Safety Science 50 (2012) 1355–1367

Contents lists available at SciVerse ScienceDirect

Safety Science
journal homepage: www.elsevier.com/locate/ssci

Review

Occupational injury and accident research: A comprehensive review

Vivek V. Khanzode a, J. Maiti b,⇑, P.K. Ray b
a
KIT’s College of Engineering, Kolhapur, India
b
Indian Institute of Technology, Kharagpur, India

a r t i c l e i n f o a b s t r a c t

Article history: In this paper, a comprehensive review of the concepts of occupational injury and accident causation and
Received 17 April 2011 prevention is presented. Starting with hazard identification, the issues on risk assessment, accident
Received in revised form 28 November 2011 causation, and intervention strategies are discussed progressively. The distinctiveness and overlaps in
Accepted 12 December 2011
accident and injury research are highlighted. Both empirical research in terms of hypotheses tested
Available online 10 February 2012
and theoretical research such as accident causation models are compared and contrasted. Finally, based
on the critical appraisal of the comprehensive review, future research directions on occupational injury
Keywords:
research are delineated.
Occupational accident
Occupational injury
Ó 2011 Elsevier Ltd. All rights reserved.
Hazard identification
Risk assessment
Accident causation
Injury mechanism
Injury intervention

Contents

1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1356
2. Taxonomy of literature . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1356
3. Methods of hazard identification . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1356
4. Methods of risk assessment. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1357
4.1. Methods based on criterion variables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1358
4.2. Methods based on modeling techniques . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1358
5. Accident causation theories . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1358
5.1. First generation: Accident proneness . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1359
5.1.1. Pure chance hypothesis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1359
5.1.2. Accident proneness theories . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1359
5.1.3. Unconscious motivation theory . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1359
5.1.4. Adjustment-stress and goal-freedom-alertness theory . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1360
5.2. Second generation: Domino theories . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1360
5.3. Third generation: Injury epidemiology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1360
5.4. Fourth generation: System models . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1360
5.4.1. Sociotechnical and macroergonomic models . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1360
5.5. Factors affecting injury . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1361
5.5.1. Individual-related factors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1361
5.5.2. Job-related factors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1361
5.5.3. Organization-related factors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1361
6. Injury mechanism models . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1362
6.1. Deviation models. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1362
6.2. Energy models . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1362
6.3. Injury severity models . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1362

⇑ Corresponding author. Tel.: +91 3222 283750; fax: +91 3222 282272.
E-mail address: [email protected] (J. Maiti).

0925-7535/$ - see front matter Ó 2011 Elsevier Ltd. All rights reserved.
doi:10.1016/j.ssci.2011.12.015
1356 V.V. Khanzode et al. / Safety Science 50 (2012) 1355–1367

7. Intervention strategies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1362

7.1. Engineering interventions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1362
7.2. Behavioral interventions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1363
7.3. Enforcement interventions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1363
8. Critical appraisal and future research directions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1363
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1364

1. Introduction comprehensive review is threefold. First, we aim to develop a uni-

fied framework for classification of injury and accident literature.
An injury (to a human being) is defined as wound or trauma; In this process, the important dimensions of the unified framework
harm or hurt; or damage inflicted on the body of the injured by are explored by navigating through the vast body of knowledge
an external force (Webster, 2002). Injury is a (suspected) bodily le- available. Second, we aim to identify a number of contemporary is-
sion resulting from acute overexposure to energy interacting with sues related to injury research. In doing this, a critical appraisal of
the body in amounts or rates that exceed the threshold of physio- the injury and accident literature is presented. Third, we aim to form
logical tolerance (ICECI, 2004). Production activity deals with dif- a basis for developing a comprehensive methodology for modeling
ferent kinds of forces and energy interactions in a controlled injury risk in worksystems, which may provide some directions
environment. On occasions, this control is lost, and an uncontrolled for the future research. The paper is organized as follows: The taxon-
energy transfer occurs, leading to an accident and/or injury inci- omy of the literature is presented in Section 2. In Section 3, literature
dent. Accident and injury are undesirable outcomes of a production on hazard identification is reviewed and related issues are dis-
process, and energy interaction is a key element in occurrence of cussed. Important techniques of injury risk assessment and their
injury. Occupational injuries are caused by two types of energy classifications are discussed in Section 4. In Section 5, accident cau-
exposures: acute and chronic. Injuries caused by sudden energy sation theories, their generations, features, and important studies
transfer to the body of victim (short duration exposure) are termed are presented. Literature related to injury mechanism is reviewed
as traumatic injuries. Injuries caused by chronic exposure to in Section 6. In Section 7, a brief review of injury intervention tech-
repetitive motions or forceful exertions are termed as cumulative niques is presented. Critical appraisal of the literature and impor-
trauma disorders (CTDs) or repetitive strain injuries (RSIs) tant research issues are presented in Section 8.
(Putz-Anderson, 1988; Tayyari and Smith, 1997). The two types
of occupational injury are shown in Fig. 1. 2. Taxonomy of literature
Injury and accident are closely related terms. Often, the terms are
used synonymously (Hale and Hale, 1972; Langley, 1988), though Accidents and injuries are the results of sequences of interact-
they are not synonyms. Every accident need not necessarily result ing events occurring in worksystems. Presence of hazard is the
in human injury, but every injury is a result of an incident that can primary condition for occurrence of an accident/injury event. A
be termed as accident. As per the definition, transfer of energy to hu- person working in the vicinity of a hazard is exposed to risk of
man body in excess of physiological threshold is a necessary prere- occupational injury. Causal factors are responsible for transforma-
quisite for injury. There is no such qualifier for an accident. Injuries tion of injury risk into injury incident. Energy interactions during
are hardly governed by chance, while an event occurring accidently injury event decide how energy is transferred to victim’s body,
connotes a chance phenomenon (Wehmeier et al., 2005). There are and what is the severity of injury? Based on this sequence, acci-
inconsistencies and overlaps in researchers’ usage of the terms acci- dent/injury literature is divided into five categories as: (i) hazard
dent and injury, and a number of research articles set to investigate identification methods, (ii) injury risk assessment methods, (iii)
accidents actually analyze injuries (Langley, 1988). accident and injury causation theories, (iv) injury mechanism
A large and diverse literature is available on workplace safety, models, and (v) accident and injury intervention methods.
accident and injury research. The available literature is fragmented The taxonomy of literature including categories and representa-
and divided in different bodies of knowledge such as ergonomics tive studies in each category are shown in Table 1. First, the hazard
and human factors engineering, organization theory, safety engi- identification methods are divided into three types according to
neering, industrial psychology, medicine, environmental sciences, search techniques as backward tracking, forward tracking and
and law. Research in injury and accident causation is largely over- morphological methods. Second, the injury risk assessment studies
lapping and contains a lot of commonalities. However, since last are divided into four types according to the criterion variables used
three decades, researchers are vouching for the need to treat injury as studies considering injury rate, lost-time injury rate, severity in-
research as a distinguished research arena from accident research dex, and occurrence probability. Further, the injury risk assessment
(McKenna, 1983; Robertson, 1998). Causal models developed for studies are classified according to modeling techniques as studies
accident analysis may not be fully applicable to injury incidents. using categorical scales, statistical distributions, and soft comput-
Moreover, different accident theories explaining accident causation ing techniques. Third, the accident causation theories are divided
are well-accepted in the literature, however no theories are categor- into four generations (generation I to IV) as accident proneness
ically known as ‘‘injury theories’’. theory, domino theories, injury epidemiology models, and system
In this paper, we present a comprehensive review of literature re- theories. Fourth, injury mechanism models are divided into three
lated to injury and accident research. The intent of presenting the types as deviation models, energy models, and injury severity
models. Finally, intervention methods are divided into three types
as engineering, behavioral, and enforcement interventions. In the
forthcoming sections, the important studies are discussed in brief.

3. Methods of hazard identification

Hazard is defined as a thing that has potential to cause harm, or

as a source of danger (Webster, 2002). Presence of hazard is the root
Fig. 1. Types of injury – occupational and non-occupational. cause of accident leading to human, economic, societal, and envi-
 V.V. Khanzode et al. / Safety Science 50 (2012) 1355–1367 1357

Table 1
Taxonomy of literature.

Broad area of research Criterion for Category Representative studies

classification
Hazard identification Search technique Backward tracking Lees (1980), Harms-Ringdahl (1993)
Forward tracking Harms-Ringdahl (1993), Bahr (1997)
Morphological Johnson (1980), ILO (1988)
Risk assessment Criterion variable Injury rate Pines et al. (1987), Jeong (1999)
Lost-time injury rate Kjellen (1995)
Severity index Boyd and Radson (1999)
Occurrence probability Duzgun and Einstein (2004)
Modeling techniques Statistical distributions Boyd and Radson (1999), Cuny and Lejeune (2003)
Soft computing Zurada et al. (1997), Papazoglou and Ale (2007)
Categorical scales Mitchell et al. (1993), Rao-Tummala and Leung (1996)
Accident causation Generations Gen-I: Accident proneness theory Greenwood and Woods (1919), Keehn (1959), Kunce (1967)
Gen-II: Domino theory Heinrich (1932), Bird and Loftus (1976), Heinrich et al. (1980)
Gen-III: Injury epidemiology Haddon et al. (1964), Kjellen (1984a), Robertson (1998)
Gen-IV: System theory Hale and Hale (1970), Ayoub (1980), Kjellen (1987), Carder and Ragan (2003)
STS theory Robinson (1982), Galli (1999), Brown et al. (2000), Morel and Chauvin (2006)
Macroergonomic theory Hendrick (1994), Carayon and Smith (2000), Kleiner (2006)
Injury mechanism Energy interaction Deviation models Kjellen and Larsson (1981), Kjellen (1984a)
Energy models Haddon (1968), Tuominen and Saari (1982)
Severity models Petersen (1989), Salminen et al. (1992)
Injury intervention Intervention type Engineering Menckel (1990), Hasan et al. (2003)
Behavioral Sulzer-Azaroff (1987), Williams and Scott-Geler (2000)
Enforcement Gun (1993), Laurence (2005)

