AN INCIDENT INVESTIGATIONPROCEDURE FOR USE IN INDUSTRYByRob StorbakkenA Research PaperSubmitted in Partial Fulfillment of theRequirements for theMasters of Science Degree inRisk ControlResearch AdvisorDr. Elbert SorrellThe Graduate SchoolUniversity of Wisconsin-StoutMenomonie, WI 54751

2The Graduate SchoolUniversity of Wisconsin-StoutMenomonie, WI 54751Abstract(Writer)Storbakken(Last Name)Rob(First)L.(Initial)An Incident Investigation Procedure for Use In Industry(Title)Risk Control(Graduate Major)Dr. Elbert Sorrell(Research Advisor)May 2002(Month/Year)48(No. of Pages)American Psychological Association Manual(Name of Style Manual Used in this Study)The purpose of the study is to develop an effective incident investigation procedure thatmay be utilized by Company XYZ as a valuable tool to reduce and prevent losses fromoccurring. In order to accomplish this, a review of pertinent literature was conductedalong with first hand participation in incident investigations at a Minnesotamanufacturing facility. By reviewing the literature related to incident investigations andidentifying shortcomings in the existing procedure, recommendations were provided toensure the incident investigation procedure will be a valuable tool to reduce and preventlosses.

3Table of ContentsAbstract2Chapter 1 Statement of the ProblemIntroductionCost of Losses in the United StatesRole of Incident InvestigationStatement of the ProblemPurpose of the StudyGoals of the StudyBackground and SignificanceLimitationsDefinition of Terms4444777788Chapter 2 Review of LiteratureIntroductionCost of InjuriesLoss Causation ModelsIncident InvestigationIncident Investigations at Company XYZSummary10101012202627Chapter 3 MethodologyIntroductionMethod of Study303030Chapter 4 Results and DiscussionIntroductionIncident Investigations at Company XYZElements of Effective Incident Investigation ProceduresA Process to Guide Incident Investigations at Company XYZ3333333538Chapter 5 Summary, Conclusions, and RecommendationsIntroductionSummary of StudyConclusionsRecommendations Related to This StudyRecommendations for Further StudyReferences41414144444647List of FiguresFigure 1 – Heinrich’s Domino TheoryFigure 2 – Cause and Effect DiagramFigure 3 – Improved Incident Investigation Flow Chart142640

4Chapter 1Statement of the ProblemIntroductionThe LocationThe location for this study was Company XYZ in Minnesota (at the request ofthe corporation, the company will remain anonymous). Company XYZ is a heavymanufacturing facility with approximately 1600 employees that operates 24 hours aday, 7 days a week.Costs of Losses in Industry in the United StatesIn the United States in 2000, there were 5200 workplace fatalities. Thiscalculates to 3.8 deaths per 100,000 workers for the year 2000 alone. Also in 2000, 3.9million American workers suffered disabling injuries. Work injuries cost Americans 131.2 billion in 2000, a figure that exceeds the combined profits of the top 13 Fortune500 companies (National Safety Council, 2000). In addition to the losses mentionedpreviously, tens of billions of dollars are also lost in destroyed equipment and materialwhere no injury was involved (Ferry, 1988).Role of Incident InvestigationInjuries that occur in the workplace have many negative effects on a company.Insurance costs increase, productivity may decrease and training costs increase asemployees are moved into unfamiliar jobs as well as other potential drains on companyresources. To preserve both the human and financial assets of a company, it is vital tohave tools in place that focus on reducing losses.

