
the patient or the • Therapy when looking at of interest to you intervention or
target disorder of • Exposure therapy questions and your patient exposure and the
interest follow-up period for
• Diagnostic test
measuring outcomes
• Prognostic factor
 Example:
PICOT-question: In adults with type 2 diabetes, does drug X,
compared to standard treatment, lead to better glycemic control
(measured by HbA1c levels), fewer side effects, more adherence,
and better quality of life over a 12-month period?
 ‫هذا السؤال سيكون بمثابة مغناطيس‬
‫تجذب به ما يعنيك من دراسات وسط‬
‫الكم الهائل من الدراسات املنشورة‪،‬‬
‫فيمنعك أن تتشتت‪،‬و يبقيك ‪.Focused‬‬
Examples of PICOT Questions
 Quiz 12
Which of the following research questions is not PICOT-formulated:
A. In adults with type 2 diabetes, does insulin glargine 100 units/mL, once
daily, compared to NPH insulin, twice daily, result in better glycemic
control (measured by HbA1c levels) over a 6-month period?
B. Does a 6-month 30 min daily aerobic exercise program improve
cardiovascular health in middle-aged men compared to no exercise?
C. Does administering a fixed-dose combination of Lisinopril 20 mg &
Hydrochlorothiazide 12.5 mg orally once daily, compared to Amlodipine
Besylate 10 mg orally once daily, lead to a greater reduction in systolic &
diastolic BP measurements over a 6-month monitoring period ?
The Hierarchy of Evidence
Outline of the Course
 6. Choosing Most
Suitable Study Design
 How are the studies in the hierarchy of evidence performed?
• Basic/highly variable.
• Examples in the upcoming videos.
• In vivo studies in late stages share design similarities with
clinical trials.
 Clinical Research
A branch of medical science that determines the safety and
effectiveness of:
• Medications
• Devices
• Diagnostic products and
• Treatment regimens
... which are intended for human use.

→ Clinical Research involves human beings.

Randomized Controlled Trials (RCTs)

• Not absolute but stratified

randomization

• Controlled due to
inclusion of control group.

• Control group can be

taking placebo or another
treatment.
 Randomized Controlled Trials (RCTs)

Good outcome
Treatment

Bad outcome
Identify study Random
subjects allocation
Good outcome
Control

Bad outcome
 Cohort Study

Good outcome
Factor present

Bad outcome
Classify
Identify study
according to
subjects
prognostic factor
Good outcome
Factor absent

Bad outcome
 Case-Control Study

Exposed
Case
(Disease present)
Unexposed
Classify according Identify study
to presence or subjects
Exposed absence of the case
Control
(Disease absent)
Unexposed
Cross-Sectional Study

No Risk factor Risk Factor

Disease Disease
Selected
Sample
No Risk Factor Risk Factor
No Disease No Disease
 Comparisons
• Which is better? Cohort or Case-Control?
• Limitations of Case-Control:
• Establishing causality: RCT >> Cohort > Case-Control > Cross-sectional
 Case Report
- An article that describes and interprets an individual case, often
written in the form of a detailed story.
- Case reports often describe:
• New side effect of a drug.
• New potential use of a drug.
• Unique case that show an important variation of a disease or
condition.
• Unique case that cannot be explained by known diseases or
syndromes.
• Examples.
 Case series
• Unlike case reports that detail a single case, case series describe
characteristics and outcomes of multiple cases.
• Examples.
• What is the difference then from a case-control?
 Narrative Reviews
• A narrative review is a type of article that provides an
overview of a topic, summarizing and integrating
information from various studies to provide a cohesive
understanding of this topic.
• Without systematic search criteria or statistical analysis.
• It offers a subjective summary and discussion based on the
author's expertise and selection of relevant studies.
Systematic Reviews
• Include systematic search criteria such as:
1. Inclusion criteria:
2. Exclusion criteria:
3. Time frame for publication of articles:
4. Specific medical databases to be searched:
5. Keywords and medical subject headings (MeSH) used in the search:

Examples.
 Meta-analysis
• In meta-analysis, several selected studies are systematically
combined to develop a single conclusion that has greater
statistical power.
• In other words, Data are pooled as if they are one data and then a
statistical analysis is made.
• This conclusion is stronger than the analysis of any single study,
due to increased numbers of subjects, greater diversity among
subjects, or accumulated effects and results.
➔ More robustness, reliability, and generalizability of the conclusion
than individual study results.
 What do you think?
• For a meta-analysis of 25 studies comparing GLP-1 agonists with
DPP-4 inhibitors for type 2 diabetes, the statistical significance
(measured as p-value) of this meta-analysis should be stronger
than that of the individual studies. This statement is:
A) True
B) False
 Clinical Practice Guidelines
• A statement produced by a panel of experts that outlines
current best practice to inform health care professionals in
making clinical decisions. The statement is produced after an
extensive review of the literature and is made by
professional associations, government agencies, and/or
private organizations.
• Good guidelines clearly define the topic; evaluate &
summarize the best evidence regarding prevention,
diagnosis, prognosis, therapy, harm, and cost-effectiveness.
• Practice guidelines should be reviewed frequently and
updated as necessary for continued accuracy and relevancy.
Primary vs. Secondary Sources
• Cohort studies, case-control studies, randomized controlled
trials, case series, and case reports are primary research
sources, as they collect original data directly.
• Meta-analyses, systematic reviews, and clinical practice
guidelines are secondary sources, as they synthesize and
analyze data from primary sources.
 Which type to choose?
Many factors:
1. Research Question and Objectives
2. Ethical Considerations
3. Time and Duration
4. Population and Sample Size
5. Generalizability of Results
6. Statistical Considerations
7. Funding and Resource Allocation
8. Feasibility and Practicality
 Which type to choose?
• By carefully considering several factors, you can choose the
most appropriate study design for your research objectives
and constraints. This holistic approach ensures that the
research is not only scientifically robust but also ethically
sound and practically feasible.
• The natural progression may start with a case report then
cross-sectional then higher level studies.
 Quiz 13
Who should typically be blinded in an RCT?
A. The research subjects
B. The researcher who collects data
C. Both
 Quiz 14
• You want to evaluate the effect of infection of diabetic foot
ulcer with S. aureus on the rate of ulcer healing. Which of
the following is the best type of research that you can
perform?
A. RCT
B. Cohort
C. Case-control
D. Cross-sectional
 Quiz 15
• Which of the following is an interventional research
“not just observational”?
A. RCT
B. Cohort
C. Case-Control
D. All of the above
E. A + B
 Quiz 16
• You want to evaluate the effect of a new anticancer drug.
Which of the following is the best type of research to be done?:
A. RCT where the control group receives placebo.
B. RCT where the control group receives the standard treatment &
the treatment group receives the new drug only.
C. RCT where the control group receives the standard treatment. &
the treatment group receives the new drug + standard treatment.
D. Cohort study.
 Quiz 17
• You wish to investigate the effects of prenatal exposure to
venlafaxine on birth weight. Which study design would be
most appropriate?
A. RCT
B. Cohort
C. Case-control
D. Cross-sectional
E. Case report
 Quiz 18
• A rare genetic disorder is believed to be more prevalent in a
particular ethnicity. The condition has an incidence of 1 in
100,000 births. Which study design is most appropriate to
identify risk factors associated with this disorder?
A. RCT
B. Cohort
C. Case-control
D. Cross-sectional
E. Case report
 Quiz 19
• Which study design would be most appropriate to evaluate the
association between smoking in pregnant women and the risk
of early-onset cancers in their newborns?
A. RCT
B. Cohort
C. Case-control
D. Cross-sectional
E. Case report
 Quiz 20
• Given the constraints of limited time and a small patient
population, but with sufficient funding, which study design
would be most effective for comparing the efficacy of a new
anticoagulant in preventing stroke in patients with atrial
fibrillation?
A. Randomized Controlled Trial (RCT)
B. Cohort Study
C. Case-Control Study
D. Cross-Sectional Study
E. Case Report
 Quiz 21
• A research team aims to assess the current prevalence of
hypertension and its correlation with the use of non-steroidal
anti-inflammatory drugs (NSAIDs) in the adult population of a
small city. Which study design would be most suitable for
capturing this information at a single point in time??
A. Randomized Controlled Trial (RCT)
B. Cohort Study
C. Case-Control Study
D. Cross-Sectional Study
E. Case Report
Revisiting the Hierarchy of Evidence
 Lessons
• Critical thinking
• Appraisal/Evaluation of evidence
• Don’t be deceived by studies lower in the
hierarchy of evidence.
‫• العلوم الشرعية في مطلع العهد اإلسالمي‬
Outline of the Course
7. How Can I Get a
Research Idea
 How to find a Research Idea?
• A skilled healthcare professional with critical thinking skills
often encounters daily phenomena that naturally spark
research ideas.
• These ideas don't require searching; they emerge as a
response to clinical observations and patient interactions.
• Pattern recognition/Ex.: ptients given a drug suffer from fever
or skin rash
‫ تبحث ملاذا‬،‫• تبقى متيقظا حاضر الذهن بحيث تالحظ املشكلة‬
.‫حصلت وكيف تصلحها‬
 An examples from my researches
Wound Healing Research
 Mephenytoin Ethotoin Mephobarbital Phenobarbital
O
F3 C
O O
Me O OMMe
e O OM e
NH O O NH H O H O
(C H 2 ) 4 N N
N C Me O NH S OM e
O H O
N S NH S C NH(C H 2 ) 4 N C (C H 2 ) 4 N N
O H O NH NH
OM e S NH Me O C (C H 2 ) 4 N
O O OO O NH
F3 C F 3 CF3C
O O
O O
F3 C
RS 102221 O