ronmental losses. Identification of hazards is the first step in assess- search is focused on potentially hazardous elements (like energy
ing risk of injury. Hazard identification includes identification of concentrations and hazardous materials), and potential targets
hazardous energy sources (for example, fuels, pressure containers, (i.e. persons or equipment). Critical or significant factors influ-
rotating machinery) and identification of hazardous processes and encing safety are focused while searching for hazards. Upon
situations (for example, oxidation, radiation, rapid pressure identifying these factors, paths leading to accident are drawn
changes, mechanical shocks) in a worksystem. Two approaches up. Following methods adopt morphological approach: (i)
are common in hazard identification (Kumamoto and Henley, Change analysis (Johnson, 1980; Ferry, 1988), (ii) energy analy-
1996): (i) a general engineering evaluation based on previous data sis (Haddon, 1980; Johnson, 1980; Harms-Ringdahl, 1993), (iii)
and operational history (an informal approach), and (ii) a formal ap- deviation analysis (Harms-Ringdahl, 1993), (iv) comparison
proach using well-developed hazard identification techniques. The analysis (Kjellen, 1995), (v) hazard and operability (HAZOP)
approaches towards hazard identification can be further divided studies (Lees, 1980; ILO, 1988; CPS, 1992; Harms-Ringdahl,
in three categories (Suokas, 1988; Willquist and Torner, 2003): 1993; Bahr, 1997), (vi) job safety and work safety analysis
(Suokas and Rouhiainen, 1984; Harms-Ringdahl, 1993), (vii)
1. Biased reactive approach, where information is analyzed after management oversight and risk tree (MORT) (Johnson, 1980),
occurrence of an accident event. This is typically an informal and (viii) operation and support hazard analysis (O&SHA) (Bahr,
approach, and general engineering evaluations fall under this 1997). Many hazard identification techniques are generic in nat-
category. ure, and can be applied to any worksystem. However, hazards
2. Biased proactive approach, where identification of hazards and are specific to worksystem under consideration (Maiti, 2005),
failure modes are based on information from similar systems or and professional experience is required to accurately identify
historic data from the same system. This is a formal approach hazards in a given worksystem. Biased proactive methods
towards hazard identification, and backward and forward track- depend heavily on available previous knowledge of the system.
ing methods fall under this category. Fault-tree analysis (FTA) Even unbiased proactive methods are not completely unbiased,
(Lees, 1980; Harms-Ringdahl, 1993; Kumamoto and Henley, as previous experience is valuable in morphological hazard
1996; Bahr, 1997) is a typical backward tracking method, while identification. Domain knowledge, thus plays an important role
event-tree analysis (ETA) (Lees, 1980; Harms-Ringdahl, 1993; in hazard identification.
Kumamoto and Henley, 1996; Bahr, 1997) is a typical forward
tracking method. Apart from FTA and ETA, following methods 4. Methods of risk assessment
adopt biased proactive approach: (i) Cause-consequence
Analysis (Lees, 1980; Henley and Kumamoto, 1981; Kumamoto Risk is defined as chance of loss, the degree of probability of
and Henley, 1996), (ii) Failure Mode Effect Analysis (FMEA) loss, a situation involving exposure to danger, or possibility that
(Harms-Ringdahl, 1993; Kumamoto and Henley, 1996; Bahr, something unpleasant will happen (Webster, 2002). Risk can also
1997; Franceschini and Galetto, 2001), and (iii) Failure Mode be defined as the ‘‘considered expected loss or damage associated
Effect and Criticality Analysis (FMECA) (Kumamoto and Henley, with the occurrence of a possible undesired event’’ (Nieuwhof,
1996). 1985). Risk assessment involves identifying potential threats, esti-
3. Unbiased proactive approach, where hazard analysis is carried mating their likelihood, and estimating the consequences. The
out without waiting for the events to occur, and without any combination of these estimates represents the risk associated with
restrictive assumptions about presence of specific hazards in the activity being evaluated. More formally, risk is perceived as a
the worksystem. This is also a formal approach and morpholog- collection of all possible pairs of likelihoods and outcomes [Li, Oi;
ical methods fall under this category. In these methods, the i = 1, 2, . . . , n], where Oi and Li denote an outcome and its likelihood,
1358 V.V. Khanzode et al. / Safety Science 50 (2012) 1355–1367

respectively (Kumamoto and Henley, 1996). Distribution pattern of ability or consequences (Chung et al., 1986; Freivalds and Johnson,
a given likelihood-outcome pair is known as risk profile. Occupa- 1990; Mallick and Mukherjee, 1996; Boyd and Radson, 1999; Cuny
tional injury risk is the likelihood of getting injured while perform- and Lejeune, 2003; Chang, 2004; Maiti and Khanzode, 2009; Maiti
ing a particular job. Phases in assessment of injury risk include et al., 2009; Khanzode et al., 2011). Beta distribution-based model
identification of hazards, evaluation of risk, and categorization of with lost workdays (consequence model) as indicator of injury risk
hazards based on risk index (Maiti, 2005). is one of the common models (Coleman and Kerkering, 2007).
Risk assessment methodologies are classified as: (i) qualitative Acute traumatic injuries occurring to individuals follow Poisson
methods, and (ii) quantitative methods (Tixier et al., 2002; Arunraj process and the inter-injury periods are exponentially distributed
and Maiti, 2007). These categories can further be classified as (Boyd and Radson, 1999). Injuries are modeled using mixed
deterministic, probabilistic, and combinatorial approaches. Quanti- Weibull distribution, considering injury as a failure, analogous to
tative risk assessment requires the estimation of frequency and reliability concepts. Such models assume differential injury liabil-
consequence severity in quantitative terms. This approach is suit- ity across individuals. Poisson regression models (Bailer et al.,
able when the risks are high, costs of detailed analyses are justified, 1997) are useful in adjusting injury rates for one or more explana-
and relevant data is available. The results are represented in the tory variables (for example, age, experience and occupation).
form of risk profiles. Qualitative risk assessment is more suitable Conditional probability-based models capture the risk of injury
when risks are low and small number of categories can cover entire through three-phase mechanism, namely pre-injury, injury and
range of consequences and likelihoods. The results are represented post-injury phases (Kjellen, 1984b,c).
in the form of risk matrices where occurrence probability and con- Third, soft computing techniques such as neural network-based
sequence severity represent the two axes. Risk assessment process models (Zurada et al., 1997) and functional block diagram-based
includes two decisions: (i) selection of criterion variables, and (ii) models (Papazoglou and Ale, 2007) are used for evaluating injury
selection of modeling techniques. As stated earlier, the modeling risk.
techniques can be subjective using scoring system to measure risk Apart from injury risk assessment methods, other techniques are
such as low, medium or high or quantify objectively the values of found in literature, which are applicable in different contexts. Rapid
likelihood and consequences using probability distributions. Use of upper limb assessment (RULA) (McAtamney and Corlett, 1993), pos-
probability models also helps in developing risk intervals instead tural loading on the upper body assessment (LUBA) (Kee and Kar-
of a point value. Based on these decisions, risk assessment methods wowski, 2001), and quick exposure checklist (QEC) (David et al.,
are divided into different categories that are explained below. 2008) are the methods of assessing risk of ergonomic hazards to
industrial workers. Risk of human errors (Reason, 1990) in a work-
4.1. Methods based on criterion variables system is evaluated by using techniques like Action Error Analysis,
Cause-consequence Analysis (Taylor, 1979; Lees, 1980; Singleton,
Different criterion variable-based indices are employed to esti- 1984a), Success Likelihood Index Method (Kirwan, 1990), and Hu-
mate risk of accident/injury. The most frequently used risk mea- man Error Probability Index (Khan et al., 2006). Probabilistic risk
sure is injury rate based on descriptive statistics (Pines et al., assessment (PRA) is the most commonly used methodology for reli-
1987; Jeong, 1999). Lost-time injury frequency rate, fatal accident ability and risk assessment in chemical, aerospace, nuclear and
rate (Kjellen and Sklet, 1995), cumulative incidence rate, severity other high-risk industries (Kumamoto and Henley, 1996). However,
rate, severity index (Boyd and Radson, 1999), disabling frequency PRA is primarily used in system- and component-level risk assess-
rate and disabling severity rate (Sheu et al., 2000) are some of ment in high-risk systems, and not for assessment of risk of injury
the indices commonly used to quantify risk. These rates are calcu- to industrial workers.
lated based on incidents per time period (for example, month or
year), man-hours worked (for example, thousand mandays or mil-
lion manhours), and number of persons (for example, per thousand 5. Accident causation theories
employees) (Kjellen and Sklet, 1995; Duzgun and Einstein, 2004).
Composite indices derived using principal component analysis A number of accident theories and models are put forth and
from more than one criterion variables (for example, number of examined by researchers over the years, but we divide these theo-
days away from work, degree of permanent disablement, and addi- ries and models into four generations as shown in Table 2. First
tional expenses) are used in some studies (Cuny and Lejeune, 1999, generation theories hold a primitive viewpoint towards accident
2003). Time between incident occurrences are also used as crite- causation. These theories hold a person’s traits and unsafe behavior
rion variable in some studies (Maiti and Khanzode, 2009; Maiti as responsible for accident (Greenwood and Woods, 1919). Second
et al., 2009). In a recent study, Maiti (2010) developed worksystem generation theories (domino theories) conceptualize a chain of
safety-capability index, similar to process capability index used in sequential events leading to an accident, and call these events as
quality control. dominos (Heinrich, 1932). Removal of any one domino from the
chain would break the chain of accident events. Domino theories
4.2. Methods based on modeling techniques are most widely used for accident mitigation in industry (Heinrich
et al., 1980). Deviation theory (Kjellen, 1984a,b) is a variation of
Different modeling techniques are proposed by researchers for domino theory where possible deviations in each domino are iden-
assessing accident/injury risk. First, ordinal scales are used for eval- tified and evaluated quantitatively. Injury epidemiology models
uating consequences of accident/injury incidents (Kejriwal, 2002). (originated in 1960s) represent third generation of accident re-
FMEA, FMECA, abbreviated injury scale (AIS) and injury severity search (Haddon et al., 1964). Injury epidemiology approach holds
scale (ISS) (Mitchell et al., 1993) employ different kinds of classifi- that accident prevention efforts do not necessarily lead to injury
cation schemes to weigh the consequences. Rao-Tummala and Leu- control in a worksystem. This approach focuses on energy transfer
ng (1996) propose a two-way priority matrix for risk assessment involved in injury incident, and tries to minimize it in order to min-
based on probability of occurrence and severity of consequences. imize the losses. System approach to accident causation (the fourth
Index of harm is a useful tool to compare occupational risk across generation) emerged in 1970s as a response to the challenge of
industries (Soloman and Alesch, 1989). maintaining safety in increasingly complex worksystems. Socio-
Second, accident/injury risk is modeled through appropriate technical systems theory (originated in 1950s) is a parallel sys-
statistical distributions, by fitting distribution of occurrence prob- tem-level approach for optimizing safety performance (Trist and
 V.V. Khanzode et al. / Safety Science 50 (2012) 1355–1367 1359

