A systematic review of safety violations in industry

Accident Analysis and Prevention 41 (2009) 739–754

Contents lists available at ScienceDirect

Accident Analysis and Prevention journal homepage: www.elsevier.com/locate/aap

A systematic review of safety violations in industry Samuel J. Alper a , Ben-Tzion Karsh b,∗ a b

University of Wisconsin-Madison, 3217 Mechanical Engineering Building, 1513 University Avenue, Madison, WI 53706, USA University of Wisconsin-Madison, 3218 Mechanical Engineering Building, 1513 University Avenue, Madison, WI 53706, USA

a r t i c l e

i n f o

Article history: Received 11 November 2008 Received in revised form 24 February 2009 Accepted 22 March 2009 Keywords: Violation Non-compliance Work-around Circumvention System safety Macroergonomics

a b s t r a c t It is widely known that intentional non-malevolent violations of safety procedures and norms occur and evidence shows that safety violations can increase the risk of accidents. However, little research about the causes of these violations in work settings exists. To help shed light on the causes, this paper systematically reviews the empirical causes of safety violations in industry. Electronic database literature searches were performed to identify relevant articles published prior to January 1, 2007. Thirteen articles met the inclusion criteria and 57 different variables were examined as predictors of safety violations. Study settings were healthcare delivery, commercial driving, aviation, mining, railroad, and construction. The predictors were categorized into individual characteristics, information/education/training, design to support worker needs, safety climate, competing goals, and problems with rules. None of the reviewed studies examined whether violations can improve system performance or safety. Methodological suggestions and a macroergonomic framework are offered for improving future studies of the epidemiology of safety violations. © 2009 Elsevier Ltd. All rights reserved.

1. Introduction In his seminal work Human Error, Reason (1990) noted “an important lesson to be learned . . . is that the term ‘error’ does not capture all the ways in which human beings contribute to accidents” (p. 194). The other contribution mechanism to which Reason was referring was through what is commonly referred to as a violation. Violations of safety rules, procedures and norms have now been the subject of extensive research, especially among recreational drivers (“recreational drivers” in this paper refers to drivers who are not driving as part of their employment) (e.g. Aberg and Rimmo, 1998; Blockey and Hartley, 1995; Parker et al., 1995; Reason et al., 1990). However, there is a limited amount of research literature that investigates rule violations in work settings, and still less exists in work settings where the causes of violations are studied. This is an alarming gap in the literature considering that some industries estimate that about 70% of their total accidents can be attributed to violations (Mason, 1997). Perhaps part of the reason that the causes of violations have not been studied in detail stems from the idea that violations are actions taken by ‘bad’ people. That the term “violation” is common and that people hold their own colloquial definition for the concept may also have contributed to the lack of systematic research

∗ Corresponding author. Tel.: +1 608 262 3002; fax: +1 608 262 8454. E-mail addresses: [email protected] (S.J. Alper), [email protected] (B.-T. Karsh). 0001-4575/$ – see front matter © 2009 Elsevier Ltd. All rights reserved. doi:10.1016/j.aap.2009.03.013

to understand both the concept and causes of violations. This paper aims to provide a specific definition for the term “violation.” Then, this paper seeks to systematically review what is currently known about the causes of safety violations in industry because efforts to secure or increase compliance with safety protocols and norms need an evidence base to guide intervention efforts. Armed with a specific definition and existing evidence about the causes of violations, this paper then re-frames violations as evidence of system problems rather than as actions taken by ‘bad’ workers. 2. Safety violations research What is known to date about violations is that they certainly exist and that some evidence suggests they can lead to unwanted outcomes such as accidents. In studies of recreational driving, the existence of violations has been demonstrated using drivers’ selfreports (e.g. Kanellaidis et al., 1995; Shinar et al., 2001), frequently using a variation of Reason et al.’s (1990) Driver Behavior Questionnaire (e.g. Aberg and Rimmo, 1998; Blockey and Hartley, 1995; Parker et al., 1995). Driving violations have further been documented during field observations (Hakkert et al., 2001; Porter and England, 2000; Retting and Williams, 1996), and uncovered through analysis of existing fatality and vehicle registration databases (Retting and Williams, 1996; Romano et al., 2006). Safety violations have also been documented in industry. In healthcare, for example, observational (Alper et al., 2008; Kobayashi et al., 2005; Patterson et al., 2002, 2006), survey (Alper et al., 2006; McKeon et al., 2006), and violation-reporting (Horning and Smith,

740

S.J. Alper, B.-T. Karsh / Accident Analysis and Prevention 41 (2009) 739–754

1991) methods have all been used to document the existence of violations. In addition, violations have been documented in aviation maintenance (e.g. Hobbs and Williamson, 2002; Wenner and Drury, 2000), mining (Laurence, 2005), railroad (Lawton, 1998), and other industries. 3. Violations: good or bad for safety? Safety violations clearly exist in industry, but it remains unclear under what circumstances violations help or hurt safety. In recreational driving, it has been established that violations can lead to outcomes such as accidents (e.g. Özkan and Lajunen, 2005; Parker et al., 1995; Reason et al., 1990), though the link between safety violations and unwanted outcomes is not as firmly established in industrial settings. If violations can lead to accidents, it provides the first reason to study the causes of violations: control the causes of violations to reduce accidents. However, violations do not necessarily lead to unwanted outcomes such as accidents and injuries. Part of the reason that a strong link between violations and unwanted outcomes has not been firmly established may be that most violations do not lead to harm. As early as 1931, Heinrich offered the idea that for every 300 accidents that do not result in an injury there are 29 minor injuries and 1 major injury (Heinrich, 1931). These accidents and injuries result from a much larger pool of errors and violations. That is, most errors and violations are not noticed because they do not result in unwanted outcomes. Despite this, there is reason to believe that safety violations may put a work system into a more vulnerable state (e.g. Reason et al., 1995, 1998). Thus, the second reason to study the causes of violations is that violations may put systems into more unsafe states, increasing their risk for an unwanted outcome. Beyond not knowing to what extent violations contribute to unwanted outcomes is the fact that it is unclear whether or not all violations are ‘wrong’ behaviors. The term “violation” certainly evokes a feeling that someone did something wrong, but when safety rules are not appropriate, violations may increase system safety (Almaberti et al., 2006; Besnard and Greathead, 2003; Reason et al., 1998). In some cases then, violations can be thought of as micro-level resilience (Hollnagel et al., 2006) where resilience is “the characteristic of managing the organization’s activities to anticipate and circumvent threats to its existence and primary goals” (Hale and Heijer, 2006, p. 35). A violation may occur when an individual, realizing that a system is in jeopardy, takes actions that are outside of normal operation to save the system. In such cases, the violations may not only improve safety, but may eventually be considered “best practice” in the situations that produced them. This gives rise to the third reason to study the causes of violations: to understand what system parameters lead to situations that require violations to maintain system safety.

All three reasons for studying the causes of violations make clear that simply blaming individuals for violating a safety policy or norm is an insufficient approach to improving safety. Unfortunately, the norm in industry seems to be to blame people for violations if a bad outcome arises, but otherwise to tacitly or explicitly approve violations (Koppel et al., 2008). But as explained, when evaluating violations, we must consider the situation the individual faced when the decision was made to violate without considering the outcomes of the action (Reason, 1998). Rather than viewing violations as a ‘bad’ individual’s actions, they could instead be viewed as an indication that a company’s rules do not meet the demands of the situations workers encounter while working. Thus, characteristics of the work system may be causes of violations. That is the question addressed by this review: what are the causes of safety violations? 4. Definitions The first step in the review was to agree upon a definition of a safety violation. Table 1 provides a list of definitions uncovered in the literature. The fact that there is variety among the definitions can lead to different conceptualizations of “violations.” The differences between definitions used in the literature as well as the absence of definitions in much of the literature necessitate a more unified approach to the definition of violations. The definitions uncovered in the literature share three similarities. First, each definition specifies that there must exist rules, guidelines, protocols, or norms to be violated. Second, a violation involves some action that is contrary to these rules, guidelines, protocols, or norms. Third, violations are intentional actions. However, even if an action is unintentional, if it is contrary to a rule it can be considered a violation. In this way, past definitions have overspecified the concept of violations. Therefore, a definition that could be used for the concept of “violation” is: an action that is contrary to a rule. This is a parsimonious and general definition that serves as a blanket for all violations. However, most violation research requires a more specific focus. This has been provided in the literature by differentiating between ‘types’ of violations. Fig. 1 depicts graphically how violation types have been distinguished (Reason, 1990). When an individual violates a rule unintentionally, it is known as an erroneous violation (Lawton, 1998; Reason, 1990) or an unintended violation (Reason et al., 1990). Erroneous/unintended violations can occur due to human error (e.g. Reason, 1990), but can also occur if an individual simply does not know the rule governing the actions (Lawton, 1998). In general, research about violations has focused more on intentional violations than on erroneous/unintended violations (e.g. Lawton, 1998; Parker et al., 1995;

Table 1 Definitions of violations that have been used in previous research. Author

Definition

Beatty and Beatty (2004) Lawton (1998)

“Intentional acts contrary to advice or best practice guidelines” (p. 528) “Deliberate departures from rules that describe the safe or approved method of performing a particular task or job” (p. 78) “Violations can be defined as any deliberate deviations from the rules, procedures, instructions or regulations introduced for the safe or efficient operation and maintenance of equipment” (p. 288) “Violations . . . may be defined as the deliberate infringement of some regulated or socially accepted code of behaviour” (p. 1036) “Deliberate – but not necessarily reprehensible – deviations from those practices deemed necessary (by designers, managers and regulatory agencies) to maintain the safe operation of a potentially hazardous system” (p. 195) “Violations can be defined as deliberate (though not necessarily reprehensible) deviations from those practices deemed necessary (by designers, managers, and regulatory agencies) to maintain the safe operation of a potentially hazardous system” (p. 1316) “Violations are . . . the deliberate deviation of actions from safe operating procedures” (p.1715) “Violations are deviations from safe operating procedures, standards, or rules” (p. 292)

Mason (1997) Parker et al. (1995) Reason (1990) Reason et al. (1990)

Reason et al. (1995) Reason et al. (1998)

S.J. Alper, B.-T. Karsh / Accident Analysis and Prevention 41 (2009) 739–754

741

Fig. 1. Categorization of violations based on Reason’s (1990) work.