vinclozolin
Rebeccamycin
 Examples on how to get research ideas
1. Repurposing Adverse Effects:
2. Validating Traditional Remedies:
3. Analyzing Local Manifestations of Global Issues:
4. Focusing on Regional Pharmacogenomics:
5. Pharmacovigilance Research:
6. Solving controversy:

Examples

Collaboration !
8. Can I be Part of
Great Discoveries?
 Doable contributions
• Gene therapy research is feasible on a smaller scale, even with
limited resources, opening doors for life-changing projects.
• In silico bioinformatics research & basic molecular biology
techniques like cloning, PCR, Western blot, and ELISA are
doable in simple labs and form the foundation of many
research projects.
• In bioinformatics you define targets & predict the outcomes of
gene therapy.
• Example: Predicting Gene Therapy Efficacy for a Specific
Disease.
Predicting Gene Therapy Efficacy for a Specific Disease
1. Collect Genetic Data:

2. Use Bioinformatics Tools:

3. Simulation of Gene Therapy:

4. Analyze Outcomes:

5. Refine Strategy:

Examples
 Doable contributions
- This approach is cost-effective and can provide vital insights into the
potential success of gene therapies before actual lab experiments.
Then in the lab:
1. Cell Culture Models:

2. Codon optimization:

3. Gene truncation:
 9. Conducting a
Literature Search
 Why is it Important to Conduct Literature Search?

1. Enhancing Research-Thinking Skills:

2. Improved Research Paper Quality:

3. Field-Specific Research Techniques:

4. Identification of Knowledge Gaps:

5. Avoiding Duplication of Work:

 Steps for Conducting an Effective Literature Search
1. Define Your Research Question:

2. Identify Keywords and Terms:

3. Select Appropriate Databases:

- PubMed:

- PubMed Clinical Queries:

4. Develop Search Strategy:

 Steps for Conducting an Effective Literature Search
5. Execute Search and Refine Results:

6. Review and Select Relevant Literature:

7. Document Your Search:

8. Evaluate, summarize and Integrate Findings:

Example
Note:
AI/ChatGPT: helps you even in choosing the keywords and search terms.
 10. Developing a
Research Proposal
 The main components of a research proposal
1. Title: A clear and concise statement of the project.
2. Abstract: A brief summary of the research's aims and objectives.
3. Introduction: Background information & the problem statement.
4. Literature Review: An overview of existing research and how this
project will contribute to the field.
5. Research Objectives/Questions: Specific goals or questions the
research aims to address.
6. Methodology: Detailed plan of the research design, data
collection, and analysis methods.
7. Implications: The potential impact or significance of the research.
8. Timeline: Estimated schedule for the project's completion.
9. Budget: Projection of the financial requirements.
10. References: List of cited literature and sources.
Outline of the Course
 11. Basics of
Statistical Analysis
 Sample Size

Placebo

Drug A Drug B

FBG
 Sample Size

Mean Placebo FBG

Mean Drug B FBG

FBG
Mean Drug A FBG
 Sample Size

Mean Placebo FBG

Mean Drug B FBG

FBG
 Sample Size

Mean Placebo FBG

Mean Drug B FBG
Mean Drug A FBG
FBG

Outlier
 The Concept of Standard Deviation

FBG
 The Concept of Standard Deviation
Therefore, the question would have been more
meaningful if it had included the standard deviation (SD)
for each group.
“Two studies were run to compare the efficacy of two
different antidiabetics. Drug "A" reduced fasting blood
glucose (FBG) by 30% (± 25%) compared with placebo.
Drug "B" reduced FBG by 18% (± 3%) compared with
placebo. Which drug is more effective?”
 The Importance of Statistical Analysis