Table 2 sociotechnical system models (Brown et al., 2000; Paul and Maiti,
Accident theories: generations. 2008) and macroergonomic models (Hendrick, 1986) also investi-
Generation Theory Important features gate this interactive sequence. These models take holistic approach
Gen-I Accident Personality traits responsible for accident, and examine job-related factors apart from organization and indi-
proneness theory Differential involvement in accident, vidual-related factors.
Behavioral interventions
Gen-II Domino theories Unsafe act and condition as immediate 5.1. First generation: Accident proneness
predecessors of accident, Interventions
focused on unsafe acts
Gen-III Injury Uncontrolled energy transfer focused, There are four prevalent theories which belong to the first gen-
epidemiology Control at pre-injury, injury and post- eration of accident causation theories as follows:
theory injury stages
Gen-IV System theories Holistic approach, Integrated safety
5.1.1. Pure chance hypothesis
systems
STS theory Interacting social and technical sub- Pure chance hypothesis (prevailed up to 1920s) is the most
systems, Job design based on STS principles primitive viewpoint of explaining accidents, and generally not re-
Macroergonomic Holistic approach like system models, garded as ‘‘theory’’ (Haddon et al., 1964). The hypothesis states
theory Organization-centered Approach that everyone in a population has an equal chance of meeting an
accident. There are no patterns in the events leading to accident,
and the accident is an Act of God. Obviously this hypothesis finds
Bamforth, 1951). Macroergonomics has evolved as a approach hardly any place in modern accident and injury research.
towards system safety improvement since late 1980s (Hendrick,
1986). 5.1.2. Accident proneness theories
There is a correspondence among four generations of accident Accident proneness theory is the first theory that systematically
theories, the prevailing accident causation themes, and types of investigates accident proneness of individuals working in a system.
causal factors. Three themes of accident causation are: person- Greenwood and Woods (1919), Greenwood and Yule (1920) and
as-cause, system-as-cause, and system-person sequence as-cause Farmer and Chambers (1929) are the proponents of the term
(Leigh et al., 1990; Brown et al., 2000; Paul and Maiti, 2008). Each ‘‘accident proneness’’. The theory suggests that there exists a
theme examines specific causal factors to explain accident events. certain subgroup within the general population that is more likely
Causal factors responsible for accident/injury are divided into three to meet accidents. This liability is caused by some innate personal-
types as: (i) individual-related, (ii) organization-related, and (iii) ity traits explained through the construct ‘‘accident proneness’’
job-related factors (Paul and Maiti, 2007, 2008; Bajpayee et al., (Davis and Coiley, 1959; Keehn, 1959; Whitlock et al., 1963; Irwin,
2004; Huang et al., 2006; Mullen, 2004; Kjellen and Hovden, 1964; Kunce, 1967; Guilford, 1973; Kunce, 1974). Initially, accident
1993). The correspondence between generations of accident theo- proneness was perceived as a non-modifiable characteristic. Later
ries, themes of accident causation and types of causal factors are studies suggested the role of transient factors like stress, socio-
shown in Table 3. psychological make-up of an individual, safety culture, and work
Person-as-cause theme holds a person’s traits as responsible for environment in explaining accident proneness. The membership
accident/injury, and examines individual-related factors for acci- of accident-prone group keeps changing, and a person’s accident
dent causation. This belief is reflected clearly in first generation proneness may change with time. Liability towards accident is also
of accident theories, especially in accident proneness theory. Sec- influenced by previous accident experience. The liability either
ond generation (domino theory) accounts for system factors like decreases (burned fingers hypothesis) or increases (contagious
unsafe conditions, but focuses on person while designing interven- hypothesis), after meeting an accident compared to the rest of
tions. Domino theory includes factors like job satisfaction among the population (Schulzinger, 1954; Kirchner, 1961; Surry, 1969;
the dominos causing accident. Typically, proponents of the first Shaw and Sichel, 1971; Verhaegen et al., 1976). Persons prone to
two generations design behavioral interventions to minimize acci- injury at workplace are also prone to injury at other places (for
dents and injury. The second theme, system-as-cause, asserts that example, home), and during other activities (for example, leisure
the reasons of failure should be traced in the system, rather than activities or sports) (Salminen and Heiskanen, 1997). Number of
holding an individual responsible (Deming, 1986). Fourth genera- individuals with repeated injuries is higher than that expected by
tion models (system models) support this belief, and examine a chance (Visser et al., 2007) indicating presence of accident prone-
number of organization-related factors for accident causation ness. Sometimes the term ‘‘differential accident involvement’’ is
(Reason, 1997; Zohar, 2010). The third theme hypothesizes an used to describe accident proneness. However, the term ‘‘accident
interactive sequence between system and person. Job-related fac- proneness’’ lacks to convey the intention of prevention (McKenna,
tors play an important role in determining the nature of system- 1983), and puts more blame on individual than system. Accident
person interactions. Third generation models investigate energy proneness models align to person-as-cause theme of injury causa-
interactions in a worksystem and examine the role of job-related tion and suggest behavioral interventions for safety improvement.
factors in uncontrolled energy release. System models, especially
5.1.3. Unconscious motivation theory
Table 3
The theory of unconscious motivation has its roots in psychoan-
Correspondence between generations and causation themes. alytic theory. Accidents are caused by subconscious processes that
include guilt, aggression, anxiety, ambition, and conflict. This the-
Generation Causation theme Types of factors examined
ory focuses on an individual and correlation of his/her perception
Gen-I Person-as-cause (unsafe acts) Individual-related of environment with personality traits. Low scores on positive or
Gen-II System-as-cause (unsafe Individual- and job-
conditions), Person-as-cause related
socially desirable traits, low opinions on jobs and work environ-
(unsafe acts) ment, and withdrawals from work are found as significant contrib-
Gen-III System-person sequence (energy Job-related (leading to utors to accidents/injuries (Hill and Trist, 1953; Castle, 1956;
interactions) energy interactions) Davids and Mahoney, 1957; Fine, 1963; Eysenck, 1964; Verhaegen
Gen-IV System-as-cause, System-person Organization-related, Job-
et al., 1976). The unconscious motivation theory is aligned with
sequence related, Individual-related
person-as-cause theme, and leads to behavioral interventions.
1360 V.V. Khanzode et al. / Safety Science 50 (2012) 1355–1367

5.1.4. Adjustment-stress and goal-freedom-alertness theory However, applications of epidemiological approach in different
Adjustment-stress theory holds that individuals who cannot ad- worksystems are scarcely reported.
just with their work and work environment tend to have more
accidents than those who can (Kerr, 1950). Failure to adjust is 5.4. Fourth generation: System models
caused by a range of physical and psychological stressors. Goals-
freedom-alertness theory (Kerr, 1957) is an extension of Over the years, organizations have become more and more
adjustment-stress theory. This theory holds that certain persons complex to meet the societal demands in terms of productivity,
are accident prone at work due to lack of alertness on the task at safety and quality. Identification of deviation from normal operat-
hand, which is a result of lack of freedom to set goals at work. ing conditions and propagation of injury event has become more
Hardly any researchers have examined and reported these complex. As linear approach is insufficient to understand such
hypotheses in the context of occupational injury. complex events (Leveson, 2004), holistic system approach is
adopted to explain accident events (Leplat, 1984; Purswell and Ru-
5.2. Second generation: Domino theories mar, 1984; Singleton, 1984b; Leveson, 2004). However, holistic
models for accident causation need to incorporate quantitative
Heinrich suggests the concept of dominos (Heinrich, 1932) to analysis to increase the strength of explanation (Attwood et al.,
explain accident causation. He perceives the following five domi- 2006). To explain accident events using a large number of system
nos: (i) ancestry and social environment, (ii) person’s fault, (iii) un- factors, simulation-based models are commonly used (Hale and
safe act and/or condition, (iv) accident, and (v) injury. Heinrich Hale, 1970; Ayoub, 1975, 1980; Brenner, 1975; Smillie and Ayoub,
argues that removal of any one domino (in most conditions, unsafe 1976; Kjellen, 1987; Woodcock et al., 2005). Rao-Tummala and
act of a worker) breaks the propagation of accident event. Bird and Leung (1996), Carder and Ragan (2003), and Bellamy et al. (2008)
Loftus (1976) replace ancestry and social environment by manage- propose system models for occupational safety and accident anal-
ment’s lack of control, and adds basic causes (personal and job fac- ysis. An integrated safety management information system is pro-
tors) and immediate causes (substandard practices, conditions or posed for analysis and decision making related to accident events
errors) to the dominos. Domino theory holds that injury severity and safety interventions (Ayoub, 1979; Kjellen, 1982). System dy-
is a result of pure chance (Heinrich et al., 1980), which is con- namic modeling and structural equation modeling approaches are
tradicted by Shannon and Manning (1980), who argues that injury employed to address coalmine safety issues (Maiti and Bhattachar-
severity is a function of worksystem characteristics (the job-re- jee, 2000; Bhattacharjee and Maiti, 2000).
lated factors). Models based on domino theory are also known as Leamon’s human machine model (Leamon, 1980) identifies
process models (Kjellen and Larsson, 1981). International Loss Con- components of worksystem, while his occupational health and
trol Institute’s (ILCI) loss causation model (Bird and Germain, safety model (Leamon, 1988) identifies interactions in worksystem
1992) is based on domino theory. The ILCI model explicitly consid- leading to occupational accidents. However, applications of Lea-
ers contact with energy as the domino leading to injury. mon’s model in analyzing occupational injuries in different indus-
Domino theory captures the dynamics of injury event. However, trial settings are not reported. A number of factors affect safety in a
the theory does not account for complex interactions that take worksystem. The relationships between these factors and the
place during injury. Issues of risk quantification across each dom- safety status of the system are multivariate in nature, and some
ino are not addressed in the theory. Focusing on only five (or six) studies examine these relationships (Brown et al., 2000; Paul and
dominos leads to an error of generalization. In practice, safety pro- Maiti, 2008). The prevailing system models analyze and explain
fessionals perceive unsafe act as the ‘‘easiest-to-blame’’ domino, only the accident causation and do not focus on injury event, en-
and leave unsafe conditions generally unattended. Out of this nar- ergy interactions, and injury severity.
row focus on unsafe acts, behavioral interventions are commonly
designed and implemented, when unsafe conditions demand more 5.4.1. Sociotechnical and macroergonomic models
attention of the management. Originated as a methodology to study productivity problem in UK
coalmines in 1950s (Trist and Bamforth, 1951), Sociotechnical sys-
5.3. Third generation: Injury epidemiology tem (STS) approach evolved as a theory of organization design in
1980s (Emery, 1969; Cherns, 1976, 1987; Susman, 1976; Clegg,
Injury epidemiology models are derived from epidemics (Rob- 1988, 2000; Appelbaum, 1997; Keating et al., 2001; Majchrzak and
ertson, 1998). These models compare and contrast injury and dis- Borys, 2001; Carayon, 2006; Reiman and Oedewald, 2007). More re-
ease. Both cause harm to human body, and both are highly cently, STS methodology is used to model, analyze, and manage
patterned. While injuries damage the human body quickly, dis- occupational safety and risk in worksystems (Robinson, 1982; Ras-
eases do so at a slower pace (Haddon, 1980). Studies of injury pat- mussen, 1997; Galli, 1999; Brown et al., 2000; Svedung and Rasmus-
terns are helpful in injury control. Injury epidemiology models sen, 2002; Morel and Chauvin, 2006). According to STS approach, a
perceive three factors to explain the injury phenomenon as (Had- system is composed of two interacting subsystems: social and tech-
don et al., 1964): the host (the person injured), the agent (the en- nical subsystem. Joint optimization of these subsystems leads to
ergy leading to injury), and the environment (physical, biological better performance of the whole system. In modeling injury risk,
and organizational). Among the environmental factors, transient joint consideration of person-level (social) factors and system-level
factors (those changing with time) are the most immediate factors (technical) factors is necessary according to STS approach. However,
causing injury. Injuries resulting from such instantaneous factors such approach is adopted by only a few researchers, for example,
are common, yet poorly understood from etiological perspective Brown et al. (2000) in steel industry, and Paul and Maiti (2008) in
(Sorock et al., 2001). Examination of transient factors through coalmining. This status of literature leaves open opportunities for
appropriate study design warrants researchers’ attention. Sorock researchers to adopt STS approach in modeling accident and injury
et al. (2004) examines transient factors for hand injuries in differ- risk and safety performance in different industrial settings. Macr-
ent occupations in US and reports malfunctioning equipment and oergonomic approach to worksystem design focuses on interactions
use of nonstandard method as significant factors. Chow et al. at organization-level (Smith and Sainfort, 1989; Hendrick, 1986;
(2007) examines transient risk factors for hand injuries in different Carayon and Smith, 2000; Robins, 2000; Kleiner, 2006). A number
occupations in Hong Kong and reports absence of hand-gloves, of premises of the macroergonomic approach are drawn from
being rushed and working overtime as significant factors. sociotechnical systems theory (Kleiner, 2006). Work compatibility
 V.V. Khanzode et al. / Safety Science 50 (2012) 1355–1367 1361