5A common tool to guide the loss reduction efforts of a company is theaccident/incident investigation. Accident/incident investigations most often occur postloss. The function of the investigation is to identify the root causes of accidents anddetermine what corrective action needs to occur to prevent future losses. The root causeis defined by Ted Ferry, an author of several books on incident investigation, as, “Themanagement system that allows the substandard practice to be committed or thesubstandard condition to exist.” Incident investigations may also be performed insituations where loss has not occurred, but has the potential to occur. Larry W. Sorrell,a CSP with significant experience in accident analysis, supports the investigation of allhazards, even concerns, to prevent an accident from happening or happening again(Sorrell, 1998). Whether an investigation occurs post-lost or to prevent a loss, the goalremains the same: identify the root cause and determine corrective action.In addition to identifying the root cause of a loss, accident/incidentinvestigations also serve other purposes. One purpose is to ensure that the injuredemployee receive the workers compensation benefits they are entitled to. A poorlywritten investigation report may result in an injured employee’s claim being denied. Aswell as protecting employees following an injury, the incident investigation process alsofunctions to protect the company from false claims (Sorrell, 1998).Carter and Menckel (1990) expressed the importance of accident/incidentinvestigation procedures by writing, “Most accident prevention efforts are based onknowledge gained from accidents and, consequently, it is important to learn as much aspossible from each accident.” By identifying the root causes of a loss producing event

6and implementing corrective action, the potential for reoccurrence and associated lossesis reduced. This is the value of the accident/incident investigation process.Company XYZ has an investigation procedure in place, but it has a lot of roomfor improvement. The general flow of an investigation begins with a loss producingevent, either injury or damage to product/equipment. The department manager willprint off the generic form from their computer, and write the basic facts of theaccident/incident. Next, the investigation will have to be scheduled which requires asignificant amount of communication to ensure that all members of the team areinformed of the date and time the investigation will take place. The location is alwaysthe same, the plant manager’s office. When the team gathers for the meeting, theaccident/incident is recounted and details provided by the employee, the departmentmanager, witnesses, engineers (if necessary), and the department safety leader. Othersin attendance include the plant manager, members of Health Services, and members ofthe Environmental, Health, & Safety department. If necessary, the team may go to theincident location. Finally, corrective action is determined, the action is written down inthe appropriate section on the printed form, and the meeting is adjourned.The inefficiencies of this process are many and include: near-misses are ignored,too much time is spent to communicate basic information to team members, the currentform is inadequate to meet the needs of the investigation, lack of an accessible databaseof past accident/incidents with corrective actions to allow follow-up of current reportsand to review past action, and the meeting location is inappropriate. From this list ofinefficiencies, it is clear that the process at Company XYZ needs to be revamped fromthe ground up to be a valuable tool to prevent the occurrence of loss.

7When properly done, Company XYZ and its employees will realize the benefitsof a thorough accident/incident investigation. However, when these investigations arenot efficiently performed, the result is a waste of time and effort for all parties involved.Statement of the ProblemThe incident investigation procedure at Company XYZ in Minnesota is not aneffective tool to address the need of the company to investigate loss producing events orconditions. As a result, incidents were either not being investigated at all, or if theywere, the investigation was incomplete. By neglecting to investigate loss producingincidents, or even near-miss activity, an opportunity to prevent/reduce future losses isnot realized.Purpose of the StudyThe purpose of the study was to develop an effective incident investigationprocedure that may be utilized by Company XYZ as a valuable tool to reduce andprevent losses from occurring.Goals of the Study1.0Analyze the current incident investigation process in place at Company XYZ.2.0Identify the elements of an effective incident investigation procedure.3.0Develop a process to guide investigations at Company XYZ that ensuresinvestigations are a useful loss prevention tool.Background and SignificanceIncident investigations have the potential to be a valuable tool when they areeffectively used in industrial settings. When properly used, they serve as an efficientmeans to identify the root problem that resulted in the incident as well as determine

8effective corrective action. Incident investigations function to continuously improveoperations at a facility to reduce human losses, financial losses, and material losses.Conversely, they also have the potential to be exercises in frustration if they areimproperly implemented and used. The current system in place at Company XYZ hasbeen identified as an area of improvement due to the determination that theaccident/incident investigation procedure does not provide an effective, efficient meansto track past accident/incidents, report new ones, and refer to past corrective action todetermine the effectiveness of recommended corrective actions.Profits and losses at Company XYZ are provided in percentages at the request ofthe company. In 2001, profits before interest and taxes were 5.6% of sales whileworkers compensation costs consumed 4.2% of profits. By reducing workerscompensation losses, these dollars will be added directly to profits, increasing the profitmargin of Company XYZ.LimitationsThis study was limited to the needs of a single manufacturing facility. While theprocess fit in well with work practices at Company XYZ, the process may need furthermodification for use in other locations.Definition of TermsAccident – An undesired event that results in harm to people, damage to property orloss to process (Bird & Germain, 1985).Accident/incident – An undesired event that, under slightly different circumstances,could have resulted in harm to people, damage to property or loss to process (Bird &Germain, 1985).

