Home > Consulting > Audits/Assessment > Predicting Injurious Events

Insights Into Predicting Injurious Events

By Roger J. Kahler and J.Nigel Ellis

This article was a feature of the Occupational Health & Safety journal of America

The question should be asked: Are there lessons for the United States based on the Australian experience with disabling injuries? We would suggest there are significant lessons. Australia is a country with comprehensive and extensive national standards, codes of practice at national and state levels, advisory and compulsory standards at a state level, relevant government departments and inspectorates, consultants, university-accredited courses for occupational health and safety, an active Ergonomics Society, etc. Yet the fundamental problem of effective reduction in the likelihood of permanent disability in the workplace remains unresolved at a national level within Australia.Progress is possibly being made in the overall likelihood of an Australian fatality. This is positive, but it is not the area where the most significant progress must be made. Management of permanent disability is the number one issue, yet it remains camouflaged in databases. Accurate descriptions remain elusive, and organizations continue to work in blissful ignorance of their exposures. Permanent disability in Australia cannot be accurately described except at very high levels of description (e.g. falls of people, body overstressing). Databases that yield significant insight are missing there. It is suggested that this is also the U.S. experience.For progress to occur, two things are needed. First, that those who are empowered to ensure the collection of accurate descriptive data and subsequent analysis, undertake this exercise. Second, that employers, both large and small, are strongly encouraged to apply a series of focusing questions to their employees that can yield significant insight into the exposures for potential permanent disability.

The Size of the Problem

It is possible to classify damage (i.e., consequence) into three classifications: Permanent (Fatal and Non-Fatal), Temporary, and Minor (Geoff McDonald & Associates, Brisbane, Qld, Australia).

	Class I--Permanent
	Class II--Temporary
	Class III -- Minor

Class I Damage alters the future of an individual permanently and includes fatality, impaired back, quadriplegia/paraplegia, disfigurement, amputation, and psychological disturbance. Class II Damage alters the future of an individual temporarily from fractures, sprains, lacerations requiring sutures, and contusions. A person is expected to fully recover from temporary damage. Class III Damage does no more than inconvenience the person from minor cuts and bruises. This damage causes discomfort but allows the person quickly to carry out normal duties.

Table I shows the relative cost and size of the various categories of personal damage arising from Australian industry. The data were produced in 1995 and still remain one of Australia's most comprehensive studies. Tragically, 50,000 Australian workers each year are permanently disabled and approximately 700 die from injury-related causes.

Table I: Number of Cases & Cost of Damage (Australia 1992-93)
Severity of Injury or Disease
	<5 days	>5 days, full duties	>5 days, reduced duties	>5 days, lower income (permanent impairment)	Permanently Incapacitated (no return to work)	Fatal
Number of occurrences	144,053	123,395	78,333	30,728	19,290	693
Percent of occurrences	36.33	31.12	19.75	7.74	4.86	0.17
Cumulative percentage	36.33	67.45	87.20	94.94	99.8	100
Cost of occurrences ($billions)	0.136	1.063	2.415	4.555	11.664	0.299
Percentage of cost	0.67	5.28	11.99	22.62	57.93	1.48
Cumulative percentage	0.67	5.95	17.98	40.56	98.49	100
Source: Industry Commission Report on Work Health and Safety--Enquiry into Occupational Health & Safety, April 1995.

The total cost derived from direct/indirect costs is $20 billion and is allocated to:

	Permanent--fatal $0.3 billion
	Permanent--non-fatal $16.4 billion
	Temporary -- $3.6 billion

Key Points of Concern

The key learning from these data can be summarized as follows. If the Australian personal damage problem were analyzed not in terms of cost, but in terms of impairment (percentage loss of whole bodily function), a similar conclusion emerges. The reader must be impressed with the notion that within Australia, the majority of the pain, suffering, cost, and impairment from work-related damage is associated with those 12 percent of non-fatal permanent disability cases. These people are not amputees; they are people who carry soft tissue injury and outwardly appear fully functional until they have to execute bodily movements and impose loads upon the soft tissue (ligaments, tendons, muscle, and intervertebral disc) that are damaged.

Analysis of Table I shows that 82 percent of the cost is associated with fatal and non-fatal Class I permanent damage coming from 13 percent of incidents. Therefore, it is clear that from a cost perspective, the priority must be the systematic identification and management of the likelihood of Class I--permanent disability. The potential for fatalities and disasters must be predicted and effectively managed because these can present sovereign risk to an organization.

Some insight into the effectiveness of Australian industry and government approaches to health and safety over the last 10 years can be gained by reviewing workplace injury statistics. Over this period, the risk management approach has been widely advocated.