Reason, 1990; Reason et al., 1998); violations are frequently discussed only as intentional actions (e.g. Beatty and Beatty, 2004; Mason, 1997; Reason et al., 1995). As Fig. 1 shows, intentional violations can be further broken down into those committed with and without intention to cause harm or damage the system (Beatty and Beatty, 2004; Parker et al., 1995; Reason, 1990). Violations that are committed with the intention of causing harm, damage to the system, or bad outcomes have been labeled sabotage (Lawton, 1998; Reason, 1990; Reason et al., 1990), terrorist action (Lawton, 1998), or malevolent acts (Beatty and Beatty, 2004). In most cases, however, violations are not undertaken with the intention to cause harm, damage the system, or produce bad outcomes (e.g. Mason, 1997; Reason et al., 1995). Such violations have received the most attention in the safety literature on violations (Mason, 1997; Parker et al., 1995; Reason, 1990). Therefore, this review specifically focuses on intentional, non-malevolent violations.

5. Methodology 5.1. Inclusion and exclusion criteria Inclusion criteria for articles in the systematic review were the following: (1) the study was empirical with a substantive focus (operationalized as articles that addressed violations in their title or in their abstract) on identifying variables that influence or predict violations of safety rules, (2) the subjects were employees and the violations were work-related, (3) the article was published prior to January 1, 2007, (4) the article was available either online or though the University of Wisconsin-Madison library and its interlibrary loan system, (5) the article was published in a refereed journal, and (6) the article was written in the English language.

Articles that discussed violations of handwashing in healthcare and articles that discussed violations of universal precautions in healthcare were excluded. While many articles on these topics met the inclusion criteria, they have already been the subjects of multiple reviews. For reviews that discuss why healthcare workers violate handwashing protocols see Bisset (2002); Gould and co-workers (Gould et al., 2007a,b); Kretzer and Larson (1998); Maskerine and Loeb (2006); Pittet (2001a,b); or Storr and ClaytonKent (2004). For reviews that discuss why healthcare workers violate universal precautions, see Beekmann and Henderson (2005) or Moore et al. (2005). Specific examples of ways articles did not meet the first two inclusion criteria included:

• the article did not report causes of violations. Many empirical papers either discussed attitudes toward violations (Lawton and Parker, 2002; Parker and Lawton, 2000) or only identified the existence of violations or provided prevalence or frequency data about violations (e.g. Horning and Smith, 1991; Kobayashi et al., 2005; McDonald, 2006); • the violation discussed in the article was not safety related, which we defined to mean there was a possibility of injury or illness. For example, violation of internet security protocols (Kraemer and Carayon, 2007), non-compliance with a survey request (Rogelberg et al., 2000), and non-compliance with journal article reprint requests (Ellis and Curless, 1985) were not included because they do not threaten injury or illness; • the subjects of the study were not employees. For example, lab studies (Zeitlin, 1994) and studies of recreational driving (e.g. Özkan et al., 2006; Parker et al., 1995; Porter and England, 2000; Reason et al., 1990) do not use employees as subjects. Other topics excluded based on this rationale included law violations, human

742

S.J. Alper, B.-T. Karsh / Accident Analysis and Prevention 41 (2009) 739–754

rights violations, ethics violations, and patients not complying with their prescribed treatment. 5.2. Procedures Articles for inclusion were identified using three methods: database searches, a search of references, and a cited reference search. 5.2.1. Database searches The primary method for identifying relevant articles was a search of online article databases using a single Boolean search string. First, searches were conducted in ISI Web of Knowledge using the search string “violati* or shortcut* or work-around* or non-complian*”, where “*” is the wildcard term used for truncation. The search resulted 31,598 results. To narrow the results, 50 subject topics that were unlikely to contain articles that met the review criteria were excluded. For example, the 7870 articles in the subject area “Physics, Particles, and Fields” and the 5671 articles in “Physics, Multidisciplinary” were excluded. There remained 9180 articles. Nine articles from this search met the inclusion criteria. Second, an additional round of database searching was completed. The databases searched were ABI/Inform, Business Source Elite, PSYCInfo, PsycARTICLES, Social Sciences Full Text, Pub Med, CINAHL, Health and Safety Science Abstracts, Proquest, and Annual Reviews. When possible, terms from the thesaurus associated with the database were used. The search yielded 3284 articles published prior to January 1, 2007. Fifty articles’ abstracts indicated that the article might meet the inclusion criteria, and they were read in their entirety. Of these, four met the inclusion criteria. 5.2.2. Search of references The citations of the articles that met the inclusion criteria were reviewed. Those appearing to meet the inclusion criteria based on their title were read in their entirety. No new articles met the inclusion criteria.

and worker participation in safety programs. “Competing goals” included situations in which the worker had goals that were competing with safety. Finally, “problems with rules” included situations in which rules were poorly designed. Two variables did not fit into one of the six categories, and are discussed as part of an ‘other’ category. These six categories do not represent definitive categories; some of the variables could fit into multiple categories. The choice of category in such cases was made based on how the variable was operationalized in its study. 6. Results Table 2 provides a summary of the 13 articles that met the inclusion criteria. 6.1. Setting of study Five of the studies were in healthcare (Beatty and Beatty, 2004; Espin et al., 2006; McKeon et al., 2006; Patterson et al., 2006; Taxis and Barber, 2003), two were in commercial driving (Beilock, 1995; Wills et al., 2006), two were in aviation (Li and Baker, 1995; Rebok et al., 2005), one was in aviation maintenance (Hobbs and Williamson, 2002), one studied mine workers (Laurence, 2005), one studied railroad workers (Lawton, 1998), and one was in construction (Baiche et al., 2006). 6.2. Causes of violations In the reviewed articles, 57 different variables were tested to determine their association with violations. Table 3 shows the groupings of all 57 variables into categories and Table 4 summarizes results. 7. Discussion

5.2.3. Cited reference search Ten seminal articles were identified that focused on violations either theoretically or empirically (Battman and Klumb, 1993; Besnard and Greathead, 2003; Hobbs and Williamson, 2002; Lawton, 1998; Mason, 1997; Parker et al., 1992, 1995; Reason et al., 1990, 1998; Zeitlin, 1994). Using the cited reference search feature in the ISI Web of Knowledge database, a search was conducted for articles that referenced any of these key articles. No new articles were identified based on the cited reference search.

The main goal of this review was to answer the question, “What causes safety violations in industry?” Thirteen papers addressed this and 57 variables were examined as causes of safety violations. All of the reviewed studies implicitly or explicitly defined violations as an unwanted, though in some cases necessary, behavior. None of the reviewed studies examined whether violations can improve system performance and safety. With that caveat in mind, the discussion proceeds with a discussion of the predictors of violations, and then provides directions for future research, including a new macroergonomic conceptual framework of violations.

5.3. Analysis

7.1. Variables that predict violations

Articles meeting the inclusion criteria were reviewed and summarized. Factors that were tested as causes of violations were analyzed based on whether they had a positive, negative, or nonsignificant association with violations. The variables were further grouped according to common themes. For the purposes of this review, the variables examined as potential causes of violations were grouped under six categories: individual characteristics, information/education/training, design to support worker needs, safety climate, competing goals, and problems with rules. “Individual characteristics” included workers’ trait characteristics, variables related to workers’ experience or actions, and workers’ perceptions of their work. “Information/education/training” included the amount of information or training a worker possessed and workplace communication. “Design to support worker needs” included physical work system design characteristics that influenced the likelihood of violations. “Safety climate” included worker perceptions of safety climate, worker perceptions of management,

Variables that were examined as causes of safety violations were grouped into six categories: individual characteristics, information/education/training, design to support worker needs, safety climate, competing goals, and problems with rules. 7.1.1. Individual characteristics Among all of the individual characteristics tested (see Table 3), attitude towards compliance, habit to comply, perceived behavioral intention to comply, and previous accident involvement had significant associations with violations. Age and experience had conflicting associations with violations. Perceptions that workers were lazy or flouted the rules were also stated to predict violations, but these were reasons workers gave for why other workers violate and were not tested statistically. The relationships found between attitude towards compliance, habit to comply, perceived behavioral intention to comply and violations were tested in the same study (Beatty and Beatty, 2004)

S.J. Alper, B.-T. Karsh / Accident Analysis and Prevention 41 (2009) 739–754

743

Table 2 Summary of reviewed articles. Author, year; domain; subjects

Method

Study variables: results

Baiche et al., 2006; Construction; Housing development residents and building control inspectors

Observations and building inspections.

From building inspections, non-compliance deemed the result of:

Survey questionnaires. Semi-structured interviews.

• Shortcomings in workmanship. • Not of “flouting of the rules to take short cuts (p. 282). • “Omissions caused by failures in site organization” (p. 282). • Not checking the procedures carefully enough.

Notes: • Violations of construction protocols are somewhat unique in that they can be identified retrospectively based on the building structure. • Predictors of violations in this paper are attributions made by individuals who did not commit the violations.

From interviews with inspectors:

• Inspectors believed that tradesmens’ training and knowledge were associated with compliance/non-compliance with building regulations. In results section (not clear from which method these results were drawn), causes of non-compliance were: • Poor workmanship (lack training and/or skills, take shortcuts, do not know standards). • Ignorance of details of regulations (due to lack of training). • Use of incorrect or non-certified materials. • Poor management. • Conflict/confusion between trades. • Pressure to complete work. • Changes to standard approved designs. • Unfamiliarity with design. • Complicated, labour intensive details requiring work from several trades. • Lack of detailed calculation of ultimate site levels around dwellings. Beatty and Beatty, 2004; Healthcare; Anesthetists

Voluntary scenario-based survey questionnaires.

Intention to comply was associated with: • Attitude towards behavior (+, except for silencing of alarms). • Subjective norm (+). • Perceived behavioral intention (+). • Habit (+). For all three scenarios, self-report of violating was associated with: • Attitude towards compliance behavior (−). • Subjective norm to comply (−). • Perceived behavioral intention for compliance (−). • Habit of compliance (−).

Beilock, 1995; Commercial truck driving; Solo drivers

Survey questionnaire about drivers’ destinations and deadlines.

Schedule considered violation-inducing if not possible to reach destination prior to deadline without breaking hours-of-service rules.

• “Between 17% [assuming a driving speed of 65 miles/h] and 30% [assuming a driving speed of 55 miles/h] of drivers had violation-suspect schedules.” • “Between 14% [assuming a driving speed of 65 miles/h] and 26% [assuming a driving speed of 55 miles/h] [of drivers] had violation-inducing” schedules. • Higher time pressure for drivers with refrigerated loads (50–80% more likely to have violation-suspect or violation-inducing schedules assuming a driving speed of 65 and 55 miles/h, respectively). • Age (N.S.). • Driver training (N.S.). • Experience (N.S.). • Participation in ongoing safety program (N.S.). • Pay type: salary versus productivity-based (N.S.).