9Immediate Cause – The circumstances that immediately precede the accident.Immediate causes may be broken down into two categories, substandard practices andsubstandard conditions (Bird & Germain, 1985).Basic Cause/Root Cause – The management system that allows the substandardpractice to be committed or the substandard condition to exist (Bird & Germain, 1985).Corrective Action – Activity that is conducted to prevent the existence of asubstandard condition or prevent the substandard act from being committed.

10Chapter 2Review of LiteratureIntroductionThe purpose of the literature review is to provide an examination of the financialimpact of loss to organizations across the United States, in the state of Minnesota, andat Company XYZ. After that, two models of loss causation will be reviewed to gain abetter understanding of the loss process. Further, the literature review will define theincident investigation process and describe its value as an effective tool to prevent lossby interrupting the loss process. Next, a discussion of root cause analysis and cause andeffect diagramming is provided to review their relationship with incident investigations.Finally, a review of the incident investigation procedure that is currently utilized bycompany XYZ in Minnesota will be provided.Cost of InjuriesIn the United States in 2000, there were 5200 workplace fatalities. Thiscalculates to 3.8 deaths per 100,000 workers for the year 2000 alone. Also in 2000, 3.9million American workers suffered disabling injuries. Work injuries cost Americans 131.2 billion in 2000, a figure that exceeds the combined profits of the top 13 Fortune500 companies (National Safety Council, 2000).In 1998, work injuries killed 84 Minnesota workers and approximately 164,000Minnesotans were hurt or became ill from job related causes. Also in 1998, the totalcost of workers compensation for Minnesota companies was estimated at 1 billiondollars (U.S. Bureau of Labor Statistics, 2002). Costs not included in the worker’scompensation figure are indirect costs such as delays in production, administrative costs

11of hiring and training new workers, pain and suffering, and economic costs to injuredworkers and their families not covered by workers compensation (Minnesota SafetyCouncil, 1999). From 1994-1996, traumatic injuries and disorders accounted for 89.6%of the total days away from work with about half of these (46.1%) due to strains andsprains (U.S. Bureau of Labor Statistics, 2002).Profits and losses at Company XYZ are provided in percentages at the request ofthe company. In 2001, profits before interest and taxes were 5.6% of sales whileworker’s compensation costs consumed 4.2% of profits. By reducing worker’scompensation losses, these dollars will be added directly to profits, increasing the profitmargin of Company XYZ.The drains on a company when an employee is injured may be significant,especially when considering the direct and indirect costs of the injury. Direct costs ofan injury include medical costs and worker’s compensation costs while indirect costsmay be far in excess of the direct costs after considering the costs of replacementemployees, extra training, legal costs, overtime, lost product, decreased productivity,and so on. Companies have a financial interest in preventing injuries to theiremployees. The employee also has an interest preventing his or her own injury.Temporary or permanent loss of abilities, disfigurement, and pain that are a result of anon the job injury are just a few of the burdens that the employee will have to bear,potentially for the balance of their life. When actions are taken to reduce injuries, boththe company and the employee enjoy the benefits of the effort. The causes of loss in anorganization can prove to be elusive if the source is not methodically investigated. To