Table II data are derived from New South Wales worker's compensation statistics (Statistics Branch, New South Wales Workers Compensation Statistical Bulletin 1998-1999, WorkCover New South Wales, 2000). They illustrate the likelihood--expressed as one case per "X year of work" or "worker years"--of a worker sustaining a permanent disability and or injury resulting in more than six months off work.

Table II
Likelihood of Permanent Disability for New South Wales
Year	Likelihood
1991-92	1:384
1992-93	1:357
1993-94	1:270
1994-95	1:208
1995-96	1:208
1996-97	1:196
1997-98	1:185
1998-99	1:192
(Note: Does not include travel/commuting or disease).

Between 1991-92 and 1998-99, the following picture emerges when the NSW compensation database is analysed with respect to the likelihood of damage. It is the only Australian database that allows such analysis.

	All classes of injury: 7 percent decrease
	Class II--less than six months off work: 25 percent decrease
	Class I--Six months and more off work: Double
	Class I--Permanent disability: More than double
	Class I -- Fatality: Possible slight decline (not convincing)
	Cost of claims: 80 per cent decrease

It is suggested that readers complete the following calculation for their organizations.

Number of employee years worked in the last 10 years = ? Number of employee impairment injuries in the last 10 years = ?

Likelihood of permanent impairment in any one year = Number of permanent injury cases divided by Number of employee years = ?

There are no published guidelines for acceptable figures of the likelihood of permanent disability. In Australia, the likelihood of a fatality is approximately 1:14,000 employee years. The U.S. Labor Department's Bureau of Labor Statistics states that in the year 2000, for private industry, there were 5,344 fatalities for a workforce of 110 million workers. The likelihood of a fatality in the United States is 1:21,000 employee years. The likelihood of a work-related fatality in the private industry in the U.S.would appear to be similar to that in Australia.The likelihood of permanent disability in the State of New South Wales (Australia) is approximately 1:200 employee years. Research of the BLS does not allow a determination of this likelihood ratio, but we can say the likelihood of a U.S. private-sector employee being off work for 31 days or more is 1:314 employee years.More research is required to answer such a basic question as to the likelihood of permanent disability (Class I damage) for U.S. industries. We would suggest the U.S. experience with respect to the likelihood of permanent disability is far more frequent than it should desirably be. If a goal were to be set of a fatality and permanent disability likelihood of 1:100,000 employee years and 1:30,000 employee years, respectively, considerable progress is necessary.For Australia, industry and government approaches to health and safety have reduced incidents producing temporary damage but have failed to address the high-cost Class I problem associated with non-fatal permanent damage.There is a permanent disability paradox. There are so many, and yet they are so rare. In the life of an organization permanent disability does not occur frequently, but within Australian industry it occurs 1,000 times per week. How often does it occur in U.S. industry?

The Nature Of The Problem

Damage to people can be considered to be a consequence of an energy exchange. Energy is simply the capacity to do work. Damage to people occurs when the energy exposures exceed the tolerable limits of the person. Energy can be loosely, but usefully, classified as follows.

· Human Energy--Physical muscular exertion of varying intensity and duration
· Gravitational Energy--People and objects falling
· Vehicular Energy--Single- and multi-vehicle collisions, pedestrian strikes, vibration, jolting and jarring
· Machine Energy--Fixed and portable machinery in operation
· Object Energy--Person contacted by moving objects either constrained or unconstrained in their movement path (e.g., projectiles)
· Electrical Energy--Contact with electricity
· Thermal Energy--Extremes of temperature
· Chemical Energy--Chemicals damage the body through absorption, inhalation, ingestion, or contact with skin/muscle tissue
· Noise Energy--Exposure to a noise source of varying intensity and duration
· Other Energy Sources--Animal, biological, radiation, etc.

An energy exchange can be considered to have a time/intensity relationship (dose) and has been grouped by McDonald & Associates into one of three classifications.

Type A Damage Single, traumatic energy exchange

Examples:

	Electric shock
	Hit by fast-moving object
	Burned by flames

Type B Damage A series of discrete energy exchanges, each not affecting the function or generating pain, but each reducing the damage limit. The cumulative effect is damage.

Examples:

Lifting, pushing, or pulling tasks leading to back damage

Type C Damage Continuous exposure to small energy exchanges, which produce cumulative damage.

Examples:

Continuous exposure to:

	Repetitive movements leading to repetitive strain injuries (occupational overuse syndrome)
	Noise
	Prolonged postural displacement
	Chemicals

Ride vibration leading to back damage

Analysis of data sets within Australia reveals a very consistent pattern. Human, gravitational, and vehicular energies are those few energy sources that contribute 80 percent of permanent disability.

Table III is typical of work completed by InterSafe and is extracted from a classification of 1,231 cases of permanent, non-fatal damage for the New South Wales and Queensland coal mining industry. It shows the very high involvement of human engergy, gravitational energy, and machine energy in permanently damaging people.