744

S.J. Alper, B.-T. Karsh / Accident Analysis and Prevention 41 (2009) 739–754

Table 2 (Continued ) Author, year; domain; subjects

Method

Study variables: results

Espin et al., 2006; Healthcare; Subjects were surgeons, nurses, and anesthesiologists

Scenario-based semi-structured interviews.

Work pressures, time pressures, and conflicting demands were identified by the authors as reasons for violating, based on interviewee responses. For example, interviewees felt that there were more important things to worry about than complying with the rules for counting sponges.

Hobbs and Williamson, 2002; Aviation maintenance; Subjects were aviation mechanics

Created and used a “maintenance behavior questionnaire” (MBQ). Factor analysis led to a four factor solution: 1. Routine violations: “actions that have become a normal way of working” (p. 871).

Exceptional violations were associated with:

2. Skill-based errors. 3. Mistakes. 4. Exceptional violations: “rare actions that occur in unusual situations” (p. 873).

• • •

Years in industry (+). Supervisory position (+). Working at a major airline (N.S.).

• •

Working at regional airline (+). Working at charter operator (+).

• •

Working for general aviation (+). Quality incidents (+).

•

Injury at work (N.S.).

Regression analyses to determine demographic factors that led to factors of the MBQ. Note that quality incidents were not a predictor of violations. They were a dependent variable.

• Age (−).

Routine violations were associated with: • Age (−). • Years in industry (+). • Supervisory position (N.S.). • Working at a major airline (+). • Working at regional airline (+). • Working at charter operator (+). • Working for General Aviation (+). • Quality incidents (+). Injury at work (N.S.) Laurence, 2005; Mining; Subjects were mine workers.

Survey questionnaire.

The most common reported reasons for violating were: • to lower the risk ∼35%, • problems with rules and regulations ∼33%, and • to save time ∼18% • Manager commitment lacking ∼7%. • It would save energy ∼6%.

Lawton, 1998; Railroad workers in the United Kingdom; Subjects were railroad workers

Pilot study

Pilot study

Survey questionnaire.

Association between perceived risk and frequency of violations was (−).

Preliminary interviews. Main study

Cluster analysis.

Main study Following are the motivators for violating that were developed in preliminary interviews. If the motivator was one of the top five motivators for a cluster, the cluster number is included after the motivator. • Time pressure (1, 2, 3, 4, 5). • High work load (1, 2, 4, 5). • Quicker way of working (1, 2, 3, 4, 5). • Inexperience (1, 2, 3, 5). • A skilled shunter can work safely this way (4, 5). • The shunter is lazy (1, 2). • Rule is outdated (1). • Rule impossible to work to (3). • Management turn a blind eye (4). • Design of the sidings makes the violation necessary (3). • No one understand the rule. • More exciting way of working. • Physical exhaustion. • It is a macho way to work.

S.J. Alper, B.-T. Karsh / Accident Analysis and Prevention 41 (2009) 739–754

745

Table 2 (Continued ) Author, year; domain; subjects

Method

Study variables: results

Li and Baker, 1995; Aviation; Subjects were pilots

Data collected from Federal Aviation Administration’s Medical Accident System.

From logistic regression, violations were associated with: • Having been involved in an accident before (+). • Age (N.S.).

Chi-square tests used to compare the control group with the crash group.

McKeon et al., 2006; Healthcare; Subjects were nurses

“Unconditional logistic regression models were fitted to estimate the adjusted relative risks” (p. 1132). Survey questionnaires. Provided correlations between the different variables.

Structural equation modeling was used to develop a path model of the r variables.

• •

From the path model: • Workload led to expectation by doctor (+) and violations (+). • Reference material led to expectation by doctor (−). • Level of Knowledge led to violations (−). •

Patterson et al., 2006; Healthcare; Subjects were nurses

Total flight hours (N.S.). Medical class (N.S.).

Expectation by doctor led to violations (+).

Ethnographic observations.

Design features of the BCMA system that made violations more likely in long-term care units than in acute care units:

Interviews. “Collected electronically generated medication administration data for the observed nurses at the medium and large hospital.”

• Medication carts larger in long-term care units to accommodate higher average number of medications, making them more difficult to maneuver.

Put observation data into a standard activity protocol and analyzed the protocols to find work-around strategies.

• Batteries for medication carts (with scanners) did not last an entire medication pass for long-term care units. This made moving computers (to complete a scan) more difficult because it required moving the plug. • Wristbands were worn longer in long-term care units, making them less likely to scan correctly. • Nurses tended to know patients better in long-term care units, which may have lessened the perceived risk of mis-identification. A greater proportion of nurses pre-poured medications (a violation) in long-term care than in acute care. Statistically significant based on Fisher’s exact test. Potential reasons to pre-pour medications: • Makes it look like medications were given on time • Saves time by allowing the bundling of tasks

Rebok et al., 2005; Aviation; Subjects were pilots

Taxis and Barber, 2003; Healthcare; Subjects were nurses and one physician

Data collected from “the biannual medical certification data and surveillance systems managed by the NTSB as well as the FAA’s Aviation Medical Examiner System and Medical Accident System (MAS)” (p. 365).

Used relative risk to determine relations between violations and pilot characteristics. Used Cox proportional hazards model to model the relationship between cumulative pilot flight hours and violations. Conducted observations and informal interviews.

Notes:

• “98% of observed violations were fast administration of bolus doses (injections administered faster than the recommended speed of 3–5 min).”

• Experience was negatively associated with violations for lower levels of experience (pilots with <5000 h flight time were at higher risk than pilots with between 5000 and 9999 h flight time). However, “protective” effect of experience diminished as pilots had greater than 10,000 cumulative hours of flight time. • Relative risk of age was 1.10 (i.e. 10% increase in rate of violations for each additional year of age).

Causes of violations (only fast administration of bolus errors): • Nurses perceived low risk. • Poor supervision (ex. senior nurse laughed at a junior nurse for administering a dose over the correct amount of time). • In some cases, it was technically difficult to administer the dose over the proper amount of time.

746

S.J. Alper, B.-T. Karsh / Accident Analysis and Prevention 41 (2009) 739–754

Table 2 (Continued ) Author, year; domain; subjects

Method

Study variables: results

Wills et al., 2006; Commercial driving; Subjects were commercial drivers

Survey questionnaires.

From hierarchical regression analysis, violations were associated with: • Age (+). • Sex (N.S.). • Average hours driven per week (N.S.). • Communication (N.S.).

Used hierarchical regression analysis. Notes: • Authors do not provide definitions of the six safety climate factors • Higher scores on the safety climate factors indicate higher perceptions of safety.

•

Work pressures (N.S.).

Cannot determine direction of association for safety rules because safety rules is not defined.

•

Relationships (N.S.).

• • •

Driver training (N.S.). Management commitment (N.S.). Safety rules (significant).

using the theory of planned behavior as a basis. The significant relationship between subjective norm to comply and violations rounds out the model (mentioned here to convey the full theory of planned behavior; subjective norm to comply was grouped in the safety climate category). Given that the theory of planned behavior is a model for predicting intentional behavior, it makes sense to use it to study intentional violation behaviors. It has successfully been used to predict violations in aircraft maintenance (Fogarty and Shaw, 2003) and in recreational driving (Parker et al., 1992), and it shows promise as a theoretical basis for understanding violations. The reviewed literature also found a direct relationship between previous accident involvement and violations (Li and Baker, 1995). This is consistent with findings in recreational driving in which previous accident involvement is generally associated with violating (Blockey and Hartley, 1995; Kontogiannis et al., 2002; Mesken et al., 2002). However, a study of commercial drivers by Sullman et al. (2002) showed that the relationship between previous accident involvement and violations may depend on the type of violation. While a relationship was found between previous accident involvement and “violations,” no relationship was found between previous accident involvement and “aggressive violations.” If it is accepted that there are different types of violations (e.g. Lawton, 1998; Mason, 1997), then further study is needed to determine how previous accident involvement relates to different types of violations.

The conflicting findings for age found in the reviewed literature are consistent with conflicting findings in the universal precautions literature where it has been found that younger health care providers violate universal precautions protocols more (Chan et al., 2002; Gershon et al., 1999), that older health care providers violate more (Ben-David and Gaitini, 1997), and that no relationship between age and violations exists (Ji et al., 2005; Naing et al., 2001). On the other hand, findings in the recreational driving literature have found that being younger increases the likelihood of violating (e.g. Mesken et al., 2002; Reason et al., 1990). The mixed relationships found in the reviewed literature between experience and violation likelihood differed from findings from the recreational driving (Aberg and Rimmo, 1998; Kanellaidis et al., 1995; Schwebel et al., 2006) and universal precautions (Gershon et al., 1999; Naing et al., 2001) literatures on violations where no relationship between experience and violations has been found. In the reviewed literature, the findings were in opposite directions; Hobbs and Williamson (2002) found a positive association between experience and violations and Rebok et al. (2005) found a negative association. On a related note, Lawton (1998) found that railroad workers reported experience as a factor related to why workers violate. If an association does exist between experience and violations, it seems unlikely that the association is direct or linear. Rather, the association between experience and violations may interact with other variables such as risk perception or skill.

Fig. 2. Macroergonomic framework of safety violations. Bold variables are those indentified in the review.

Table 3 Groupings of variables. Organized as: “Variable. (Description of variable. Method of data collection.)”. Information/education/ training

Design to support worker needs

Safety climate

Competing goals

Problems with rules

Other

Experience. (Time with current firm; time in the industry; time in the specific job. Self-report; worker reported reason; review of records.)

Design makes necessary. (Design of a tool or environment makes compliance impossible. Self-report; observation.)

Poor management. (Lack of supervision, experienced workers giving junior workers advice to violate. Retrospective evaluation; interviews.)

Airline type. (Worker worked at regional, charter, or major airline. Self-report.)

Changes to standard approved designs. (Not defined further. Retrospective evaluation.)

No one understands rule. (Worker provided reason a worker may violate. Self-report.)

Lack of calculations of site levels. (Not defined further. Retrospective evaluation.)

Previous accident. (Worker had been involved in a previous accident. Review of records.)

Use incorrect materials. (Use of incorrect materials in construction. Retrospective evaluation.)

Management turns a blind eye. (Workers stated that a reason for violating is that management ignores the violation. Self-report.) Subjective norm to comply. (Worker’s perceptions of how relevant others would perceive compliance. Self-report.) Participation in safety programs. (Participation in ongoing safety program. Self-report.)