model is a recent methodology of assessment of occupational risk in worksystems containing more hazards. There are more chances
(Genaidy et al., 2007a,b, 2008, 2009), which is fundamentally a of getting injured while performing certain activities in a worksys-
macroergonomic approach. tem. Manual material handling is one of the most common activi-
ties leading to workplace injury (Davies et al., 2003). Manual
5.5. Factors affecting injury material handling is also the commonest activity responsible for
musculoskeletal disorders, strains and sprains. Individuals are
Accident causation theories and causation themes (Table 3) more prone to injury when working in night shifts than in day
identify several factors contributing towards the occurrence of shifts. Workplace factors like poor housekeeping and space avail-
accidents/injuries in worksystems. Over the years, a large number ability are found to be responsible in causing accidents in construc-
of individual-related, job-related, and organizational-related tion industry. Job stress (Paul and Maiti, 2005), job dissatisfaction
factors are examined in different studies. In the following para- (Paul and Maiti, 2005), job responsibility (Ferguson et al., 1984,
graphs findings on effect of these factors on accident/injury event 1985) and work performance (Ferguson et al., 1984, 1985) are
are presented in brief. some more job-related factors that correlate strongly with injury.
High stress, high dissatisfaction, less responsibility, and poor work
5.5.1. Individual-related factors performance are associated with more injuries. Technology change
Individual-related factors are examined the most commonly by may reduce injury risk in one area and increase in other unknown
first generation theorists. Age and work experience are the most areas. Effect of technology change on injury is more prevalent
investigated factors. Increase in age increases risk of injury (Fotta among younger workers (Blank et al., 1996). A study of Turkish
and Bockosh, 2000; Jeong, 1999). However, contrary findings also coal mines shows that mechanization changes the riskiest loca-
exist (Bennett and Passmore, 1984a). Inadequate work experience tions from face to out-by-face, and reduces number of injuries (Sari
(less than 1 year) poses high risk of injury (Keyserling, 1983; Butani, et al., 2004).
1988). Young workers (under 25 years) are at higher risk in nonfatal
injury incidents, while older workers are at higher risk in fatal injury 5.5.3. Organization-related factors
incidents (Salminen, 2004). Warr (1993) classifies occupational Since 1970, organizational factors have been considered as
tasks into four types as: age-impaired, age-counteracted, age-neu- important causal factors of accident occurrences (e.g., (Powell
tral, and age-enhanced tasks. The relationship of injury risk with et al., 1971)). Subsequently, the term ‘safety climate’ was defined
age is respectively increasing, inverted-U shaped, unaffected, and by Zohar (1980) as the summary of molar perceptions that employ-
decreasing for these four types (Laamme and Menckel, 1995). Injury ees share about their work environment. Some of the significant
risk is high in initial years of employment, diminishes when an indi- organization-related factors are: workgroup size (Guastello and
vidual acquires sufficient work experience, and rises again with Guastello, 1987), management support (Gillen et al., 2002; Maiti
aging. Gun and Ryan (1994) find absence of correlation between et al., 2004), coworker support and supervisory support (Gillen
age/experience and injury. However, such results are exceptions, et al., 2002; Maiti et al., 2004), management’s commitment to
rather than trends, and are applicable only to age-neutral tasks. safety (O’Toole, 2002), workplace safety status (Lindell, 1997; Gil-
Age, experience, education (Leigh et al., 1990) and mental ability len et al., 2002), and workgroup size (Guastello and Guastello,
correlate well with injury as well as with poor work performance 1987). Higher management, coworker and supervisory support,
(Ferguson et al., 1984). Living habits (alcohol and tobacco consump- higher commitment towards safety, safer perception of system sta-
tion) are found to have an association with injury events (Dawson, tus, and smaller workgroup sizes (less than 15), are found to be
1994; Zwerling et al., 1996). Negative affectivity (the chronic expe- associated with low injury risks. Workers, who experience their
rience of negative emotional states and lack of emotional stability) workplace as safe, also perceive management and coworker
correlates strongly with injury (Verhaegen et al., 1985; Maiti et al., support as high (Gillen et al., 2002). Cheyne et al. (2002) measured
2004; Paul and Maiti, 2005). Skill-based errors correlate strongly employee safety attitudes and examined relationships between
with injuries (Reason, 1990; Hobbs and Williamson, 2002) as com- components of organizational safety elements, perceptions of phys-
pared to violations and mistakes. Absenteeism, risk-taking nature ical work environment and perceptions of workplace hazards.
and injuries correlate with each other (Verhaegen et al., 1976). Neg- Organizational variables influence environmental and job-related
ative affectivity and risk-taking nature correlate with violations variables, which in turn influence precursors to safe behavior. How-
leading to injury (Verhaegen et al., 1985). Negative affectivity and ever, the focus of this work is only behavioral. Behavioral safety
risk-taking behavior along with job dissatisfaction are strong predic- models hypothesize a relationship between safe/unsafe behavior
tors of injury in Indian coalmines (Paul et al., 2005; Paul and Maiti, and injury. Safer the behavior at work, lesser is the risk of injury.
2007). According to Hill and Trist (1953), the negative traits A number of studies aimed at finding drivers of safety behavior find
(absenteeism, negative affectivity, risky behavior and violations) that unsafe behavior correlates with higher injury (Andriessen,
reflect the quality of relationship of an individual with the organiza- 1978; Prussia et al., 2003).
tion and contribute to injury. Role overload (Mullen, 2004), organization’s priority to perfor-
mance over safety (Mullen, 2004), influences of socialization at
5.5.2. Job-related factors work (Mullen, 2004), poor safety attitudes (Rundmo and Hale,
Injury risk differs job-wise and a number of job-related factors 2003; Mullen, 2004), perceived hazards (Seo, 2005), and high per-
are the predictors of injury (Ferguson et al., 1985). This relationship ceived risks (Rundmo, 1996; Mullen, 2004; Seo, 2005) are some of
is stronger than individual-related factors (McLeod et al., 2003). the drivers of unsafe behavior at work. An improvement in per-
Some of the commonest job-related factors examined and found ceived safety climate leads to a shared understanding of unsafe
significant by the researchers are: occupation (Leigh et al., 1990; behavior (Prussia et al., 2003), particularly in tightly-coupled sys-
Maiti and Bhattacherjee, 1999; Maiti et al., 2001), location of work tems (Perrow, 1984), which in turn seemingly reduces injury risks.
(Leigh et al., 1990; Maiti and Bhattacherjee, 1999; Maiti et al., Measurement of safety climate and its effect on safety performance
2001), hazards in worksystem (Maiti et al., 2004; Khanzode et al., has witnessed growing attention of researchers in the past few
2011), activity (Leigh et al., 1990), shift of working (Levin et al., years (Gillen et al., 2002; Prussia et al., 2003; Fullarton and Stokes,
1985; Frank, 2000), and workplace factors (Haslam et al., 2005). 2007). Instruments measuring safety climate predict injury rates
Injury risk differs with different occupations, and is affected by with varying strength and validity. Siu et al. (2004) evaluated effect
location of work where a given job is performed. Injury risk is high of safety climate on safety performance through a path model.
1362 V.V. Khanzode et al. / Safety Science 50 (2012) 1355–1367

Safety attitude and psychological distress affect injury rates (Siu hypothesis which does not account for injury severity. On the con-
et al., 2004). Negative affectivity affects safety climate and both, trary, Hale and Hale (1972), Petersen (1989) and Salminen et al.
in turn, affect injury (Maiti et al., 2004). Poor safety environment (1992) support a differential causation hypothesis for injuries of
is a factor to injury (Paul and Maiti, 2008). Industry-specific haz- varying severity. The differential causation hypothesis warrants
ards affect the relationship between safety climate and safety per- examination of injuries in the light of varying severity. Petersen
formance (Smith et al., 2006). For the last three decades, many (1989) validated the differential hypothesis in mining industry,
researchers focused their study on measuring safety climate, and while Salminen et al. (1992) obtained similar findings independent
establishing relation between safety climate and safety perfor- of occupations in Swedish industry. Severity of injury is not gov-
mance (Zohar, 2010). erned by chance, but depends on energy interaction involved in
the injury event (Shannon and Manning, 1980). In general, two
6. Injury mechanism models types of factors are important in determining injury severity (Hull
et al., 1996): (i) factors related to susceptibility of body tissue to
Transfer of energy to human body is a central issue in injury damage, and (ii) factors related to energy interaction. Type and
incident. Injury models attempt to explain how this energy is built agency of accident, activity and injured body part are found to be
in a worksystem, and what causes release of uncontrolled energy. associated significantly with injury severity in Australian coal-
Consideration of energy interactions those are present at the time mines (Hull et al., 1996) and in Indian coalmines (Khanzode
of injury gains obvious importance in this regard. The entire mech- et al., 2010). However, Hull et al. (1996) found location and occu-
anism is not simple and linear but often multiple-phased, complex pation as non-significant factors. Various other causal factors also
and unpredictable (Perrow, 1984). Till 1960s, objects carrying en- decide injury severity. For example, age, experience, job, location,
ergy were perceived as injury agents. However, Gibson’s work of- mining system are the factors found to be affecting injury severity
fered a paradigm shift in injury research (Robertson, 1998). in coalmines (Bennett and Passmore, 1984b, 1985).
Different types of energy, i.e. mechanical, thermal, chemical, and
electrical energy, and ionizing radiation are now identified as 7. Intervention strategies
agents of injury. Later, Haddon (1972) proposed Haddon Matrix
to categorize injury-related events in three phases namely: pre-in- Intervention strategies for occupational accident/injury control
jury, injury and post-injury for the host, vehicle and environment. are based on the following principles (Haddon, 1980; Heath,
Since then, multiple-phased structure of injury mechanism is 1982): (i) isolating the source of energy release, (ii) separating
examined by different researchers. The models explaining mecha- the source and the user in time–space, (iii) providing protecting
nism of injury event can be classified into three types as follows: envelope around the user, (iv) providing instruction and informa-
tion necessary to enable the users to operate rationally in the pres-
6.1. Deviation models ence of energy release. While first three strategies refer to
engineering interventions (hardware design and organization de-
Deviation is said to occur when a system characteristic attains a sign), the last one refers to education (and training), and enforce-
value beyond norms (Kjellen, 1984a; Kjellen and Hovden, 1993). ment (of safety rules and regulations).
Kjellen and Larsson (1981) identify three distinct phases occur in
injury incident as: initial phase, concluding phase and injury phase.
7.1. Engineering interventions
Haslam et al. (2005) perceive three tiers in injury mechanism as:
originating influences, shaping factors and immediate accident cir-
Engineering interventions generally adopt an approach called
cumstances. Deviation models explain injury on the basis of devi-
‘‘design for safety’’ or ‘‘integrating safety into design’’ (Menckel,
ations occurring in the system variables which lead to release of
1990; Kletz, 1990; Hasan et al., 2003; Pohjola, 2003; Hollnagel,
uncontrolled energy (Kjellen, 1984b,c). However, the deviation
2004). Most common shop floor strategies of engineering interven-
models are scarcely developed by further researchers.
tion include incorporating shields and guards with the equipment,
and providing personal protective equipment to the operator. Engi-
6.2. Energy models neering interventions are often associated with technology
changes. Accident risk increases during phase of rapid technology
Energy models describe the phases in injury mechanism in change (Saari, 1982). Better technology leads to reduction of occu-
slightly different terms as compared to deviation models. These pational hazards, but it is accompanied by increased risks in other
phases are: energy build-up phase, energy release phase and unknown areas (Singleton, 1982). Accidents in high-risk systems
impact phase (Haddon, 1968). Oleske et al. (1989) modeled acci- are usually originated and propagated through system shortcom-
dent types and sources of injury to explain the injury mechanism. ings rather than operators errors. Multiple and unexpected interac-
Energy build-up in worksystem and human interactions with sys- tions causing failures are inevitable in complex systems (Perrow,
tem energy play an important role in injury incident (Tuominen 1984). According to Perrow (1984), the intervention strategies
and Saari, 1982). Investigation of uncontrolled energy transfer as need to be shifted to system shortcomings rather than local design
a direct cause of injury is the foremost step in injury analysis variables.
(OSHA, 2002; Storbakken, 2002). Design for safety is not always a viable option for existing work-
systems. However, job redesign is a possible alternative in many
6.3. Injury severity models cases. If human errors are treated as instances of human–machine
or human-task misfits, then systematic or frequent instances of
Examination of factors affecting injury severity in different misfits are the symptoms of design errors, requiring job redesign
worksystems is helpful in minimizing the severity of impact, and (Rasmussen, 1982). On the other hand, occasional instances of mis-
reducing lost mandays under a given situation. Factors affecting in- fits are either due to system variability (system failures) or human
jury severity are not exactly the same as the factors causing injury. variability (human errors). Incidents that increase exposure to en-
Major and fatal injuries are caused by different factors as compared ergy or concentrate on different forms of energy can be modified
to minor and nonfatal injuries. Heinrich et al. (1980) and Lozada- through job redesign under certain conditions (Robertson, 1998).
Larsen and Laughery (1987) perceive an identical causation Typically, any worksystem, originally designed for safe conditions,
 V.V. Khanzode et al. / Safety Science 50 (2012) 1355–1367 1363