Understanding statistical concepts provides the tools to

correctly gather, analyze, and interpret data, ensuring
reliable and valid conclusions .
The study's design, implementation, and analysis must be
sound and unbiased, allowing for a valid interpretation
of the results, whether they support H1 or not.
Outline for Basics of Statistical Analysis
 1. Types of data/variables
- The first thing you have to do when you start analyzing data, is ask yourself 'what
kind of variable am I dealing with? The type of the variable of interest will dictate the
form of statistical analysis that you perform. Data (Variables)

Categorical
Qualitative Numerical
Quantitative

Nominal, incl. binary Ordinal

Discrete Continuous
(unordered) (ordered)
eg. Blood group, Yes or No eg. Disease stage eg. no. of fits eg. FBG
 2. Null Hypothesis vs. Alternative hypothesis
- Null Hypothesis: no difference/effect. Any observed
difference is by chance.
- Alternative hypothesis: there is difference/effect.
 3. Descriptive Statistics
- Definition: Summarizes and describes the main features of a
data set.
-Includes: Mean, median, mode, standard deviation.
-Example: Average fasting blood glucose readings in a group of
patients.
3. Descriptive Statistics

Mean - 2 SD Mean Mean + 2 SD

 4. Inferential Statistics
- Purpose: Makes inferences about populations using data about
sample drawn from this population.
- "Inferential Statistics" is named so because it involves inferring or
generalizing conclusions from a sample to the broader population.
- Key Concepts: Sampling, hypothesis testing, confidence intervals.
- Example: Estimating the average efficacy of a new drug based on a
sample of patients.
5. Probability
 6. Confidence Interval
- Definition: Range of values within which a population parameter is
estimated to lie.

• A 95% confidence interval means that if we were to take 100

different samples and compute a confidence interval for each
sample, then approximately 95 of the 100 confidence intervals will
contain the true population mean.

- Importance: Indicates the reliability of an estimate.

 6. Confidence Interval
• To apply confidence intervals (CIs) in a study comparing a
hypoglycemic drug with a placebo group, follow these steps:
1. Collect Data: Gather FBG from participants in both the
hypoglycemic drug group and the placebo group.

2. Calculate Means: Determine the mean (average) blood glucose level

for each group.
 6. Confidence Interval
3. Determine the Standard Error (SE): Calculate the SE for each group.
SE is the standard deviation divided by the square root of the sample size.

4. Calculate CIs: For a 95% CI, typically you'd use the formula: Mean ±
(1.96 × SE). This gives you the range within which you can be 95%
confident that the true mean lies.

5. Compare CIs: If the CIs of the two groups do not overlap, it suggests a
statistically significant difference between the drug and placebo.

Example
 No Overlap

Placebo

Drug

FBG
The Effect of Sample Size on Confidence Interval

Control
Control
Treatment
Treatment
 7. T-test
• A t-test compares means of two groups to assess significant differences.
• Example: Comparing blood glucose levels of a hypoglycemic drug group
and a placebo group.
• Requires knowing means, standard deviations, and sample sizes.
• Independent t-test used for distinct, unpaired groups.
• Formula calculation yields t-value; p-value indicates significance.
• Assume the following:
• Drug Group: Mean = 85 mg/dL, Standard Deviation (SD) = 5, Sample
Size (n) = 30.
• Placebo Group: Mean = 100 mg/dL, SD = 5, n = 30.
 7. T-test
• In our example, a statistically significant t-value leading to a low p-
value would indicate that the difference in FBG between the
hypoglycemic drug group and the placebo group is unlikely to be due
to chance.

• If we were to measure the blood glucose levels of the same

participants before and after taking the medication, then we would
use a dependent or paired t-test. This test compares the before-
and-after measurements within the same individuals and relies on a
different formula than the independent t-test.
 8. P-value
• The p-value is not exclusive to the t-test; it can be calculated for many
other statistical tests, including ANOVA, ANCOVA, Mann-Whitney U
test, and more. It is a measure of the strength of evidence against the
null hypothesis provided by the data in a wide variety of statistical
analyses.
• If P = 0.05, it means 5% probability of getting this result or a more extreme one
just by chance.
• If there is no difference between treatments (H0), what is the probability of finding
this or a more extreme result?
• Be aware that in research, we frequently encounter P-values which are
insignificant, but you still see a very clear trend in the data that you cannot ignore.
 Exercise on Using GraphPad Online

• https://www.graphpad.com/quickcalcs/

• https://www.graphpad.com/quickcalcs/ttest1/
9. Normal vs. Skewed Distribution
 10. Parametric vs. Non-Parametric Analyses
• T-test: Unpaired & Paired.