Injury at work. (Worker had been injured at work in previous 12 months. Self-report.) Worker is lazy. (Worker provided reason that other workers violate. Self-report.) Shortcoming in workmanship. (Work evaluated as not up to par. Retrospective evaluation.) Perceived behavioral intention to comply. (How much worker intended to comply. Self-report.) Flouting of the rules. (Intentionally not following rules. Retrospective evaluation.) Medical class. (Class of medical certificate held by pilot. Review of records.) Hours driven per week. (Number of hours driven each week. Self-report.)

Unfamiliarity with design. (Worker not familiar with design of building. Retrospective evaluation.) Failures in site organization. (Not further defined. Retrospective evaluation.)

Expectation by doctor. (How much nurse perceives doctor expects nurse to violate. Self-report.) Not checking procedures. (Not defined further. Retrospective evaluation.)

Relationships. (Not further defined. Self-report.)

Manager commitment. (Not defined further. Self-report.)

Time pressure. (Schedule tightness comparing required mileage and deadline; pressure to perform quickly; worker provided reason for violating. Self-report; interview. Pay type – salary versus productivity. (Whether driver is paid a salary or paid based on per load or per mileage. Self-report.) Work pressure. (Pressure to get work done; not further defined. Retrospective evaluation; interviews; self-report.) Perceived risk. (Amount of risk associated with specific violations; evaluation of risk of specific violations. Self-report; interview.) Workload. (Worker provided reason to violate; adverse conditions of work. Self-report.) Conflicting demands. (Having tasks the compete for worker’s time. Interview; observation.) Skilled worker can do safely. (Worker provided reason a worker may violate. Self-report.)

Rule impossible to work to. (Worker provided reason a worker may violate. Self-report.)

Attitude towards compliance. (Worker’s beliefs about outcomes likely to result from compliance. Self-report.) Total flight hours. (Number of hours a pilot had spent pilot a plane. Review of records; self-report.)

Worker level of knowledge. (Worker knowledge of rules and regulations; nurse level of knowledge of medications and ability to explain to patients. Retrospective evaluation; self-report.) Conflict/confusion between trades. (Different trade workers not communicating effectively. Retrospective evaluation.) Worker level of training. (Worker not familiar with details of regulations. Retrospective evaluation.)

Communication. (Not further defined. Self-report.)

Supervisory position. (Worker held a supervisory position. Self-report.)

Complicated design. (Not defined further. Retrospective evaluation.)

Physical exhaustion. (Worker provided reason a worker may violate. Self-report.) Quicker way of working. (Worker provided reason a worker may violate. Self-report.)

Difficult to comply. (Technically difficult to follow protocol. Interview.)

Safety rules. (Not defined further. Self-report.)

Problems with rules. (Not defined further. Self-report.)

747

To save time. (Not defined further; to complete tasks more quickly. Self-report; observation.) More exciting way to work. (Worker provided reason a worker may violate. Self-report.)

Rule is outdated. (Worker provided reason a worker may violate. Self-report.)

S.J. Alper, B.-T. Karsh / Accident Analysis and Prevention 41 (2009) 739–754

Individual characteristics

S.J. Alper, B.-T. Karsh / Accident Analysis and Prevention 41 (2009) 739–754

Macho way to work. (Worker provided reason a worker may violate. Self-report.) Type of cargo. (Whether driver’s cargo is refrigerated or non-refrigerated. Self-report.) Work schedule. (How a worker’s schedule is set for the worker. Self-report.) To lower risk. (Not defined further. Self-report.) To save energy. (Not defined further. Self-report.) Sex. (Sex of worker. Self-report.)

Habit to comply. (A worker’s personal norm to comply. Self-report.)

Age. (Age of worker. Self-report; review of records.)

Individual characteristics

Table 3 (Continued )

Information/education/ training

Design to support worker needs

Safety climate

Competing goals

Problems with rules

Other

748

The findings that worker laziness (Lawton, 1998) and shortcoming in workmanship (Baiche et al., 2006) predict violations were both based on workers’ reports of why other workers violate. Therefore, they are likely subject to attribution errors (Dejoy, 1994). In addition, a shortcoming in workmanship may be the outcome of a violation rather than a cause. Two factors for which no significant association with violations was found in the reviewed literature were exposure and sex. The lack of a signification relationship in the reviewed literature is consistent with the literature on compliance with universal precautions (Gershon et al., 1999), but differs from findings in the recreational driving literature where exposure is often found to have a positive relationship with violations (e.g. Özkan and Lajunen, 2005; Reason et al., 1990). The pattern for sex with regard to research on violations in industry, recreational driving, and universal precautions was the same as the pattern for exposure. In both the reviewed literature and in the universal precautions literature (e.g. Gershon et al., 1999), no association between sex and violations was found. However, evidence among recreational drivers shows that males are more likely to violate (e.g. Blockey and Hartley, 1995; Kanellaidis et al., 1995; Parker et al., 1995; Shinar et al., 2001). A number of the individual characteristics that have been studied for potential associations with violations in the recreational driving and universal precautions literatures were not examined in the reviewed articles. These included education (Gershon et al., 1995; Ji et al., 2005; Shinar et al., 2001), the extent to which an individual lets their mood influence their behavior (Reason et al., 1990; Schuman et al., 1967), drinking behavior (Dobson et al., 1999; Schuman et al., 1967), income (Shinar et al., 2001), and personality characteristics (Schwebel et al., 2006). Whether any of these are important predictors of industrial safety violations remains to be seen. While much work has been done trying to identify associations between individual characteristics and violations, more interesting significant effects were found for the other categories of predictors. This is an important finding in itself because the other categories represent variables that are under an organization’s control and therefore are more amenable to intervention. 7.1.2. Information/education/training Four of the variables (unfamiliarity with design, use of incorrect materials, conflict/confusion between trades, and failures in site organization) were positively associated with violations, two variables (worker level of knowledge, worker level of training) were negatively associated with violations, and the other two variables had non-significant relationships (relationships, communication). All of the significant variables in this category appear to be associated with the worker lacking information. There is some support for this finding in the universal precautions literature (Cabana et al., 2000; Hammond et al., 1990; Naing et al., 2001), though null findings exist as well (Chan et al., 2002; Ji et al., 2005). Since a lack of information may lead to violations, it would seem that training could improve compliance by ensuring that all workers are educated and have sufficient information. However, there is a possible flaw with that idea. As previously mentioned, most studies in the current review assumed that violations of safety rules were undesirable. If, in fact, there were occasions in the 13 studies where the violations of safety rules improved safety, then it is likely that those violators already were highly trained and educated and it was precisely their training and education that led them to violate. Furthermore, a longitudinal study of hand washing found that after training sessions, compliance spiked and then declined (Dubbert et al., 1990), casting doubt on the use of training as a long-term method for addressing safety violations.

S.J. Alper, B.-T. Karsh / Accident Analysis and Prevention 41 (2009) 739–754

749

Table 4 Summary of results; variables and their association with violations. Positive association. As factor increases, violations increase

Individual characteristics: Shortcoming in workmanship (Baiche et al., 2006); worker is lazy (Lawton, 1998); previous accident (Li and Baker, 1995) Information/education/ training: Unfamiliarity with design (Baiche et al., 2006); use incorrect materials (Baiche et al., 2006); conflict/confusion between trades (Baiche et al., 2006); failures in site organization (Baiche et al., 2006) Design to support worker needs: Design makes necessary (Lawton, 1998; Patterson et al., 2006); complicated designs (Baiche et al., 2006); changes to standard approved designs (Baiche et al., 2006) Safety climate: Not checking procedures (Baiche et al., 2006); expectation by doctor (McKeon et al., 2006); management turn a blind eye (Lawton, 1998); poor management (Baiche et al., 2006; Taxis and Barber, 2003) Competing goals: Macho way to work (Lawton, 1998); more exciting way to work (Lawton, 1998); physical exhaustion (Lawton, 1998); quicker way of working (Lawton, 1998); skilled worker can do it safely (Lawton, 1998); time pressure (Beilock, 1995; Espin et al., 2006; Lawton, 1998); to lower risk (Laurence, 2005); to save energy (Laurence, 2005); to save time (Laurence, 2005; Patterson et al., 2006); conflicting demands (Espin et al., 2006; Patterson et al., 2006) Problems with rules: Difficult to comply (Taxis and Barber, 2003); no one understands rule (Lawton, 1998); problems with rules (Lawton, 1998); rule impossible to work to (Lawton, 1998); rule is outdated (Lawton, 1998) Other: Lack calculations of site levels (Baiche et al., 2006)

Negative association. As factor increases, violations decrease

Individual characteristics: Attitude towards compliance (Beatty and Beatty, 2004); habit to comply (Beatty and Beatty, 2004); perceived behavioral intention to comply (Beatty and Beatty, 2004) Information/education/training: Worker level of knowledge (Baiche et al., 2006; McKeon et al., 2006); worker level of training (Baiche et al., 2006) Design to support worker needs: Safety climate: Subjective norm to comply (Beatty and Beatty, 2004) Competing goals: Perceived risk (Lawton, 1998; Taxis and Barber, 2003) Problems with rules: Other:

Non-significant. No evidence found of an association between factor and violations

Individual characteristics: Flouting of the rules (Baiche et al., 2006); sex (Wills et al., 2006); hours driven per week (Wills et al., 2006); injury at work (Hobbs and Williamson, 2002); medical class (Li and Baker, 1995) Information/education/training: Communication (Wills et al., 2006); relationships (Wills et al., 2006) Design to support worker needs: Safety climate: Participation in safety program (Beilock, 1995) Competing goals: Pay type (salary versus productivity) (Beilock, 1995) Problems with rules: Other:

Conflicting results. Evidence about association between factor and violations not consistent

Individual characteristics: Age (Beilock, 1995; Hobbs and Williamson, 2002; Li and Baker, 1995; Rebok et al., 2005; Wills et al., 2006); experience (Beilock, 1995; Hobbs and Williamson, 2002; Lawton, 1998; Rebok et al., 2005); total flight hours (Li and Baker, 1995; Rebok et al., 2005) Information/education/training: Design to support worker needs: Safety climate: Manager commitment (Laurence, 2005; Wills et al., 2006); supervisory position (Hobbs and Williamson, 2002) Competing goals: Workload (Lawton, 1998; McKeon et al., 2006); work pressure (Baiche et al., 2006; Espin et al., 2006; Wills et al., 2006) Problems with rules: Other:

Finding has no direction. Factor was not defined such that it can increase or decrease

Individual characteristics: Information/education/training: Design to support worker needs: Safety climate: Competing goals: Type of cargo (Beilock, 1995); work schedule (Beilock, 1995) Problems with rules: Safety rules (Wills et al., 2006) Other: Airline type (Hobbs and Williamson, 2002)