Table III
Classification of 1,231 Permanent Disability Cases for Open Cut and Underground Coal Mines of Queensland and New South Wales
Energy Type	Number of People
Human Energy	491
Gravitational Energy	398
Machine Energy	254
Object Energy	36
Thermal Energy	16
Chemical Energy	12
Susceptible Part	7
Anxiety/Stress Disorder	7
Oxygen Deprivation	1
Heart Attack	1
Biological Energy	1
Specialised Shape	2
Insufficient information	5

Therefore, the Pareto principle (the 80/20 Rule) applies at least twice in understanding where the majority of damage arises in the industrial work environment. First, the majority (more than 80 percent) comes from permanent, non-fatal damage. Second, a review of many databases shows human energy, gravitational energy, and vehicular energy are over-involved in producing non-fatal damage.

There is a need to understand the pattern of non-fatal permanent damage. We commenced an extensive search for quality descriptive data of non-fatal permanent disability of any type. It could not be obtained because of confidentiality and/or extremely poor-quality descriptive data. Coded data often are not suitable for taxonomic analysis and do not yield satisfactory insight. It is possible to obtain quality descriptive data with respect to fatalities, and InterSafe has been successful in obtaining quality descriptive data from OSHA and completed pattern analysis of fatalities, which yielded significant insight and assisted a number of Australian organizations to make progress. However, it is beholden on insurers and appropriate government agencies to collect and report on comprehensive descriptive data with respect to permanent disability.

What We're Missing, and Why

This paper already has suggested the absolute necessity of relevant government and research organizations to collect and publish pattern analysis of non-fatal permanent disability data. Industry can embark upon other activities while waiting for such progress by others. It is, therefore, necessary for industry to establish where the actual and potential damage experience of an organization exists within that organization.

With respect to incident records, it is important to realize the database within an organization is essentially of Class II and Class III (temporary and minor damage) and is unlikely to yield insight into future permanent damage. You must engage other strategies. Further, the incident database can contain significant distortion with respect to predicting future damage. The table below is taken from an analysis of the incident database of an open cut coal mine. It shows the ratio of personal damage incidents to all recorded incidents for the dominant energy types that permanently damage people. This pattern has emerged in at least three studies completed by InterSafe.

Table IV
Ratio of Damaging Incidents to All Reported Incidents by Energy Type
	Personal Damage Incidents	All Recorded Incidents	Ratio
Vehicular Energy	54	366	1:7
Gravitational Energy	106	235	1:2
Human Energy	336	389	1:1

It can be observed that there is an apparent sensitization to vehicle-related incidents but a significant de-sensitization to human energy and gravitational energy incidents. There are many reasons why this takes place. With respect to human energy, people often do not report a lifting/pushing/pulling incident unless it generates pain. However, there is significant difference between the concept of damage and pain with respect to that frequently and permanently damaged structure, the intervertebral disc, and, in particular, the lumbar spine. That is, a person can experience damage without any perception of pain. This is an important understanding because it means there is a store of knowledge within the workforce about heavy lifting/pushing/pulling tasks, which are predictably overloading the musculoskeletal structure and generating longer-term damage.With respect to gravitational energy, there appears to be a desensitization to work situations that have the potential to generate permanent damage from people falling either to the same level (through loss of grip at heel strike or contaminant present), falling while climbing up or down (fixed or mobile plant), and people falling from height. It would be fair to say that some sections of industry have an increasing sensitivity to work at height. However, industry has not yet come to understand the issues associated with poor footholds and hand holds while completing tasks or while accessing equipment.

While the recorded incident databases can be useful, there will be a store of knowledge within the workforce that is not perceived as an incident, but simply is lying dormant within the person's experience. It then remains to determine how one accesses that store of knowledge. It is achieved by interviewing either individuals or small groups using a framework of focusing questions, which cause a person to have a definite frame of reference in organizing his thinking patterns during the interviewing process. For example, imagine the difference in the information that can be obtained by asking, where do you think you are most likely to be injured on this site? versus:Question A: Please describe to me task you complete that you would subjectively describe as heavy or very heavy lifting/pushing or pulling tasks.Question B: Please describe to me where you work at height where, if you were to overbalance or fall to accommodate some critical information at an appropriate time, you could fall 1 meter or more and be seriously injured.Question C: Please describe to me surfaces in your workplace where your foot has slipped forward rapidly as you were walking and/or working.The difference in the questioning is very simple but profound in terms of the results it produces. It is possible to develop a set of focusing questions against the energy types known to damage people - electrical, mobile equipment, human, gravitational, chemical, etc.