deteriorates first to a meta-stable unsafe condition (Mol, 2003),

and later to an unstable condition, leading to uncontrolled energy
release (Hale, 1990). Redesign activity need to be carried out at the
meta-stable condition before the system enters into the unstable
zone. Several studies reported that job design and ergonomic inter-
ventions lead to improvement in safety, increased job satisfaction
(for example, Aborg et al. (1998)) and reduction in musculoskeletal
disorders (Geneidy, 1990; Christmansson et al., 1999; Hess et al.,
2004; Ferguson et al., 2005; Holmstrom and Ahlborg, 2005; Colom-
bini and Occhipinti, 2006; Torma-Krajewski et al., 2007). However,
studies examining effect of job redesign on injury events are not
reported.

7.2. Behavioral interventions

Studies related to education and training interventions mainly Fig. 2. Gaps in injury and accident literature.
focus on behavioral modification (for example, Heath (1982),
Sulzer-Azaroff (1987), Krause et al. (1999), Williams and Scott-Gel- FMEA, and FMECA are good examples of this. However, the acci-
er (2000), Whysall et al. (2006), Paul and Maiti (2007)). However, dent causation models are not interfaced with hazard identifica-
effects of behavioral interventions on injury prevention are not tion and risk assessment methodologies. Interfaces between risk
conclusively examined and alternate viewpoints exist. Designers assessment techniques and accident causation models would be
of safe behavior programs assume unsafe behavior as the prime helpful in examining whether an accident will occur under given
cause of accidents, and overlook other causal factors (Hopkins, conditions. On the other hand, injury mechanism models and in-
2006). Behavioral scientists assume human beings as objects train- jury severity models are well interfaced with each other. However,
able by reward and punishable to do virtually anything, and gener- these models are not interfaced with accident causation models.
ally overlook energy interactions at workplace (Robertson, 1998). Development of such interfaces will enhance understanding of
Behavioral approach ignores biological limitations as well as inter- how the risk is transferred first to an accident event and then to
mediate cognition between training and expected behavior, and an injury.
therefore may yield poor results. Behavioral interventions can be The complex system-level accident causation models are more
successful only if they decrease exposure to hazards. However, suitable for analyzing event chains that lead from component-level
behavioral interventions do not include components for hazard to system-level failures and cause accidents. These complex event
exposure reduction. The belief that all safety problems can be chains lead to high severity-low frequency accidents and not to
solved by training is a myth (Mason, 1973). Training generally en- low severity-high frequency incidents. However, from injury re-
hances skill and knowledge on the job, and is effective in reducing search point of view, low severity-high frequency incidents are
injury if lack of skill or knowledge is a strong causal factor for in- more important. Methodologies available for analyzing high sever-
jury. However lack of skill or knowledge is not always a causal fac- ity-low frequency incidents are not particularly useful for identify-
tor of injury. Unless the occurrence of energy interactions is ing injury patterns and effectively mitigate them. This ineffective
understood properly, the effectiveness of behavioral interventions ness is reflected in astronomical figures of occupational injuries.
may be limited. Unilateral interventions stressing on behavioral It is therefore essential to develop methodologies that address
change may lead to improvement in perceived safety without the injury problem specifically.
reducing actual injury risk. Thus, behavioral interventions have a Any injury causation model, in order to be effective, needs to
limited usefulness in injury reduction (Robertson, 1998). comprise of four components as shown in Fig. 3. First, as injury in-
volves transfer of energy to human body, models that do not con-
sider energy interactions can only be of limited use in
7.3. Enforcement interventions
understanding injury. Second, transient risk factors are important
in realization of injury risk. These factors are the most immediate
With regard to assessment of enforcement interventions to-
to injury incident and vary between the injured and non-injured
wards injury prevention, status of the literature is even poorer. A
person, when most of the other system variables are the same (Sor-
few researchers examined effects of rules and regulations on occu-
ock et al., 2001, 2004; Chow et al., 2007). Therefore, an injury mod-
pational safety (Hale, 1990; Gun, 1993; O’Toole, 1999; Laurence,
el should provide a scheme for identification of transient risk
2005). A general inclination of authorities to impose more regula-
factors in a worksystem. Third, an injury model needs to establish
tions to improve workplace safety should be curbed as this may be
links of causal factors with energy transfer event. Stronger the rela-
counter-productive (Hale, 1990). As more industries are imple-
tionship, more important is the role of that factor in injury occur-
menting injury interventions, researchers need to critically evalu-
rence. Fourth, injury models use different response variables to
ate significance of effects of such interventions on injury reduction.
measure the injury phenomenon. For the findings to be useful in
understanding injury, the response variable should have a good
8. Critical appraisal and future research directions content validity with injury event. Injury rates, accident rates,
and man-days lost are direct and good measures of injury risk.
There are a number of critical issues that are derived from the However, safe behavior (Brown et al., 2000; Cheyne et al., 2002;
review. As explained earlier (in Section 2), an injury incident fol- Prussia et al., 2003; Rundmo and Hale, 2003; Seo, 2005), safety atti-
lows a chain of events starting with presence of hazard and leading tude (Siu et al., 2004), and safety climate (Hopkins, 2006) are indi-
to injuries of varying severity. The review reveals that there are rect measures of injury incident. The validity of results needs to be
methodological gaps in analyzing this sequence of events, which assessed critically when injury models use indirect measures as
are shown in Fig. 2. response variables.
Fig. 2 shows that hazard identification and risk assessment Effects of different accident prevention strategies on injury
methodologies are well interfaced with each other. ETA, FTA, control and prevention are not evaluated objectively and reported
1364 V.V. Khanzode et al. / Safety Science 50 (2012) 1355–1367