7.1.3. Design to support worker needs Three variables in this category (design makes necessary, complicated design, and changes to standard approved designs) were positively associated with violations. As the design made violations more necessary, grew more complicated, or differed more from standard approved designs, violations increased. This category of variables demonstrates that violations may result from

systems that are not supporting workers doing their work. In the violations literature, this idea has been expressed using a concept referred to as a situational violation (e.g. Mason, 1997) in which a worker violates because it is necessary to complete job tasks. Situational violations may therefore define the overlap between the related, but distinct, concepts of “violations” and “work-arounds.” Where violations are defined with respect to rules, work-arounds

750

S.J. Alper, B.-T. Karsh / Accident Analysis and Prevention 41 (2009) 739–754

are defined with respect to barriers in workflow (Halbesleben et al., 2008). When a rule is broken to deal with a barrier to workflow, the action is both a violation and a work-around. If a worker is satisfied with, or supported by good product and process designs, then the situational barrier to workflow may not exist, eliminating the need to violate or work-around. The universal precautions literature is replete with evidence that a worker not being supported in completing his or her work is more likely to violate. In health care, a lack of support can be as basic as not having access to necessary safety equipment or tools (Adegboye et al., 1997; Cutter and Jordan, 2004; Naing et al., 2001; Nelsing et al., 1997). Having tools that do not support a worker can also lead to violations. Healthcare workers have reported that they do not wear gloves due to a reduction in tactile sensation (Adegboye et al., 1997; Naing et al., 2001), because the gloves interfere with their dexterity (Henry et al., 1994; Naing et al., 2001; Taveira et al., 2002), or because the gloves do not fit (Naing et al., 2001; Nelsing et al., 1997). Generally, when workers do not have access to tools that support them in completing their work, violations seem to be more likely. The lack of support does not stem solely from problems with tools and technologies, but also from other components of the work system: the organization, the task, and the environment (Carayon et al., 2006; Smith and Carayon-Sainfort, 1989). Though none of the studies reviewed explored how these factors may contribute to violations, research in related literatures demonstrates they are important contributors. Staffing level, an organizational factor that also influences workers’ tasks, was found to be related to violations of handwashing protocols such that for each additional patient a nurse had to care for, there was a significant increase in violations (Brown et al., 2003) and each patient fewer a nurse had to care for was associated with an increase in compliance (Harbarth et al., 2002). Similarly, Hugonnet et al. (2002) found that when nurses had greater than 60 opportunities for handwashing in an hour, they were more likely to violate than when they had 0–20 opportunities for handwashing in an hour. Finally, Patarakul et al. (2005) found that the design of the environment could also have an influence on violation behaviors; healthcare professionals indicated that inconveniently located handwashing sinks and alcohol-based hand rubs were a reason for violating handwashing protocols. It is imperative that research on industrial safety violations explores how human–tool, human–job, human–environment, and human–organization interactions affect the likelihood of violations. Poor design or fit among all of these interfaces is well-known to contribute to an array of performance problems and it is highly likely that poorly designed interfaces also contribute to violations. Some evidence for this certainly exists (Carayon et al., 2007; Halbesleben et al., 2008; Koppel et al., 2008; Patterson et al., 2006), though there is a clear need for quantitative modeling to clarify the nature of the evidence. Intuitively, better designed systems that make work easier for workers should lead to fewer violations. However, there are likely complex relationships among the costs and benefits of different system design variables (e.g. it is easy to comply using the tool, but the supervisor wants his/her staff to rush) that require further exploration. 7.1.4. Safety climate Perceptions of safety climate are shaped by “policy and procedural actions of top management and from supervisory actions exhibited by shop-floor or frontline supervisors” (Zohar and Luria, 2005, p. 516), and these perceptions are used to guide behavior (Zohar, 1980, 2002). As such, the link between safety climate and violations makes sense since safety climate is linked to safety behaviors (Pousette et al., 2008). Of the variables related to safety climate, four were positively associated with violations (not check-

ing procedures, management turns a blind eye, poor management, expectation by doctor), one was negatively associated with violations (subjective norm to comply), one was not significantly associated with violations (participation in safety program), and two had conflicting results (manager commitment, supervisory position). The variables with positive, significant associations with violations demonstrate that poor actions of management can result in workers believing that management condones violation behaviors, such as management turns a blind eye or expectation by a doctor for a nurse to violate. “It perhaps comes as no surprise to discover that violations are highly susceptible to management influences – since many of the underlying causes of violations are created, often inadvertently, by management itself” (Mason, 1997, p. 291). Safety climate has also been found to predict violations in the universal precautions compliance literature (Gershon et al., 1999, 1995) and in the hand hygiene compliance literature (Ji et al., 2005). 7.1.5. Competing goals The vast majority of variables in this category were positively associated with violations. These variables included time pressure, conflicting demands, to lower risk (a worker believes that violating will lower risk in a situation, which competes with a desire to comply), skilled worker can do [the task] safely, to save energy, physical exhaustion, quicker way of working, to save time, more exciting way to work, and macho way to work. Perceived risk was the only variable negatively associated with violations, such that when a worker perceived an increasing amount of risk, violations were less likely. Two variables had conflicting results (work pressure, workload), one was not significant (pay type – salary versus productivity), and two did not have a directional finding (type of cargo in commercial driving, work schedule). Consistent with the reviewed literature, studies of violations among recreational drivers (Aberg and Rimmo, 1998; Porter and Berry, 2001), compliance with universal precautions (Cutter and Jordan, 2004; Henry et al., 1994) and handwashing (Patarakul et al., 2005) also show that competing goals predict violations. Conflicting goals can lead to violations when an individual has a goal of compliance with the rules, and a different goal that conflicts with compliance. For example, Alper et al. (2008) described a situation in which a nurse had to choose between administering a medication late and administering a medication that was intended for another patient (in this case, the nurse perceived both actions as violations). The nurse’s competing goals were (1) to give the medication to the patient on time and (2) to avoid borrowing a medication from a different patient. The nurse was only able to meet one of those goals. The situations in which goal competition leads to compliance with, or violation of, safety rules are still unclear. Woods and Patterson (2001) escalation principle suggests that more difficult situations reduce slack in a system; the reduction of slack increases demands and may lead to safety violations. On the other hand, difficult or risky situations may increase compliance with the safety goal, at the expense of a competing goal such as productivity, if the worker perceives that the risk of the situation outweighs the possible benefits of increased productivity or speed (Battman and Klumb, 1993; Hope and Pate, 1988). It may alternatively be that when a worker has competing goals, the worker chooses the goal that he/she is more committed to (Locke and Latham, 2002). Two important determinants of goal commitment are the importance of the goal and an individual’s selfefficacy. Thus, explaining the importance of compliance and helping workers feel that they have the ability to comply could increase commitment to compliance. Additionally, performance on assigned goals is moderated by the goals an individual sets for him/herself (Locke and Latham, 2002). This is in keeping with the finding in this

S.J. Alper, B.-T. Karsh / Accident Analysis and Prevention 41 (2009) 739–754

review that an individual’s personal goals, such as a desire to complete a task using less effort (e.g. to save energy) or to complete a task in a more fun way (e.g. more exciting way to work), can impact an individual’s goal of compliance with safety rules. 7.1.6. Problems with rules Each of the variables in this category (difficult to comply, problems with rules, no one understands rule, rule impossible to work to, rule is outdated) was positively associated with violations, except one that did not have a stated direction (safety rules). These findings are intriguing and difficult to interpret. While violations are generally viewed negatively (Lawton and Parker, 1999; Parker and Lawton, 2000), if a rule is bad, then a violation may be the correct action (Reason et al., 1998). Since rules do not always prescribe behaviors that will lead to optimal results (Dekker, 2003; Hope and Pate, 1988; Reason et al., 1998), in some cases violating a safety protocol could improve results. To violate safely, however, may require that a worker, “(a) has an accurate knowledge about the task at hand (b) has enough information available (c) and processes this information accurately” (Besnard and Greathead, 2003, p. 277). It was not clear from the studies reviewed that examined problems with rules whether the problems with rules reported by workers were valid or not. Similar evidence exists in the universal precaution compliance literature, where healthcare professionals have reported violating because they did not think protocols were necessary (Hammond et al., 1990), because they did not think that taking measures prescribed by rules was effective (Nelsing et al., 1997), or because they disagreed with protocols (Cabana et al., 2000). Two questions regarding rules are thus raised: “why do rules fail?”, and “how should organizations address problems with rules?” Rules can fail for a wide variety of reasons. Hope and Pate describe this stating, “a directive that once may have been perfectly logical or rational may become ineffective as tasks, people, structures, and technologies change” (1988, p. 741). In other words, rules fail because the systems, both internal and external to an organization, are in a constant state of change. As such, static rules cease to keep the system within its safety boundaries – boundaries which defend the line between safety and disaster (Woods, 2006) – which means that when rules fail, workers may be forced to violate to maintain system safety (Dougherty, 1995). So the question becomes what should an organization do to maintain safety, not what should an organization do to eliminate violations. To maintain safety an organization should consistently monitor its safety boundaries to determine if its rules and processes continue to facilitate safety (Almaberti, 2001; Polet et al., 2003; Rasmussen, 1997; Woods, 2006). Organizations must understand the gap between prescribed procedures and practice (Dekker, 2003). The presence of violations may serve as a litmus test that an organization’s processes have ceased to effectively protect the safety boundary – either because the rule makes the situation less safe (e.g. Dekker, 2003) or because the rule makes completing the work impossible (Lawton, 1998). To address problems with rules, Besnard and Greathead offer sage advice, “design workable instead of exhaustive procedures” (2003, p. 280). 7.2. Directions for future research Intentional non-malevolent violations, by their very nature, are difficult to study for a variety of reasons. First, they are socially undesirable. Thus, violations may be underreported or hidden. Second, the determination of whether or not a behavior is a violation depends on the rules that are considered relevant. In this review, different researchers used different rules, which has led to different sets of behaviors constituting violations. The use of different rules when studying violations in organizations is necessary because