• Welch's t-test: Use when you cannot assume the variances of the
two groups are equal.

• Mann-Whitney U test: Use when your data are not normally

distributed or when the data are ordinal
 11. ANOVA (Analysis of Variance)
• Purpose: Tests differences between the means of three or
more groups.
• Parametric.
• Example
• Kruskal-Wallis test is the nonparametric equivalent (for
skewed data).
 12. Correlation:
• Measures the strength and direction of a relationship
between two variables.
• It is typically represented by a correlation coefficient
(ranging from -1 to 1).
• It is very important to notice that correlation
does not indicate causation.
• Examples: Level of
- Correlation between age and blood pressure. Anxiety
- Sleep duration and stress level.
- Screen time and level of anxiety.
Screen Time
 13. Regression Analysis
• A statistical method that serves two purposes:
1. It quantifies the relationship between independent and dependent
variables, indicating how much the dependent variable is expected to
change with a one-unit change in the independent variable.
2. It allows for prediction of the dependent variable based on the
independent variable, within the range of the data.
➔ Regression analysis often implies a cause-and-effect relationship
• It measures R-squared (R²) value.
• Example
13. Regression Analysis
 14. Sample Size Calculation
• To make sure that your research is not underpowered
• To calculate the sample size for research on continuous data, you'll
need:
1. Effect size:
2. Standard deviation:
3. Power:
4. Significance level:
• Use these values in a software calculator designed for sample size
determination. This will give you the number of participants needed
to detect an effect, if one exists, with desired power and significance.
 15. Statistical significance vs. Clinical Significance
• Statistical significance does not always equate to clinical significance, and
vice versa.
Examples:
• Statistical Significance Without Clinical Significance: A large study might
find a statistically significant reduction in blood pressure from a new drug.
• Clinical Significance Without Statistical Significance: A small pilot study
on a new cancer treatment might show a substantial increase in survival time
for patients.
• In essence, statistical significance refers to the likelihood that a result is not
due to chance, whereas clinical significance pertains to the practical
importance or real-world relevance of a finding.
 Quiz 22
You conducted a study to predict the effectiveness of a new
cholesterol-lowering drug based on different dosages. Which
statistical test should she use?
A) Correlation analysis
B) Paired t-test
C) Regression analysis
D) ANOVA
 Quiz 23
You want to examine the relationship between hours of sleep and
blood glucose levels in diabetic patients. Which statistical test is
most suitable?
A) Regression analysis
B) Correlation analysis
C) Paired t-test
D) ANOVA
 Quiz 24
You are investigating the impact of a new antihypertensive
medication by measuring patients' blood pressure before and
after the treatment. What is the appropriate statistical test for this
study?
A) ANOVA
B) Correlation analysis
C) Regression analysis
D) Paired t-test
 Quiz 25
You are comparing the efficacy of three different anti-
inflammatory drugs on reducing pain in arthritic patients. Which
statistical method should be used for the analysis?
A) Paired t-test
B) Regression analysis
C) Correlation analysis
D) ANOVA
 Quiz 26
If a study shows a new drug's effect on FBG with an SD of ±5,
what does this suggest about the data's distribution?
A) High variability
B) Low variability
C) Skewed distribution
D) No variability
 Quiz 27
In comparing two antidiabetics, which statistical test would
allow for the consideration of additional covariates like patient
age?
A) Independent t-test
B) ANCOVA
C) Chi-square test
D) Mann-Whitney U test
 Quiz 28
You compared two pain-relieving drugs in two separate groups;
in each group you measured the reduction in pain severity after
administering the drug. You got the results in the figure. Which
statistical test should be used to evaluate the difference in
effectiveness among the two drugs? 100
90
A) Independent t-test 80
70
B) ANCOVA 60
No. of 5040
C) Welch’s Test Patients 30
D) Mann-Whitney U test 20
10

0 1 2 3 4 5 6 7 8 9 10
Decrease in Pain
 ‫يمكنك االشتراك في هذه الدورة‬
‫القيمة من خالل الصفحة املخصصة‬
:MedOne Academy ‫لها في‬
https://www.medone.academy/
course/medical-research-rec
‫شكرا لكم على املتابعة‪ ،‬ونرجو‬
‫لكم التوفيق في البحث العلمي‪.‬‬

‫أساسيات البحث الطبي‬

‫البروفيسور إياد قنيبي و البروفيسور طالل إبراهيم‬