Ayoub, M., 1975. The problem of occupational safety. Industrial Engineering 7, 16–
23.
Ayoub, M., 1979. Integrated safety management information system: Part I: design
and architecture. Journal of Occupational Accidents 2, 135–157.
Ayoub, M., 1980. Simulation modeling and analysis in safety: Part I: planning and
design. Journal of Occupational Accidents 3, 3–20.
Bahr, N., 1997. System Safety Engineering and Risk Assessment – A Practical
Approach. Taylor & Francis, Washington DC.
Bailer, A., Reed, L., Stayner, L., 1997. Modeling fatal injury rates using Poisson
regression: a case study of workers in agriculture, forestry, and fishing. Journal
of Safety Research 28, 177–186.
Bajpayee, T., Rehak, T., Mowrey, G., Ingram, D., 2004. Blasting injuries in surface
mining with emphasis on flyrock and blast area security. Journal of Safety
Research 35, 47–57.
Fig. 3. Components of effective injury causation model. Bellamy, L., Geyer, T., Wilkinson, J., 2008. Development of a functional model which
integrates human factors, safety management systems and wider
organizational issues. Safety Science 46, 461–492.
Bennett, J., Passmore, D., 1984a. Correlates of coalmine accidents and injuries: a
conclusively. Post-trauma behavior of injury victims, long term literature review. Accident Analysis and Prevention 16, 37–45.
Bennett, J., Passmore, D., 1984b. Days lost from work due to injuries in US
effects of injury on the victims and their peers and the costs of underground bituminous coal mines. Journal of Occupational Accidents 5, 265–
long-term effects to society are not well researched. Effects of 278.
enforcement of safety laws on injury reduction are not thoroughly Bennett, J., Passmore, D., 1985. Multinomial logit analysis of injury severity in US
underground bituminous coal mines. Accident Analysis and Prevention 17,
researched and reported. Lack of comprehensive data on injury sur-
399–408.
veillance from organized and unorganized sector imposes limita- Bhattacherjee, A., Maiti, J., 2000. A new look into quantitative analysis of mine
tions on injury research. Timely availability of reliable injury data safety studies. Transactions of the Society for Mining, Metallurgy, and
Exploration, Inc. 308, 83–90.
will help researchers to analyze injury problems from different
Bird, F., Germain, G., 1992. Practical Loss Control Leadership. International Loss
perspectives and find out ‘‘patterns’’ in injury causation. Control Institute Inc., Loganville, Georgia.
To the best of the knowledge of the authors, there is no scheme Bird, F., Loftus, R., 1976. Loss Control Management. Institute Press, Loganville,
that models and controls recurring nature of operational hazards Georgia.
Blank, V., Laflamme, L., Diderichsen, F., 1996. The impact of major transformations
in worksystems. The study by Khanzode et al. (2011), can be consid- of a produc-tion process on age-related accident risks: a study of an iron ore
ered an initiation to this end. There exists a research gap as regards mine. Accident Analysis and Prevention 28, 627–636.
to quantification of occurrence–removal–recurrence characteristic Boyd, A., Radson, D., 1999. Statistical analysis of injury severity rates. IIE
Transactions 31, 207–216.
of injury-causing hazards, although there is a lot of literature Brenner, L., 1975. Accident investigations: multilinear events sequencing methods.
available for evaluation of risks of catastrophe-causing hazards. As Journal of Safety Research 1, 67–73.
hazards are at the root of injury-event chain, a scheme for character- Brown, K., Willis, P., Prussia, G., 2000. Predicting safe employee behavior in the steel
industry: development and test of a sociotechnical model. Journal of Operations
ization of injury-causing hazards is a useful first step towards mod- Management 18, 445–465.
eling injury risk. Butani, S., 1988. Relative risk analysis of injuries in coal mining by age and
Risk assessment techniques focus on system failures leading to experience at present company. Journal of Occupational Accidents 10, 209–216.
Carayon, P., 2006. Human factors of complex sociotechnical systems. Applied
high severity – low frequency accidents and quantify risks caused
Ergonomics 37, 525–535.
by such accidents. PRA and other risk assessment techniques do Carayon, P., Smith, M., 2000. Work organization and ergonomics. Applied
not quantify risk of occupational injury to an individual, per se. Ergonomics 31, 649–662.
Carder, B., Ragan, P., 2003. A survey-based system for safety measurement and
Injury risk profiles across different jobs and worksystems are neither
improvement. Journal of Safety Research 34, 157–165.
documented extensively nor reported widely in published litera- Castle, P., 1956. Accidents, absence and withdrawal from the work situation.
ture. To some extent, deviation models (for example, Kjellen Human Relations 9, 223–233.
(1984a)) provide a methodology to quantify injury risks. However, Chang, W., 2004. A statistical model to estimate the probability of slip and fall
incidents. Safety Science 42, 779–789.
such attempts are scarce. Hazard identification and risk assessment Cherns, A., 1976. The principles of sociotechnical design. Human Relations 29, 783–
techniques are more used for assessing chemical and process- 792.
related hazards and accident risks as compared to mechanical Cherns, A., 1987. Principles of sociotechnical design revisited. Human Relations 40,
153–162.
hazards and traumatic injury risks. Cheyne, A., Oliver, A., Tomas, J., Cox, S., 2002. The architecture of employee attitudes
In summary, with respect to injury research, there are signifi- to safety in the manufacturing sector. Personnel Review 31, 649–670.
cant challenges to and scope for future researchers. We will follow Chow, C., Lee, H., Lau, J., Yu, I., 2007. Transient risk factors for acute traumatic hand
injuries: a case-crossover study in Hong Kong. Occupational Environmental
the traditional path of risk assessment from hazard identification Medicine 64, 47–52.
to risk prioritization, as well as for accident and injury causation. Christmansson, M., Friden, J., Sollerman, C., 1999. Task design, psycho-social work
However, it is the time is now to focus on high-frequency and climate and up-per extremity pain disorders: effects of an organizational
redesign on manual repetitive assembly jobs. Applied Ergonomics 30, 463–472.
low-severity incidents under injury research while accident re-
Chung, M., Wu, S., Herrin, G., 1986. The use of mixed Weibull model in occupational
search on high-severity and low-frequency incidents to be contin- injury analysis. Journal of Occupational Accidents 7, 239–250.
ued. The gaps identified based on the review can be explored Clegg, C., 1988. Appropriate technology for manufacturing: some management
issues. Applied Ergonomics 19, 25–34.
further in the directions mentioned.
Clegg, C., 2000. Sociotechnical principles for system design. Applied Ergonomics 31,
463–477.
Coleman, P., Kerkering, J., 2007. Measuring mining safety with injury statistics: lost
References workdays as indicators of risk. Journal of Safety Research 38, 523–533.
Colombini, D., Occhipinti, E., 2006. Preventing upper limb work-related
Aborg, C., Fernstrom, E., Ericson, M., 1998. Work content and satisfaction before and musculoskeletal disorders (UL-WMSD): new approaches in job (re)design and
after a reorganization of data entry work. Applied Ergonomics 29, 473–480. current trends in standardization. Applied Ergonomics 37, 441–450.
Andriessen, J., 1978. Safe behavior and safety motivation. Journal of Occupational CPS, 1992. Guidelines for Hazard Evaluation Procedures. Center for Chemical
Accidents 1, 363–376. Process Safety, American Institute of Chemical Engineers, New York.
Appelbaum, S., 1997. Socio-technical systems theory: an intervention strategy for Cuny, X., Lejeune, M., 1999. Occupational risks and the value and modelling of a
organizational development. Management Decision 35, 452–463. measurement of severity. Safety Science 31, 213–229.
Arunraj, N., Maiti, J., 2007. Risk-based maintenance: techniques and applications. Cuny, X., Lejeune, M., 2003. Statistical modelling and risk assessment. Safety Science
Journal of Hazardous Materials 142, 653–661. 41, 29–51.
Attwood, D., Khan, F., Veitch, B., 2006. Occupational accident models: where we David, G., Woods, V., Li, G., Buckle, P., 2008. The development of the Quick Exposure
have been and where are we going? Journal of Loss Prevention in the Process Check (QEC) for assessing exposure to risk factors for work-related
Industries 19, 664–682. musculoskeletal disorders. Applied Ergonomics 39, 57–69.
 V.V. Khanzode et al. / Safety Science 50 (2012) 1355–1367 1365

Davids, A., Mahoney, J., 1957. Personality dynamics and accident proneness in an Hale, A., 1990. Safety rules O. K.? Possibilities and limitations in behavioral safety
industrial setting. Journal of Applied Psychology 41, 303–305. strategies. Journal of Occupational Accidents 12, 3–20.
Davies, J., Kemp, G., Frostick, S., McElwaine, C.D.J., 2003. Manual handling injuries Hale, A., Hale, M., 1970. Accidents in perspective. Occupational Psychology 44, 115–
and long term disability. Safety Science 41, 611–625. 122.
Davis, D., Coiley, P., 1959. Accident proneness in motor-vehicle drivers. Ergonomics Hale, A., Hale, M., 1972. A review of the industrial accident research literature. Tech.
2, 239–246. rep., Her Majesty’s Stationery Office, London.
Dawson, D., 1994. Heavy drinking and the risk of occupational injury. Accident Harms-Ringdahl, L., 1993. Safety Analysis: Principles and Practice in Occupational
Analysis and Prevention 26, 655–665. Safety. Elsevier Applied Science, London.
Deming, W., 1986. Out of the Crisis. MIT Center for Advanced Engineering Study, Hasan, R., Bernard, A., Ciccotelli, J., Martin, P., 2003. Integrating safety into the
Cambridge. design process: elements and concepts relative to the working situation. Safety
Duzgun, M.A., Einstein, H., 2004. Assessment and management of roof fall risks in Science 41, 155–179.
underground coal mines. Safety Science 42, 23–41. Haslam, R., Hide, S., Gibb, A., Gyi, D., Pavitt, T., Atkinson, S., Duff, A., 2005.
Emery, F., 1969. Systems Thinking. Penguin, London. Contributing factors in construction accidents. Applied Ergonomics 36, 401–
Eysenck, I., 1964. The personality of drivers and pedestrians. In: Haddon, W., 415.
Suchman, E., Klein, D. (Eds.), Accident Research: Methods and Approaches. Heath, E., 1982. Worker training and education in occupational safety and health- a
Harper and ROW, New York. report on practice in six industrialized western nations: part two of a four part
Farmer, E., Chambers, E., 1929. A study of personal qualities in accident proneness series. Journal of Safety Research 13, 73–87.
and proficiency. Tech. Rep., Industrial Health Research Board, His Majesty’s Heinrich, H., 1932. Industrial Accident Prevention, first ed. McGraw Hill, New York.
Stationery Office, London. Heinrich, H., Petersen, D., Ross, N., 1980. Industrial Accident Prevention, fifth ed.
Ferguson, J., McNally, M., Booth, R., 1984. Individual characteristics as predictors of McGraw-Hill, New York.
accidental injuries in naval personnel. Accident Analysis and Prevention 16, 55– Hendrick, H., 1986. Macroergonomics as a preventive strategy in occupational
62. health: an organizational level approach. In: Hendrick, H., Bradley, G. (Eds.),
Ferguson, J., McNally, M., Booth, R., 1985. Accidental injuries among naval personnel Human Factors in Organizational Design and Management II. Elsevier Science.
by occupation, duty status, and pay-grade. Accident Analysis and Prevention 17, Henley, E., Kumamoto, H., 1981. Reliability Engineering and Risk Assessment.
79–86. Prentice-Hall Inc, New Jersey.
Ferguson, S., Marras, W., Burr, D., 2005. Workplace design guidelines for asymptotic Hess, J., Hecker, S., Weinstein, M., Lunger, M., 2004. A participatory ergonomics
vs. low-back-injured workers. Applied Ergonomics 36, 85–95. intervention to reduce risk factors for low-back disorders in concrete laborers.
Ferry, T., 1988. Modern Accident Investigation and Analysis. Wiley, New York. Applied Ergonomics 35, 427–441.
Fine, B., 1963. Introversion-extroversion and motor vehicle driver behavior. Hill, J., Trist, E., 1953. A consideration of industrial accidents as a means of
Perceptual and Motor Skills 16, 95–100. withdrawal from the work situation- a study of their relation to other absences
Fotta, B., Bockosh, G., 2000. The aging workforce: an emerging issue in the mining in an iron and steel works. Human Relations 6, 357–380.
industry. Hobbs, A., Williamson, A., 2002. Unsafe acts and unsafe outcomes in aircraft
Franceschini, F., Galetto, M., 2001. A new approach for evaluation of risk priorities of maintenance. Ergonomics 45, 866–882.
failure modes in FMEA. International Journal of Production Research 39, 2991– Hollnagel, E., 2004. Barriers and Accident Prevention. Ashgate Publishing Limited,
3002. Aldershot Hants.
Frank, A., 2000. Injuries related to shift work. American Journal of Preventive Holmstrom, E., Ahlborg, B., 2005. Morning warming-up exercise: effects on
Medicine 18, 33–36. musculoskeletal fitness in construction workers. Applied Ergonomics 36, 513–
Freivalds, A., Johnson, A., 1990. Time series analysis of industrial accident data. 519.
Journal of Occupational Accidents 13, 179–193. Hopkins, A., 2006. What are we to make of safe behavior programs? Safety Science
Fullarton, C., Stokes, M., 2007. The utility of a workplace injury instrument in 44, 583–597.
prediction of work place injury. Accident Analysis and Prevention 39, 28–37. Huang, Y., Ho, M., Smith, G., Chen, P., 2006. Safety climate and self-reported injury:
Galli, E., 1999. A sociological case study of occupational accidents in the Brazilian assessing the mediating role of employee safety control. Accident Analysis and
petrochemical industry. Accident Analysis and Prevention 31, 297–304. Prevention 38, 425–433.
Genaidy, A., Karwowski, W., A-Rehim, A., 2007a. The work compatibility Hull, B., Leigh, J., Driscoll, T., Mandryk, J., 1996. Factors associated with occupational
improvement framework: preliminary findings of a case study for defining injury severity in the New South Wales underground coal mining industry.
and measuring the human-at-work system. Ergonomics 50, 1771–1808. Safety Science 21, 191–204.
Genaidy, A., Rinder, M., A-Rehim, A., 2008. The work compatibility improvement ICECI, 2004. International Classification of External Causes of Injuries. ICECI
framework: an assessment of the worker-work environment interaction in the Coordination and Maintenance Group, World Health Organization.
manufacturing sector. Ergonomics 51, 1195–1218. ILO, 1988. Major Hazard Control – A Practical Manual. International Labour
Genaidy, A., Salem, S., Karwowski, W., Paez, O., Tuncle, S., 2007b. The work Organization, Geneva.
compatibility improvement framework: an integrated perspective of the Irwin, J., 1964. The personal factor in accidents- a review article. Journal of Royal
human-at-work system. Ergonomics 50, 3–25. Statistical Society- A 127, 438–451.
Genaidy, A., Sequeira, R., Rinder, M., A-Rehim, A., 2009. Determinants of business Jeong, B., 1999. Comparisons of variables between fatal and nonfatal accidents in
sustainability: an ergonomics perspective. Ergonomics 52, 273–301. manufacturing industry. International Journal of Industrial Ergonomics 23,
Geneidy, A., 1990. Physical ergonomics job design: an accident prevention approach 565–572.
to control upper extremity cumulative trauma disorders in manufacturing Johnson, W., 1980. MORT Safety Assurance Systems. National Safety Council, New
industry. Journal of Occupational Accidents 13, 303–320. York.
Gillen, M., Baltz, D., Gassel, M., Kirsch, L., Vaccaro, D., 2002. Perceived safety climate, Keating, C., Fernandez, A., Jacobs, D., Kauffmann, P., 2001. A methodology for
job demands, and coworker support among union and nonunion injured analysis of complex sociotechnical processes. Business Process Management
construction workers. Journal of Safety Research 33, 33–51. Journal 7, 33–49.
Greenwood, M., Woods, H., 1919. The incidence of industrial accidents upon Kee, D., Karwowski, W., 2001. LUBA: an assessment technique for postural loading
individuals with special reference to multiple accidents. Tech. rep., Industrial on the upper body based on joint motion discomfort and maximum holding
Fatigue Research Board, London. time. Applied Ergonomics 32, 357–366.
Greenwood, M., Yule, G., 1920. An inquiry into the nature of frequency distributions Keehn, J., 1959. Factor analysis of reported minor personal mishaps. Journal of
representative of multiple happenings, with particular reference to the Applied Psychology 43, 311–314.
occurrence of multiple accidents or disease or repeated accidents. Journal of Kejriwal, B., 2002. Safety in Mines: A Survey of Accidents, Their Causes and
Royal Statistical Society 83, 255. Prevention. Lovely Prakashan, Dhanbad.
Guastello, D., Guastello, S., 1987. The relationship between work group size and Kerr, W., 1950. Accident proneness of factory departments. Journal of Applied
occupational accidents. Journal of Occupational Accidents 9, 1–9. Psychology 34, 167–170.
Guilford, J., 1973. Prediction of accidents in a standardized home environment. Kerr, W., 1957. Complementary theories of safety psychology. Journal of Social
Journal of Applied Psychology 57, 306–313. Psychology 45, 3–9.
Gun, R., 1993. The role of regulations in the prevention of occupational injury. Keyserling, W., 1983. Occupational injuries and work experience. Journal of Safety
Safety Science 16, 47–66. Research 14, 37–42.
Gun, R., Ryan, C., 1994. A case-control study of possible risk factors in the causation Khan, F., Amyotte, P., DiMattia, D., 2006. HEPI: a new tool for human error
of occupational injury. Safety Science 18, 1–13. probability calculation for offshore operation. Safety Science 44, 313–334.
Haddon, W., 1968. The changing approach to the epidemiology, prevention, and Khanzode, V., Maiti, J., Ray, P., 2011. A methodology for evaluation and monitoring
amelioration of trauma: the transition to approaches ethologically rather than of recurring hazards in underground coal mines. Safety Science 49, 1172–1179.
descriptively based. American Journal of Public Health 58, 1431–1438. Khanzode, V., Maiti, J., Ray, P., Tewari, V., 2010. Injury severity assessment for
Haddon, W., 1972. A logical framework for categorizing highway safety underground coalmine workers. Applied Ergonomics 41, 242–250.
phenomenon and activity. Journal of Trauma 12, 193–207. Kirchner, W., 1961. The fallacy of accident proneness. Personnel 38, 34–37.
Haddon, W., 1980. The basic strategies for reducing damage from hazard of all Kirwan, B., 1990. A practical ergonomics methodology. In: Wilson, J., Corlett, E.
kinds. Hazard Prevention 16, 8–12. (Eds.), Human Reliability Assessment. Taylor and Francis, London.
Haddon, W., Suchman, E., Klein, D., 1964. Accident Research: Methods and Kjellen, U., 1982. An evaluation of safety information systems at six medium-sized
Approaches. Harper & ROW, New York. and large firms. Journal of Occupational Accidents 3, 273–288.
1366 V.V. Khanzode et al. / Safety Science 50 (2012) 1355–1367