751

organizational rules vary across organizations. Social interactions complicate the selection of which rules apply even further: should one consider a violation to be breaking the written protocol, or breaking the social norm that has developed for how work is completed? Third, many times a violation is completed and no adverse event occurs. This makes violations difficult to detect. Nonetheless, further research on violations in organizations is necessary as violations present a path to accidents that has been largely unaddressed. 7.2.1. Methodological considerations The methodology chosen for a study must be appropriate for determining whether, and what, rules were violated. When using an observational methodology, it must be possible to visually assess whether or not a rule was followed, and this is not always a simple task. Taking red-light running behavior as an example, very specific criteria were necessary to determine what constituted a violation (Porter and England, 2000; Retting and Williams, 1996). However, observational studies preclude assessment of intentionality because the violator’s intentions are unknown. While counting violations may be straightforward in an observational study, determining the causes of those violations may not be possible. To determine a violator’s intentions and to determine the causes of violations, it may be necessary to use a method that allows subjects to specify causes and intentions. Interviews, focus groups, or surveys allow subjects to provide researchers with their thoughts about violations. As mentioned above, subjects may have a difficult time participating honestly based on the social undesirability of violations. Also, clearly defined rules are just as important when asking subjects about violations as when observing violations. If rules are not clearly defined, when subjects discuss violations they may be discussing violations of rules that researchers do not know about or are not interested in. For example, if a researcher wants to know about violations of a written rule and asks subjects about violations in general, the subjects may respond with respect to social norms. If the social norm is to violate the written rule, the subject will report no violations when, based on the researcher’s interest, violations may be common. Many of the studies reviewed utilized self-report or other-report methodologies. These studies offer an important contribution, not least because they utilize workers’ expertise and knowledge as a means for identifying the causes of violations. Workers have proven adept at identifying hazards in their work settings (Smith, 2002), and it is possible that they are also able to identify causes of violations. This identification could be realized using processes such as participatory ergonomics (Haims and Carayon, 1998; Imada, 1991; Noro and Imada, 1991). Further research could be directed toward understanding the extent to which worker-identified reasons for violations can be validated using other methodologies. More generally, because both qualitative methodologies and quantitative methodologies have strengths and weaknesses, use of multiple methods is recommended to allow convergent support from multiple sources. A few studies have demonstrated such a use of multiple methodologies (Conner et al., 2007; Koppel et al., 2008). 7.2.2. Theoretical frameworks to study violations One reason that so many potential causes of violations have been explored may be a lack of theoretical guidance. The few studies that were based on theory used different theories (Beatty and Beatty, 2004; Espin et al., 2006; Taxis and Barber, 2003), perhaps because there is no unifying theory of violations at this time or perhaps because the causes of violations are multi-factorial. Theoretical frameworks used in the driving literature suggested individual characteristics that predispose individuals to violate (e.g. Castellá and Pérez, 2004). Theoretical frameworks used in the organiza-

752

S.J. Alper, B.-T. Karsh / Accident Analysis and Prevention 41 (2009) 739–754

tional literature seemed to focus on variables that influenced the decision to violate (e.g. Beatty and Beatty, 2004). As discussed above, frameworks that focus on individuals’ characteristics may be useful in determining which individuals are likely to violate. However, when studying intentional, non-malevolent violations, it is assumed that the individual chose to violate. Thus, the use of decision-making theories to understand why individuals choose to violate may shed light on what variables are salient when choosing whether or not to violate (e.g. Battman and Klumb, 1993; Hope and Pate, 1988). These theories recognize that deciding to violate has both benefits and costs; benefits could include reduced time pressure or increased productivity while costs could include acceptance of a greater level of risk. When an individual perceives that the benefits outweigh the costs, the decision to violate is ‘justified.’ Using behavior theories also makes sense in the study of violations because violations are, in fact, behaviors (e.g. Beatty and Beatty, 2004; Parker et al., 1992). The theory of planned behavior is one such theory (Ajzen, 1991) which postulates that intention and perceived behavioral control directly influence behavior. It also states that attitude towards a behavior, subjective norm, and perceived behavioral control influence behavior through intention. Given the findings in this review, it is clear that the causes of violations are manifold. It is also clear that in actual practice these causes likely work in concert to create situations in which an individual chooses to violate. Because the causes of violations can come from any level of the work system, we propose a macroergonomic (Hendrick and Kleiner, 2001) framework for studying and addressing violations. Using this approach helps to reinforce that to address violations it may be necessary to make changes as specific as the way an individual interfaces with a technology, or as broad as the safety climate of the organization. A macroergonomic systems framework of violations is presented in Fig. 2. The framework identifies a number of variables that were empirically tested in the reviewed studies (in bold) and variables that were not tested that theoretically could be associated with violations. It suggests recognition of the multilevel nature of work systems (Carayon et al., 2006; Karsh et al., 2006) to define situations encountered by workers as part of a framework explaining violations. At the most basic level, the framework shows that an individual encounters a situation, which is defined by the dynamics of the work system (left-side of model), then makes a decision about whether or not to violate in that situation, and finally acts on that decision. Considering only this, the framework is similar to other models that have been proposed that also address the compliance/violation decision, the applicability of the rule, and the individual’s preference (Hope and Pate, 1988; Reason et al., 1998). However, the results of this review demonstrate that the decision of whether or not to comply with/violate a rule is more complex. The unique contribution of Fig. 2, and the way it differs from models that have been proposed in the past, is that it accounts for the various influences of the work system on both the situation the individual encounters and on the decision that the individual makes. Once the compliance/violation decision has been made and the individual has taken action, the results of that action feed back into the work system, which includes the individual, to influence future situations and to influence future decisions-making. To obtain desired performance with regard to violations, inputs to the situation and to the compliance/violation decision, represented by the system pyramid (Fig. 2), should be designed to support employees making a good decision. Importantly, a good decision may not always be compliance. This is the case when rules do not fit the situation (Reason et al., 1998), and may suggest the need for an iterative, macroergonomic process for rule creation. This could be

accomplished through, for example, participatory ergonomic processes (Haims and Carayon, 1998; Imada, 1991; Noro and Imada, 1991). 7.2.3. Addressing intentional non-malevolent violations The evidence reviewed would suggest that better design to support worker performance, elimination of production/safety goal conflicts, supportive safety climates, and better designed rules could all help reduce the need to violate safety rules. But, it remains unclear if these are the best strategies. Based on the reviewed evidence it is clear that too little is known at this time about causes of violations to provide specific instruction on how to address violations. This is true for at least two reasons. First, there is inconclusive evidence in the literature regarding which variables cause violations. This is because of a lack of quantitative modeling, lack of replication, and uncertainty about whether entire classes of variables have gone unmeasured. The latter problem can be addressed by the use of strong theory. Second, not enough is known about violations to assess which variables lead to good outcomes (potentially averting poor outcomes), and which violations lead to poor outcomes. However, there exist well-specified methods for analyzing and designing work systems to support safety, such as hazard management (Smith et al., 2001, 2003; Smith, 2002), Macroergonomics (Hendrick and Kleiner, 2001), cognitive work analysis (Vicente, 1999), participatory ergonomics (Haims and Carayon, 1998; Imada, 1991; Noro and Imada, 1991) and others which can be used to minimize the likelihood of producing systems and situations that require unsafe violations. Each method accomplishes this by, in its own way, attempting to design systems to support workers in a way that achieves the system’s multiple goals, such as safety, efficiency, and quality. Each method is guided implicitly or explicitly by a recognition that work systems are multilevel structures, as shown in Fig. 2, and the vertically nested levels must fit together. 8. Conclusion The purpose of this paper was to provide a review of the empirical literature studying causes of violations of safety rules in organizational settings. Thirteen articles studying a variety of work settings were identified that met the review criteria, and 57 potential predictors of violations were examined. The predictors of violations appeared to be multi-factorial, and were generally related to the individual worker, the organization, the worker’s task, or the organization’s rules. Some of the predictors identified suggested measures that could be taken to address violations. However, much more research is needed to determine which variables consistently predict unsafe violations and to determine the mechanism of action. In addition, the conversation about violations needs to continue to evolve from one of automatically assigning blame to one that views violations as symptoms of system design problems. It has been stated before that most workers do not violate with an intention to cause harm (Mason, 1997; Reason et al., 1995); however, it is easy to blame workers for violating because rules dictate that certain actions are ‘wrong’ and certain actions are ‘right.’ Reason (2008) calls blaming one of three pathologies that can make a system more vulnerable to adverse events and further states that it “is the most tenacious and the most harmful” (p. 74) of the three. We propose practicing systems thinking, or what others have called ‘attributional charity’ (Griffin and Ross, 1991); let us proceed with the understanding that violating can, in some cases, be the correct action (Reason et al., 1998), and instead of blaming workers for violating, strive to understand why they violate so that we can design their work environments to eliminate, or reduce the need