12provide a model to guide investigation procedures, several loss causation models exist,two of which are detailed in the following section.Loss Causation ModelsThere are numerous accident and loss causation models in existence. The twothat will be detailed in this section will be H.W. Heinrich’s Domino Theory and theILCI Loss Causation Model. Loss causation models are used as models for safety andaccident prevention theory. Loss causation models provide a direction of focus for theindividual interested in reducing injuries in an organization. Heinrich’s principles dateback to 1932 and encourage focusing on near misses instead of injury-related incidentsto prevent significant losses from occurring (Fulwiler, 2002). The International LossControl Institute developed their own model in 1985, the ILCI Loss Causation model, toprovide users a tool to control the vast majority of accidents and loss control problems(Bird & Germain, 1985). The ILCI model encourages focusing on development ofstandards, the measurement and evaluation of standards to ensure they are beingfollowed through by members of the organization, and the continuous update ofstandards to provide a means to prevent injuries in an organization. A more detailedaccount of each of these loss causation models follows starting with Heinrich’s theory.H.W. Heinrich’s Domino Theory of Loss CausationIn his 1932 book “Industrial Accident Prevention”, H.W. Heinrich wrote thatthere are five factors in the accident sequence. The first factor is the social environmentand ancestry. Traits such as recklessness, stubbornness, avariciousness, and otherundesirable character traits may be passed along through inheritance. The second factoris the fault of the person. This factor states that inherited or acquired traits of the

13person; such as violent temper, lack of consideration, ignorance of safe practice, etc.,are responsible for the person committing unsafe acts or allowing the existence ofmechanical or physical hazards. The third factor is the unsafe act and/or mechanical orphysical hazard. Unsafe acts include standing under suspended loads, failure to adhereto lock-out/tag-out policy, horseplay, and removal of safeguards. Mechanical orphysical hazards include such items as unguarded machinery, unguarded pinch points,and insufficient light. The fourth factor is the accident. The accident includes eventssuch as slips and trips, being struck by flying objects, being caught in machinery, orcoming into contact with high energy sources. Finally, the fifth and last factor is theinjury. Injuries include fractures, lacerations, etc., that result directly from accidents(Heinrich, 1932).Heinrich then arranges these five factors in a domino fashion such that the fall ofthe first domino results in the fall of the entire row. The domino arrangement illustratesHeinrich’s notion that each factor leads to the next with the end result being the injury.It also illustrates that if one of the factors (dominos) is removed, the sequence is unableto progress and the injury will not occur. While it may be difficult or impossible tochange a person’s attitude (the first and second domino), proper supervision can guidethe person’s behavior so that they do not perform a substandard act or allow asubstandard condition to exist (the third domino) which leads to an accident (the fourthdomino) that leads to an injury (the fifth and final domino).Heinrich also developed a pyramid shaped model to explain the relationship ofnear-miss accidents to minor injuries and major injuries. Heinrich’s pyramid states thatfor every 300 near-miss incidents, there will be 29 minor injuries, and 1 major injury.

14This pyramid summarizes Heinrich’s belief that-near miss incidents must be preventedin order to eliminate the possibility of reaching each successive level of the pyramid.Figure 1 depicts Heinrich’s pyramid.1 MajorInjury29 MinorInjuries300Near Miss IncidentsFigure 1While the terminology and thinking found in Heinrich’s Domino Theory ofAccident Prevention are dated, the process remains worth review. Properly trainingemployees and managers insures that the management system is working in concertwith employees to reduce the occurrence of near misses that, in turn, reduces theopportunity for more severe injuries to occur. A major flaw with Heinrich’s model isthat it relies on a single cause that leads to an incident. Rarely is an incident the productof a single cause; more likely, an incident is the result of several factors that occurredsimultaneously to produce the incident or loss. A more up to date and complete modelof loss causation is the ILCI Loss Causation Model that dates to 1985. Where Heinrichfocused on reducing the incidence of near misses, the ILCI model focuses ondevelopment of performance standards and enforcement of standards to ensure thatemployees are performing their work in a safe manner. With emphasis on performance

15standards, the ILCI model takes a proactive approach to loss prevention and suggeststhat losses are due to a breakdown in these standards. A closer look at the ILCI modelfollows.ILCI Loss Causation ModelThe International Loss Control Institute has developed the ILCI Loss CausationModel. Like Heinrich’s Domino Theory, the ILCI model is based on a sequence ofevents that leads up to an eventual loss. The events in sequential order are, Lack ofControl, Basic Causes, Immediate Causes, Incident/Contact, and Loss (Bird & Germain,1985).Each event has a role in continuing the loss process to its conclusion, the Loss.To facilitate a better understanding of the ILCI model, the events will be reviewed inreverse, starting from the end with the Loss (injury or damage to property) and workingback to Lack of Control (inadequate program or inadequate compliance