Kjellen, U., 1984a. The deviation concept in occupational accident control – I: Mason, A., October 1973. How to determine your safety and health training needs.
definitions and classification. Accident Analysis and Prevention 16, 289–306. In: Proceedings of 28th Annual Federal Safety Conference. Chicago, Illinois.
Kjellen, U., 1984b. The deviation concept in occupational accident control – II: data McAtamney, L., Corlett, E., 1993. RULA: a survey method for the investigation of
collection and assessment of significance. Accident Analysis and Prevention 16, work-related upper limb disorders. Applied Ergonomics 24, 91–99.
307–323. McKenna, F., 1983. Accident proneness: a conceptual analysis. Accident Analysis
Kjellen, U., 1984c. The role of deviations in accident causation and control. Journal and Prevention 15, 65–71.
of Occupational Accidents 6, 117–126. McLeod, R., Stockwell, T., Rooney, R., Stevens, M., Phillips, M., Jelinek, G., 2003. The
Kjellen, U., 1987. Simulating the use of a computerized injury and near accident influence of extrinsic and intrinsic risk factors on the probability of sustaining
information system in decision making. Journal of Occupational Accidents 9, an injury. Accident Analysis and Prevention 35, 71–80.
87–105. Menckel, E., 1990. Safety engineers and accident prevention- an inventory of
Kjellen, U., 1995. Integrating analyses of the risk of occupational accidents into the activities within one industrial sector in Sweden. Journal of Occupational
design process – part II: method for prediction of the LTI-rate. Safety Science 19, Accidents 12, 271–282.
3–18. Mitchell, C., Cloeren, M., Schwartz, B., 1993. Application of an injury surveillance
Kjellen, U., Hovden, J., 1993. Reducing risks by deviation control – a retrospection system to injuries at an industrial facility. Accident Analysis and Prevention 25,
into a research strategy. Safety Science 16, 417–438. 453–458.
Kjellen, U., Larsson, T., 1981. Investigating accidents and reducing risks- a dynamic Mol, T., 2003. Productive Safety Management. Butterworth Heinemann.
approach. Journal of Occupational Accidents 3, 129–140. Morel, G., Chauvin, C., 2006. A socio-technical approach of risk management applied
Kjellen, U., Sklet, S., 1995. Integrating analyses of the risk of occupational accidents to collisions involving fishing vessels. Safety Science 44, 599–619.
into the design process – part I: a review of types of acceptance criteria and risk Mullen, J., 2004. Investigating factors that influence individual safety behavior at
analysis methods. Safety Science 18, 215–227. work. Journal of Safety Research 35, 275–285.
Kleiner, B., 2006. Macroergonomics: analysis and design of work systems. Applied Nieuwhof, G., 1985. Risk: a probabilistic concept. Reliability Engineering 10, 183–
Ergonomics 37, 81–89. 188.
Kletz, T., 1990. The need for friendly plants. Journal of Occupational Accidents 13, 3– Oleske, D., Brewer, R., Doan, P., Hahn, J., 1989. An epidemiologic evaluation of the
13. injury experience of a cohort of automative parts workers: a model for
Krause, T., Seymour, K., Sloat, K., 1999. Long-term evaluation of a behavior-based surveillance in small industries. Journal of Occupational Accidents 10, 239–253.
method for improving safety performance: a meta-analysis of 73 interrupted OSHA, 2002. Small Business Handbook. OSHA-2209-02R.
time-series replications. Safety Science 32, 1–18. O’Toole, M., 1999. Successful safety committees: participation not legislation.
Kumamoto, H., Henley, E., 1996. Probabilistic Risk Assessment and Management for Journal of Safety Research 30, 39–65.
Engineers and Scientists, second ed. IEEE Press. O’Toole, M., 2002. The relationship between employees’ perceptions of safety and
Kunce, J., 1967. Vocational interests and accident proneness. Journal of Applied organizational culture. Journal of Safety Research 33, 231–243.
Psychology 51, 223–225. Papazoglou, I., Ale, B., 2007. A logical model for quantification of occupational risk.
Kunce, J., 1974. SVIB scores and accident proneness. Measurement and Evaluation in Reliability Engineering and System Safety 92, 785–803.
Guidance 7, 118–121. Paul, P., Maiti, J., 2005. Development and test of a sociotechnical model for accident/
Laamme, L., Menckel, E., 1995. Aging and occupational accidents: a review of the injury occurrences in underground coalmines. The Journal of the South African
literature of the last three decades. Safety Science 21, 145–161. Institute of Mining and Metallurgy 105, 1–12.
Langley, J., 1988. The need to discontinue the use of the term ‘‘accident’’ when Paul, P., Maiti, J., 2007. The role of behavioral factors on safety management in
referring to unintentional injury events. Accident Analysis and Prevention 20, underground mines. Safety Science 45, 449–471.
1–8. Paul, P., Maiti, J., 2008. The synergic role of sociotechnical and personal
Laurence, D., 2005. Safety rules and regulations on mine sites: the problem and a characteristics on work injuries in mines. Ergonomics 51, 737–767.
solution. Journal of Safety Research 36, 39–50. Paul, P., Maiti, J., Dasgupta, S., Forjuoh, S., 2005. An epidemiological study of injury
Leamon, T., 1980. The organization of industrial ergonomics: a human-machine in mines: implications for safety promotion. Injury Control and Safety
model. Applied Ergonomics 11, 223–226. Promotion 12, 157–165.
Leamon, T., 1988. Human-workplace model: an integrated approach to Perrow, C., 1984. Normal Accidents: Living with High-risk Technologies, second ed.
occupational injury and illness. International Journal of Industrial Ergonomics Princeton University Press.
2, 211–219. Petersen, D., 1989. Techniques of Safety Management: A Systems Approach, third
Lees, F., 1980. Loss Prevention in the Process Industries. Butterworth and Co Ltd, ed. Aloray Goshen, New York.
London. Pines, A., Halfon, S., Prior, R., 1987. Occupational accidents in the construction
Leigh, J., Mulder, H., Want, G., Farnsworth, N., Morgan, G., 1990. Personal and industry of Israel. Journal of Occupational Accidents 9, 225–243.
environmental factors in coal mining accidents. Journal of Occupational Pohjola, V., 2003. Fundamentals of safety conscious process design. Safety Science
Accidents 13, 233–250. 41, 181–218.
Leplat, J., 1984. Occupational accident research and systems approach. Journal of Powell, P.I., Hale, M., Martin, J., Simon, M., 1971. Accidents: A Shop OOR Study of
Occupational Accidents 6, 77–89. Their Causes. National Institute for Industrial Psychology, London.
Leveson, N., 2004. A new accident model for engineering safer systems. Safety Prussia, G., Brown, K., Willis, G., 2003. Mental models of safety: do managers and
Science 42, 237–270. employees see eye to eye? Journal of Safety Research 34, 143–156.
Levin, L., Oler, J., Whiteside, J., 1985. Injury incident rates in a paint company on Purswell, J., Rumar, K., 1984. Occupational accident research: where we have been
rotating production shifts. Accident Analysis and Prevention 17, 67–73. and where are we going? Journal of Occupational Accidents 6, 219–228.
Lindell, M., 1997. Occupational safety and health inspection scores predict rates of Putz-Anderson, V., 1988. Cumulative Trauma Disorders: A Manual for
workers’ lost time injuries. Accident Analysis and Prevention 29, 563–571. Musculoskeletal Diseases of the Upper Limbs. Taylor and Francis, London.
Lozada-Larsen, S., Laughery, K., 1987. Do identical circumstances precede minor and Rao-Tummala, V., Leung, Y., 1996. A risk management model to assess safety and
major injuries? In: Rising to New Heights with Technology, vol. 1. Human reliability risks. International Journal of Quality and Reliability Management 13,
Factors Society, New York, pp. 200–204. 53–62.
Maiti, J., 2005. Risk assessment and safety evaluation of mining system. Journal of Rasmussen, J., 1982. Human errors: a taxonomy for describing human malfunction
Institution of Engineers (India) – Mining 85, 33–41. in industrial installations. Journal of Occupational Accidents 4, 311–333.
Maiti, J., 2010. Development of worksystem safety capability index (WSCI). Safety Rasmussen, J., 1997. Risk management in a dynamic society: a modeling problem.
Science 48, 1369–1379. Safety Science 27, 183–213.
Maiti, J., Bhattacharjee, A., 2000. A causal model for evaluation of mine safety. Reason, J., 1990. Human Error. Cambridge University Press.
Transaction of the Institute of Mining and Metallurgy (Section A – Mining Reason, J., 1997. Managing the Risks of Organizational Accidents. Adershot, UK.
Technology) 109, A55–A59. Reiman, T., Oedewald, P., 2007. Assessment of complex sociotechnical systems:
Maiti, J., Bhattacherjee, A., 1999. Evaluation of risk of occupational injuries among theoretical issues concerning the use of organizational culture and
underground coal mine workers through multinomial logit analysis. Journal of organizational core task concepts. Safety Science 45, 745–768.
Safety Research 30, 93–101. Robertson, L., 1998. Injury Epidemiology: Research and Control Strategies. Oxford
Maiti, J., Bhattacherjee, A., Bangdiwala, S., 2001. Loglinear model for analysis of University Press, New York.
cross-tabulated coal mine injury data. Injury Control and Safety Promotion 8, Robins, S., 2000. Organization Theory: Structure, Design, and Applications. Prentice
229–236. Hall of India, New Delhi.
Maiti, J., Chatterjee, S., Bangdiwala, S., 2004. Determinants of work injuries in Robinson, G., 1982. Accidents and sociotechnical systems: principles for design.
mines: an application of structural equation modelling. Injury Control and Accident Analysis and Prevention 14, 121–130.
Safety Promotion 11, 29–37. Rundmo, T., 1996. Associations between risk perception and safety. Safety Science
Maiti, J., Khanzode, V., 2009. Development of a relative risk model for roof and side 24, 197–209.
fall fatal accidents in underground coal mines in India. Safety Science 47, 1068– Rundmo, T., Hale, A., 2003. Managers’ attitudes towards safety and accident
1076. prevention. Safety Science 41, 557–574.
Maiti, J., Khanzode, V., Ray, P., 2009. Severity analysis of Indian coal mine accidents- Saari, J., 1982. Accidents and progress of technology in Finnish industry. Journal of
a retrospective study for 100 years. Safety Science 47, 1033–1042. Occupational Accidents 4, 133–144.
Majchrzak, A., Borys, B., 2001. Generating testable socio-technical systems theory. Salminen, S., 2004. Have young workers more injuries than older ones? An
Journal of Engineering Technology Management 18, 219–240. international literature review. Journal of Safety Research 35, 513–521.
Mallick, S., Mukherjee, K., 1996. An empirical study for mines safety management Salminen, S., Heiskanen, M., 1997. Correlations between traffic, occupational, sports
through analysis on potential for accident reduction. Omega 24, 539–550. and home accidents. Accident Analysis and Prevention 29, 33–36.
 V.V. Khanzode et al. / Safety Science 50 (2012) 1355–1367 1367