S.J. Alper, B.-T. Karsh / Accident Analysis and Prevention 41 (2009) 739–754

for violations or allow violations to happen safely, when they are necessary. Acknowledgements This study was funded in part by a grant from the Agency for Healthcare Research and Quality (R01 HS013610, Karsh PI). The authors would like to thank Dr. Pascale Carayon and Dr. Douglas A. Wiegmann for helpful discussions on the topic of violations, and Dr. Todd Loushine, Emily Davenport, and Scot Barnett for their helpful suggestions. The authors would also like to thank Amy Kindschi and Emily Wixson for their help on searches and databases. Finally, the authors thank the anonymous reviewers for their helpful suggestions. References Aberg, L., Rimmo, P.A., 1998. Dimensions of aberrant driver behaviour. Ergonomics 41 (1), 39–56. Adegboye, A.A., Roy, P.K., Emeka, C., 1997. Glove utilization and reasons for poor compliance by health care workers in a Nigerian teaching hospital. Tropical Doctor 27, 93–97. Ajzen, I., 1991. The theory of planned behavior. Organizational Behavior and Human Decision Processes 50, 179–211. Almaberti, R., 2001. The paradoxes of almost totally safe transportation systems. Safety Science 37 (2–3), 109–126. Almaberti, R., Vincent, C., Auroy, Y., Maurice, G.d.S., 2006. Violations and migrations in health care: a framework for understanding and management. Quality and Safety in Health Care 15, 66–71. Alper, S.J., Karsh, B., Holden, R.J., Scanlon, M.C., Patel, N., Kaushal, R., 2006. Protocol violations during medication administration in pediatrics. In: Paper Presented at the Human Factors and Ergonomics Society 50th Annual Meeting, San Francisco. Alper, S.J., Scanlon, M.C., Murkowski, K., Patel, N., Kaushal, R., Karsh, B., 2008. Routine and situational violations during medication administration. In: Paper Presented at the 9th International Symposium on Human Factors in Organizational Design and Management, Guarujá, São Paulo, Brazil. Baiche, B., Walliman, N., Ogden, R., 2006. Compliance with building regulations in England and Wales. Structural Survey 24 (4), 279–299. Battman, W., Klumb, P., 1993. Behavioral economics and compliance with safety regulations. Safety Science 16, 35–46. Beatty, P.C.W., Beatty, S.F., 2004. Anaesthetists’ intentions to violate safety guidelines. Anaesthesia 59, 528–540. Beekmann, S.E., Henderson, D.K., 2005. Protection of healthcare workers from bloodborne pathogens. Current Opinion in Infectious Diseases 18, 331–336. Beilock, R., 1995. Schedule-induced hours-of-service and speed limit violations among tractor-trailer drivers. Accident Analysis and Prevention 27 (1), 33–42. Ben-David, B., Gaitini, L., 1997. Compliance with gloving in anesthesia: an observational study of gloving practice at induction of general anesthesia. Journal of Clinical Anesthesia 9, 527–531. Besnard, D., Greathead, D., 2003. A cognitive approach to safe violations. Cognition, Technology, and Work 5, 272–282. Bisset, L., 2002. Can alcohol hand rubs increase compliance with hand hygiene? British Journal of Nursing 11 (16), 1072–1077. Blockey, P.N., Hartley, L.R., 1995. Aberrant driving behaviour: errors and violations. Ergonomics 38 (9), 1759–1771. Brown, S.M., Lubimova, A.V., Khrustalyeva, N.M., Shulaeva, S.V., Tekhova, I., Zueva, L.P., et al., 2003. Use of an alcohol-based hand rub and quality improvement interventions to improve hand hygiene in a Russian neonatal intensive care unit. Infection Control and Hospital Epidemiology 24 (3), 172–180. Cabana, M.D., Ebel, B.E., Cooper-Patrick, L., Powe, N.R., Rubin, H.R., Rand, C.S., 2000. Barriers pediatricians face when using asthma practice guidelines. Archives of Pediatrics & Adolescent Medicine 154, 685–693. Carayon, P., Hundt, A.S., Karsh, B., Gurses, A.P., Alvarado, C.J., Smith, M., et al., 2006. Work system design for patient safety: the SEIPS model. Quality & Safety in Health Care 15 (Suppl. I), i50–i58. Carayon, P., Wetterneck, T.B., Hundt, A.S., Ozkaynak, M., Desilvey, J., Ludwig, B., et al., 2007. Evaluation of nurse interaction with bar code medication administration technology in the work environment. Journal of Patient Safety 3 (1), 34–42. Castellá, J., Pérez, J., 2004. Sensitivity to punishment and sensitivity to reward and traffic violations. Accident Analysis and Prevention 36, 947. Chan, R., Molassiotis, A., Chan, E., Chan, V., Ho, B., Lai, C.Y., et al., 2002. Nurses’ knowledge of and compliance with universal precautions in an acute care hospital. International Journal of Nursing Studies 39 (2), 157–163. Conner, M., Lawton, R., Parker, D., Chorlton, K., Manstead, A.S.R., Stradling, S., 2007. Application of the theory of planned behavior to the prediction of objectively assessed breaking of posted speed limits. British Journal of Psychology 98, 429–453. Cutter, J., Jordan, S., 2004. Uptake of guidelines to avoid and report exposure to blood and body fluids. Journal of Advanced Nursing 46 (4), 441–452. Dejoy, D.M., 1994. Managing safety in the workplace: an attribution theory analysis and model. Journal of Safety Research 25 (1), 3–17.

753

Dekker, S., 2003. Failure to adapt or adaptations that fail: contrasting models on procedures and safety. Applied Ergonomics 34, 233–238. Dobson, A., Brown, W., Ball, J., Powers, J., McFadden, M., 1999. Women drivers’ behavior, socio-demographic characteristics and accidents. Accident Analysis and Prevention 31, 525–535. Dougherty, E., 1995. ‘Violation’—does HRA need the concept? Reliability Engineering and System Safety 47, 131–136. Dubbert, P.M., Dolce, J., Richter, W., Miller, M., Chapman, S.W., 1990. Increasing ICU staff handwashing: effects of education and group feedback. Infection Control and Hospital Epidemiology 11 (4), 191–193. Ellis, L., Curless, I., 1985. Psychology of the scientist: LI. Noncompliance with reprint requests among behavioral scientists. Psychological Reports 56, 403–406. Espin, S., Lingard, L., Baker, G.R., Regehr, G., 2006. Persistence of unsafe practice in everyday work: an exploration of organizational and psychological factors constraining safety in the operating room. Quality & Safety in Health Care 15, 165–170. Fogarty, G.J., Shaw, A., 2003. Safety climate and the theory of planned behavior: towards the prediction of unsafe behavior. In: Paper Presented at the 5th Australian Industrial & Organizational Psychology Conference, Melbourne, Australia, June 26–29. Gershon, R.R.M., Karkashian, C.D., Vlahov, D., Kummer, L., Kasting, C., GreenMcKenzie, J., et al., 1999. Compliance with universal precautions in correctional health care facilities. Journal of Occupational and Environmental Medicine 41 (3), 181–189. Gershon, R.R.M., Vlahov, D., Felknor, S.A., Vesley, D., Johnson, P.C., Delclos, G.L., et al., 1995. Compliance with universal precautions among health care workers at three regional hospitals. American Journal of Infection Control 23 (4), 225–236. Gould, D., Chudleigh, J.H, Moralejo, D., Drey, N., 2007a. Interventions to improve hand hygiene compliance in patient care. Cochrane Database of Systematic Reviews, Issue 2. Art. No.: CD005186. doi:10.1002/14651858.CD005186.pub2. Gould, D.J., Hewitt-Taylor, J., Drey, N.S., Gammon, J., Chudleigh, J., Weinberg, J.R., 2007b. The CleanYourHandsCampaign: critiquing policy and evidence base. Journal of Hospital Infection 65, 95–101. Griffin, D.W., Ross, L., 1991. Subjective construal, social inference, and human misunderstanding. Advances in Experimental Psychology 24, 319–359. Haims, M.C., Carayon, P., 1998. Theory and practice for the implementation of ‘inhouse’ continuous improvement participatory ergonomic programs. Applied Ergonomics 29 (6), 461–472. Hakkert, A.S., Gitelman, V., Cohen, A., Doveh, E., Umansky, T., 2001. The evaluation of effects on driver behavior and accidents of concentrated general enforcement on interurban roads in Israel. Accident Analysis and Prevention 33, 43–63. Halbesleben, J.R.B., Wakefield, D.S., Wakefield, B.J., 2008. Work-arounds in health care settings: literature review and research agenda. Health Care Management Review 33 (1), 2–12. Hale, A., Heijer, T., 2006. Defining resilience. In: Hollnagel, E., Woods, D., Leveson, N. (Eds.), Resilience Engineering: Concepts and Precepts. Ashgate Publishing Company, Burlington, VT. Hammond, J.S., Eckes, J.M., Gomez, G.A., Cunningham, D.N., 1990. HIV, trauma, and infection control: universal precautions are universally ignored. The Journal of Trauma 30 (5), 555–561. Harbarth, S., Pittet, D., Grady, L., Zawacki, A., Potter-Bynoe, G., Samore, M.H., et al., 2002. Interventional study to evaluate the impact of an alcohol-based hand gel in improving hand hygiene compliance. The Pediatric Infectious Disease Journal 21, 489–495. Heinrich, H.W., 1931. Industrial Accident Prevention: a Scientific Approach, 1st ed. McGraw-Hill Insurance Series, New York and London. Hendrick, H.W., Kleiner, B.M., 2001. Macroergonomics: an Introduction to Work System Design. Human Factors and Ergonomics Society, Santa Monica, CA. Henry, K., Campbell, S., Collier, P., Williams, C.O., 1994. Compliance with universal precautions and needle handling and disposal practices among emergency department staff at two community hospitals. American Journal of Infection Control 22 (3), 129–137. Hobbs, A., Williamson, A., 2002. Unsafe acts and unsafe outcomes in aircraft maintenance. Ergonomics 45 (12), 866–882. Hollnagel, E., Woods, D., Leveson, N. (Eds.), 2006. Resilience Engineering: Concepts and Precepts. Burlington, VT, Ashgate Publishing Company. Hope, J.W., Pate, L.E., 1988. A cognitive-expectancy analysis of compliance decisions. Human Relations 41 (10), 739–751. Horning, L.A., Smith, P.W., 1991. Infection control violations. Infection Control and Hospital Epidemiology 12, 375–672. Hugonnet, S., Perneger, T.V., Pittet, D., 2002. Alcohol-based handrub improves compliance with hand hygiene in intensive care units. Archives of Internal Medicine 162, 1037–1043. Imada, A.S., 1991. The rationale and tools of participatory ergonomics. In: Noro, K., Imada, A.S. (Eds.), Participatory Ergonomics. Taylor & Francis, London, pp. 39–49. Ji, G., Yin, H., Chen, Y., 2005. Prevalence of and risk factors for non-compliance with glove utilization and hand hygiene among obstetrics and gynaecology workers in rural China. Journal of Hospital Infection 59, 235–241. Kanellaidis, G., Golias, J., Zarifopoulos, K., 1995. A survey of drivers’ attitudes toward speed limit violations. Journal of Safety Research 26 (1), 31–40. Karsh, B., Holden, R.J., Alper, S.J., Or, C., 2006. A human factors engineering paradigm for patient safety—designing to support the performance of the health care professional. Quality and Safety in Health Care 15 (Suppl. I), i59–i65. Kobayashi, M., Fussell, S.R., Xiao, Y., Seagull, F.J., 2005. Work coordination, workflow, and workarounds in a medical context. CHI Extended Abstracts, 1561–1564.