Harvesting Employees' Knowledge There are a number of recorded techniques in the literature for harvesting the store of information within a workforce. The one recorded over the longest duration is a technique known as Critical Incident Recall (in excess of 90 years). Focused Recall and Perception Analysis are other techniques. The process involves the following three steps:

Problem identification of potentially permanent damage based on a workforce's experience set against a framework of focusing questions.

Prioritizing problems followed by analysis using an appropriate model to generate solutions.

Implementation of solutions followed by an audit to determine effectiveness.

A significant factor in this predictive process is the correct combination of site knowledge with outside expertise. That expertise either can be employed into an organization or is transferable into the organization so that the organization itself combines expertise with site knowledge. Expertise should be inherently organized and communicable--experience is not. Therefore, the combination of expertise and experience has the potential to document the detail of experience against the generalized pattern of permanent personal damage.One of the fundamental principles underlying the harvesting of experience is as follows:

Principles:

	10 people with 15 years of experience each
	150 years of exposure
	Possible likelihood of one permanent disability case.

Harvesting the information contained in the store of knowledge of a workforce is a problem-identification process. However, it is necessary to have appropriate goals when embarking upon such a process. A summary of appropriate management goals could be as follows:Goal 1. We will direct 80 percent of our safety effort toward the prediction and management of future potential permanent personal damage.Goal 2. We will maximize our understanding of future potential high-consequence incidents. This requires that those who are allocated to problem-solving be prepared to challenge their own experience base, as well as the industry norms with respect to how tasks are completed.Goal 3. We will apply a multi-factorial model in understanding potential future damage, as opposed to a single factorial model.A single factorial model often uses "cause/effect" terminology in which people try to understand basic causes, root causes, main causes, etc. However the application of a model where understanding and insight is obtained through asking non-value, non-judgmental, non-emotive questions: What did people do and what did people not do that could be essential in the propagation of damage? What features of equipment are present and what features of equipment are absent that could be essential in the propagation of damage?Note that the terms "safe" and "unsafe" aren't used in these questions. They are value judgment terms and could result in rejection of information before it is recorded. Judgments with respect to "safe" and "unsafe" are different between cultures, different between and within organizations, and different between and within individuals.Expectation plays an important role in processing information with respect to future potential damage, and particularly with respect to selecting appropriate solutions. A common expectation is that 88 percent of accidents are caused by human error, 10 percent by machine design, and 2 percent by Acts of God (the 88:10:2 rule). Not only is this scientifically nonsensical, it is also theologically nonsensical. The only correct statement that can be made is, in actual and potential incidents, behaviour factors, design factors, and environmental factors were either present or absent in 100 percent of cases. The correct ratio is 100:100:100. Therefore, when harvesting experience, expect to observe contributions from people, equipment, and the working environment.

If you were to approach a study of tasks that had the potential to create future permanent personal damage with the following four major areas of control as their dominant information organizers, their expectation will have a significant influence on the final recommendation.

	Control Measure 1: The person was not adequately trained.
	Control Measure 2: The person was not following procedures.
	Control Measure 3: The procedures were inadequate.
	Control Measure 4: The person was not wearing appropriate PPE.

These categories of control measures are commonly observed on Incident Report Forms. They create an expectation with respect to control measures when incidents are being analysed. If one takes the major potential damaging energy source of human energy, it is common to find that statements are made with respect to the person not following correct procedures, not being adequately trained, the training was not adequate.Having investigated many hundreds of human energy/simple lifting permanent damage cases, it is most frequently the case that well-established scientific guidelines with respect to acceptable moments (load x distance) of lift are exceeded. The foregoing set of expectations with respect to training and procedures will not yield the necessary gains with respect to the management of human, gravitational, and vehicular energies.It is possible the focus on people as a control measure at the point of task has plateaued in terms of its ability to influence future personal permanent damage, and that a different and more effective hierarchy of controls has to be more widely applied.

Summary

This paper has attempted to quantify the size and the nature of the industrial personal damage problem, in terms of not only classifying that damage but also showing the predominant damaging energy types. We have shown the majority of the cost of work-related personal damage (in Australia) is from non-fatal permanent disability. It is not possible to make this statement as clearly for the United States.We can find no clear identification of the permanently disabled group for U.S. industry within the national dataset. This could indicate it is being camouflaged, as it was in Australia until definitive work was completed in the early 1990s.The enormous value of the knowledge of the workforce has been expounded. The collection and recording of such knowledge requires a very structured approach using a set of focusing questions that are established against a backdrop of what is known to damage people permanently. In particular, the questions must contain reference to human energy (heavy lifting/pushing/pulling tasks), gravitational energy (falls of people, objects), and machine energy (vehicle/vehicle, vehicle/pedestrian, vehicle/environment).Having harvested the store of knowledge of a workforce, it is necessary to have appropriate goals and expectations in responsibly handling the information you have obtained.

◄ Back