Salminen, S., Saari, J., Saarela, K., Rasanen, T., 1992. Fatal and non-fatal occupational Svedung, I., Rasmussen, J., 2002. Graphic representation of accident scenarios:
accidents: identical versus differential causation. Safety Science 15, mapping system structure and the causation factors. Safety Science 40, 397–
109–118. 417.
Sari, M., Duzgun, H., Karpuz, C., Selcuk, A., 2004. Accident analysis of two Turkish Taylor, J., 1979. A Background to Risk Analysis. Electronics Department, RisØ
underground coal mines. Safety Science 42, 675–690. National Laboratory, Denmark.
Schulzinger, M., 1954. A closer look at accident proneness. National Safety News 69, Tayyari, F., Smith, J., 1997. Occupational Ergonomics: Principles and Applications.
32–33. Kluwer Academic Publishers, Boston.
Seo, D., 2005. An explicative model of unsafe work behavior. Safety Science 43, 187– Tixier, J., Dusserre, G., Salvi, O., Gaston, D., 2002. Review of 62 risk analysis
211. methodologies of industrial plants. Journal of Loss Prevention in the Process
Shannon, H., Manning, D., 1980. Differences between lost-time and non-lost time Industries 15, 291–303.
industrial accidents. Journal of Occupational Accidents 2, 265–272. Torma-Krajewski, J., Steiner, L., Lewis, P., Gust, P., Johnson, K., 2007. Implementation
Shaw, L., Sichel, H., 1971. Accident proneness. International Series of Monographs in of an ergonomic process at a US surface coal mine. International Journal of
Experimental Psychology, vol. II. Pergamon Press, Oxford. Industrial Ergonomics 37, 157–167.
Sheu, J., Hwang, J., Wang, J., 2000. Diagnosis and monetary quantification of Trist, E., Bamforth, K., 1951. Some social and psychological consequences of the
occupational injuries by indices related to human capital loss: analysis of a steel long-wall method of coal-getting. Human Relations 4, 3–38.
company as an illustration. Accident Analysis and Prevention 32, 435–443. Tuominen, R., Saari, J., 1982. A model for analysis of accidents and its application.
Singleton, W., 1982. Accidents and progress of technology. Journal of Occupational Journal of Occupational Accidents 4, 263–273.
Accidents 4, 91–102. Verhaegen, P., Strubbe, J., Vonck, R., van den Abeele, J., 1985. Absentism, accidents
Singleton, W., 1984a. Application of human error analysis to occupational accident and risk taking. Journal of Occupational Accidents 7, 177–186.
research. Journal of Occupational Accidents 6, 107–115. Verhaegen, P., Vanhalst, B., Derijcke, H., Hoecke, M.V., 1976. The value of some
Singleton, W., 1984b. Future trends in accident research in European countries. psychological theories on industrial accidents. Journal of Occupational
Journal of Occupational Accidents 6, 3–12. Accidents 1, 39–45.
Siu, O., Phillips, D., Leung, T., 2004. Safety climate and safety performance among Visser, E., Pijl, Y., Stolk, R., Neeleman, J., Rosmalen, J., 2007. Accident proneness, does
construction workers in Hong Kong- the role of psychological strains as it exist? A review and meta-analysis. Accident Analysis and Prevention 39, 556–
mediators. Accident Analysis and Prevention 36, 359–366. 564.
Smillie, R., Ayoub, M., 1976. Accident causation theories: a simulation approach. Warr, P., 1993. In what circumstances does job performance vary with age?
Journal of Occupational Accidents 1, 47–68. European Work and Organizational Psychologist 3, 237–249.
Smith, G., Huang, Y., Ho, M., Chen, P., 2006. The relationship between safety climate Webster, 2002. Webster’s Third New International Dictionary: Unabridged.
and injury rates across industries: the need to adjust for injury hazards. Merriam Webster Inc.
Accident Analysis and Prevention 38, 556–562. Wehmeier, S., McIntosh, C., Turnbull, J., Ashby, M., 2005. Oxford Advanced Learner’s
Smith, M., Sainfort, P., 1989. A balance theory of job design for stress reduction. Dictionary of Current English, seventh ed. Oxford University Press.
International Journal of Industrial Ergonomics 4, 67–79. Whitlock, G., Clouse, R., Spencer, W., 1963. Predicting accident proneness. Personnel
Soloman, K., Alesch, K., 1989. The index of harm: a measure for comparing Psychology 16, 35–44.
occupational risk across industries. Journal of Occupational Accidents 11, 19– Whysall, Z., Haslam, C., Haslam, R., 2006. A stage of change approach to reducing
35. occupational ill health. Preventive Medicine 43, 422–428.
Sorock, G., Lombardi, D., Courtney, T., Cotnam, J., Mittleman, M., 2001. Epidemiology Williams, J., Scott-Geler, E., 2000. Behavior-based intervention for occupational
of occupational acute traumatic hand injuries: a literature review. Safety safety: critical impact of social comparison feedback. Journal of Safety Research
Science 38, 241–256. 31, 135–142.
Sorock, G., Lombardi, D., Hauser, R., Eisen, E., Herrick, R., Mittleman, M., 2004. A Willquist, P., Torner, M., 2003. Identifying and analysing hazards in manufacturing
case-crossover study of transient risk factors for occupational acute hand injury. industry- a review of selected methods and development of a framework for
Occupational Environmental Medicine 61, 305–311. method applicability. International Journal of Industrial Ergonomics 32, 165–
Storbakken, R., 2002. An Incident Investigation Procedure for Use in Industry. 180.
Master’s thesis, The Graduate School, University of Wisconsin-Stout, Woodcock, K., Drury, C., Smiley, A., Ma, J., 2005. Using simulated investigations for
Menomonie, WI. accident investigation studies. Applied Ergonomics 36, 1–12.
Sulzer-Azaroff, B., 1987. The modification of occupational safety behavior. Journal of Zohar, D., 1980. Safety climate in industrial organizations: theoretical and applied
Occupational Accidents 9, 177–197. implications. Journal of Applied Psychology 65, 96–101.
Suokas, J., 1988. The role of safety analysis in accident prevention. Accident Analysis Zohar, D., 2010. Thirty years of safety climate research: reflections and future
and Prevention 20, 67–85. directions. Accident Analysis and Prevention 42, 1517–1522.
Suokas, J., Rouhiainen, V., 1984. Work Safety Analysis: Method Description and Zurada, J., Karwowski, W., Marras, W., 1997. A neural network-based system for
User’s Guide. Technical Research Center of Finland, Espoo, Finland. classification of industrial jobs with respect to risk of low back disorders due to
Surry, J., 1969. Industrial Accident Research: A Human Engineering Appraisal. Tech. workplace design. Applied Ergonomics 28, 49–58.
rep., Labour Safety Council, Ministry of Labour, Government of Canada. Zwerling, C., Sprince, N., Wallace, R., Davis, C., Whitten, P., Heering, S., 1996. Alcohol
Susman, G., 1976. Autonomy at Work: A Sociotechnical Analysis of Participative and occupational injuries among older workers. Accident Analysis and
Management. Praeger Publishers, New York. Prevention 28, 371–376.