754

S.J. Alper, B.-T. Karsh / Accident Analysis and Prevention 41 (2009) 739–754

Kontogiannis, T., Kossiavelou, Z., Marmaras, N., 2002. Self-reports of aberrant behaviour on the roads: errors and violations in a sample of Greek drivers. Accident Analysis and Prevention 34, 381–399. Koppel, R., Wetterneck, T.B., Telles, J.L., Karsh, B.T., 2008. Workarounds to barcode medication administration systems: their occurrences, causes, and threats to patient safety. Journal of the American Medical Informatics Association 15 (4), 408–423. Kraemer, S., Carayon, P., 2007. Human errors and violations in computer and information security: the viewpoint of network administrators and security specialists. Applied Ergonomics 38, 143–154. Kretzer, E.K., Larson, E.L., 1998. Behavioral interventions to improve infection control practices. American Journal of Infection Control 26 (3), 245–253. Laurence, D., 2005. Safety rules and regulations on mine sites—the problem and a solution. Journal of Safety Research 36, 39–50. Lawton, R., 1998. Not working to rule: understanding procedural violations at work. Safety Science 28 (2), 75–95. Lawton, R., Parker, D., 1999. Procedures and the professional: the case of the British NHS. Social Science and Medicine 28, 353–361. Lawton, R., Parker, D., 2002. Judgements of the rule-related behavior of health care professionals: an experimental study. British Journal of Health Psychology 7, 253–265. Li, G., Baker, S.P., 1995. Crash and violation experience of pilots involved in prior commuter and air taxi crashes: a historical cohort study. Aviation, Space, and Environmental Medicine 66 (12), 1131–1135. Locke, E.A., Latham, G.P., 2002. Building a practically useful theory of goal setting and task motivation: a 35-year odyssey. American Psychologist 57 (9), 705–717. Maskerine, C., Loeb, M., 2006. Improving adherence to hand hygiene among health care workers. The Journal of Continuing Education in the Health Professions 26, 244–251. Mason, S., 1997. Procedural violations—causes, costs and cures. In: Redmill, F., Rajan, J. (Eds.), Human Factors in Safety-Critical Systems. Butterworth-Heinemann, Oxford, UK, pp. 287–318. McDonald, C.J., 2006. Computerization can crease safety hazards: a bar-coding near miss. Annals of Internal Medicine 144 (7), 510–516. McKeon, C.M., Fogarty, G.J., Hegney, D.G., 2006. Organizational factors: impact on administration violations in rural nursing. Nursing and Healthcare Management and Policy 55 (1), 115–123. Mesken, J., Lajunen, T., Summula, H., 2002. Interpersonal violations, speeding violations, and their relation to accident involvement in Finland. Ergonomics 45 (7), 469–483. Moore, D., Gamage, B., Bryce, E., Copes, R., Yassi, A., Group, a. o. m. o. T. B. I. R. P. S., 2005. Protecting health care workers from SARS and other respiratory pathogens: organizational and individual factors that affect adherence to infection control guidelines. American Journal of Infection Control 33 (2), 88–96. Naing, L., Nordin, R., Musa, R., 2001. The prevalence of, and factors related to, compliance with glove utilization among nurses in Hospital Universiti Sains Malaysia. The Southeast Asian Journal of Tropical Medicine and Public Health 32 (3), 636–642. Nelsing, S., Nielsen, T.L., Nielsen, J.O., 1997. Noncompliance with universal precautions and the associated risk of mucocutaneous blood exposure among Danish physicians. Infection Control and Hospital Epidemiology 18, 692–698. Noro, K., Imada, A.S., 1991. Participatory Ergonomics. Taylor & Francis, London. Özkan, T., Lajunen, T., 2005. Why are there sex differences in risky driving? The relationship between sex and gender-role on aggressive driving, traffic offences, and accident involvement among young Turkish drivers. Aggressive Behavior 31, 547–558. Özkan, T., Lajunen, T., Chliaoutakis, J.E., Parker, D., Summula, H., 2006. Cross-cultural differences in driving behaviours: a comparison of six countries. Transportation Research Part F 9, 227–242. Parker, D., Lawton, R., 2000. Judging the use of clinical protocols by fellow professionals. Social Science and Medicine 51, 669–677. Parker, D., Manstead, A.S.R., Stradling, S.G., Reason, J.T., 1992. Intention to commit driving violations: an application of the theory of planned behavior. Journal of Applied Psychology 77 (1), 94–101. Parker, D., Reason, J.T., Manstead, A.S.R., Stradling, S.G., 1995. Driving errors, driving violations and accident involvement. Ergonomics 38 (5), 1036–1048. Patarakul, K., Tan-Khum, A., Kanha, S., Padungpean, D., Jaichaiyapum, O., 2005. Cross-sectional survey of hand-hygiene compliance and attitudes of health care workers and visitors in the intensive care units at King Chulalongkorn Memorial Hospital. Journal of the Medical Association of Thailand 88 (Suppl. 4), S287–S293. Patterson, E.S., Cook, R.I., Render, M.L., 2002. Improving patient safety by identifying side effects from introducing bar coding in medication administration. Journal of the American Medical Informatics Association 9 (5), 540–553. Patterson, E.S., Rogers, M.L., Chapman, R.J., Render, M.L., 2006. Compliance with intended use of bar code medication administration in acute and long-term care: an observational study. Human Factors 48 (1), 15–22. Pittet, D., 2001a. Compliance with hand disinfection and its impact on hospitalacquired infections. Journal of Hospital Infection 48 (Suppl. A), S40–S46. Pittet, D., 2001b. Improving adherence to hand hygiene practice: a multidisciplinary approach. Emerging Infectious Diseases 7 (2), 234–240. Polet, P., Vanderhaegen, F., Amalberti, R., 2003. Modelling border-line tolerated conditions of use (BTCU) and associated risks. Safety Science 41, 111–136. Porter, B.E., Berry, T.D., 2001. A nationwide survey of self-reported red light running: measuring prevalence, predictors, and perceived consequences. Accident Analysis and Prevention 33, 735–741.

Porter, B.E., England, K.J., 2000. Predicting red-light running behavior: a traffic safety study in three urban settings. Journal of Safety Research 31 (1), 1–8. Pousette, A., Larsson, S., Torner, M., 2008. Safety climate cross-validation, strength and prediction of safety behavior. Safety Science 46, 398–404. Rasmussen, J., 1997. Risk management in a dynamic society: a modelling problem. Safety Science 27 (2–3), 183–213. Reason, J., 1990. Human Error. Cambridge University Press, Cambridge, UK. Reason, J., 1998. Achieving a safe culture: theory and practice. Work and Stress 12 (3), 293–306. Reason, J., 2008. The Human Contribution: Unsafe Acts, Accidents, and Heroic Recoveries. Ashgate Publishing Company, Burlington, VT. Reason, J., Manstead, A., Stradling, S., Baxter, J., Campbell, K., 1990. Errors and violations on the roads: a real distinction? Ergonomics 33 (10–11), 1315–1332. Reason, J., Parker, D., Lawton, R., 1995. A systems-approach to organizational error. Ergonomics 38 (8), 1708–1721. Reason, J., Parker, D., Lawton, R., 1998. Organizational controls and safety: the varieties of rule-related behavior. Journal of Occupational and Organizational Psychology 71, 289–304. Rebok, G.W., Qiang, Y., Baker, S.P., McCarthy, M.L., Li, G., 2005. Age, flight experience, and violation risk in mature commuter and air taxi pilots. The International Journal of Aviation Psychology 15 (4), 363–374. Retting, R.A., Williams, A.F., 1996. Characteristics of red light violators: results of a field investigation. Journal of Safety Research 27 (1), 9–15. Rogelberg, S.G., Luong, A., Sederburg, M.E., Cristol, D.S., 2000. Employee attitude surveys: examining the attitudes of noncompliant employees. Journal of Applied Psychology 85 (2), 284–293. Romano, E., Voas, R., Trippetts, S., 2006. Stop sign violations: the role of race and ethnicity on fatal crashes. Journal of Safety Research 37, 1–7. Schuman, S.H., Pelz, D.C., Ehrlich, N.J., Selzer, M.L., 1967. Young male drivers: impulse expression, accidents, and violations. Journal of the American Medical Association 200 (12), 102–106. Schwebel, D.C., Severson, J., Ball, K.K., Rizzo, M., 2006. Individual difference factors in risky driving: the roles of anger/hostility, conscientiousness, and sensationseeking. Accident Analysis and Prevention 38, 801–810. Shinar, D., Schechtman, E., Compton, R., 2001. Self-reports of safe driving behaviors in a relationship to sex, age, education and income in the US adult driving population. Accident Analysis and Prevention 33, 111–116. Smith, M.J., Carayon-Sainfort, P., 1989. A balance theory of job design for stress reduction. International Journal of Industrial Ergonomics 4, 67–79. Smith, M.J., Carayon, P., Karsh, B., 2001. Design for occupational health and safety. In: Salvendy, G. (Ed.), Handbook of Industrial Engineering: Technology and Operations Management, 3rd ed. John Wiley and Sons, New York, pp. 1156–1191. Smith, M.J., Karsh, B., Carayon, P., Conway, F.T., 2003. Controlling occupational safety and health hazards. In: Quick, J.C., Tetrick, L.E. (Eds.), Handbook of Occupational Health Psychology. American Psychological Association, Washington, DC, pp. 35–68. Smith, T.J., 2002. Macroergonomics and hazard management. In: Hendrick, H.W., Kleiner, B. (Eds.), Macroergonomics: Theory, Methods, and Applications. Lawrence Erlbaum Associates, Mahwah, NJ. Storr, J., Clayton-Kent, S., 2004. Hand hygiene. Nursing Standard 18 (40), 45–51. Sullman, M.J.M., Pajo, K., Meadows, M.L., 2002. The impact of transport company safety climate on truck crashes. In: de Waard, K.A.B.D., Moraal, J., Toffetti, A. (Eds.), Human Factors in Transportation, Communication, Health, and the Workplace. Shaker, Maastricht, The Netherlands, pp. 121–129. Taveira, A., Hable, P., Karsh, B., 2002. Investigation of the impact of workers’ perceptions on compliance with standard safety precautions in a rural hospital. In: Paper Presented at the Human Factors and Ergonomics Society 46th Annual Meeting, pp. 1467–1471. Taxis, K., Barber, N., 2003. Causes of intravenous medication errors: an ethnographic study. Quality & Safety in Health Care 12 (5), 343–347. Vicente, K.J., 1999. Cognitive Work Analysis: Toward Safe, Productive, and Healthy Computer-Based Work. Lawrence Erlbaum Associates, Mahwah, NJ. Wenner, C.A., Drury, C.G., 2000. Analyzing human error in aircraft ground damage incidents. International Journal of Industrial Ergonomics 26, 177–199. Wills, A.R., Watson, B., Biggs, H.C., 2006. Comparing safety climate factors as predictors of work-related driving behavior. Journal of Safety Research 37, 375–383. Woods, D., 2006. Essential characteristics of resilience. In: Hollnagel, E., Woods, D., Leveson, N. (Eds.), Resilience Engineering: Concepts and Precepts. Ashgate Publishing Company, Burlington, VT. Woods, D., Patterson, E.S., 2001. How unexpected events produce an escalation of cognitive and coordinative demands. In: Hancock, P.A., Desmond, P.A. (Eds.), Stress, Workload, and Fatigue. Lawrence Erlbaum Associates, Publishers, Mahwah, NJ, pp. 290–302. Zeitlin, L.R., 1994. Failure to follow safety instructions: Faulty communication or risky decisions? Human Factors 36 (1), 172–181. Zohar, D., 1980. Safety climate in industrial organizations: theoretical and applied implications. Journal of Applied Psychology 65 (1), 96–102. Zohar, D., 2002. The effects of leadership dimensions, safety climate, and assigned priorities on minor injuries in work groups. Journal of Organizational Behavior 23, 75–92. Zohar, D., Luria, G., 2005. A multilevel model of safety climate: cross-level relationships between organization and group-level climates. Journal of Applied Psychology 90 (4), 616–628.