This module introduces you to the concepts of assessment and analysis as they relate to the accident investigation process. We'll review some theories of accident causation and discuss the process of developing and analyzing the sequence of events occurring prior to, during, and immediately after an accident.
So far, you have collected a lot of factual data and it's strewn all over your desk. The task now is to turn that data into useful information. You've got to somehow take this data and make some sense of it.
It's important to know that you're not gathering all of this information just to conduct an assessment of what was and was not present immediately prior to the accident. You're actually conducting an analysis to determine specifically how surface causes (behaviors and conditions), and the underlying root causes (system weaknesses) contributed to the accident. To better understand this, let's take a closer look at what the process of "analysis" is.
Webster defines analysis as the, "separation of an intellectual or substantial whole into its parts for individual study."
When an accident occurs, we need to separate or "break down" the "whole" accident process into its component "parts" for study to determine how they relate to the whole accident.
Since the accident, itself, is the main event, its component parts may be thought of as the individual events leading up to and including the main event or the accident.
The accident investigator's challenge is to effectively assess each event to identify the presence or absence of behaviors and conditions, and then analyze those behaviors and conditions in each event to determine if and how they contributed to the accident. To do this we need to make some basic assumptions about the factors that cause or contribute to accidents.
Over the past century, safety professionals have tried to more effectively explain how and why accidents occur. During the early years the initial explanations were at first rather simplistic. Theorists gradually realized that it was not sufficient to explain away workplace accidents as simple cause-effect events. They developed new theories that better explained the result of complicated interactions taking place among conditions, behaviors and systems. With this in mind, let's take a look at some of these theories.
"Common sense" leads us to this explanation. An accident is thought to be the result of a single, one-time easily identifiable, unusual, unexpected occurrence that results in injury or illness. Some still believe this explanation to be adequate. It's convenient to simply blame the victim when an accident occurs. For instance, if a worker cuts her hand on a sharp edge of a work surface, her lack of attentiveness may be explained as the cause of the accident. ALL responsibility for the accident is placed squarely on the shoulders of the employee. An accident investigator who has adopted this explanation for accidents will never look beyond perceived personal employee flaws to discover the underlying system weaknesses that may have contributed to the accident.
This explanation describes an accident as a series of related occurrences which lead to a final event which results in injury or illness. Like dominoes, stacked in a row, the first domino falling sets off a chain reaction of related events that result in an injury or illness.
The accident investigator who has adopted this approach will assume that by eliminating any one of those actions or events, the chain will be broken and the future accident prevented. In the example above, the investigator may recommend removing the sharp edge of the work surface (an engineering control) to prevent any future injuries. This explanation still ignores important underlying system weaknesses or root causes for accidents.
This explanation takes us beyond the rather simplistic assumptions of the single event and domino theories. Once again, accidents are not assumed to be simple events. They are the result of a series of random related or unrelated actions that somehow interact to cause the accident. Unlike the domino theory, the investigator realizes that eliminating one of the events does not assure prevention of future accidents. Removing the sharp edge of a work surface does not guarantee a similar injury will be prevented at the same or other workstation. Many other factors may have contributed to an injury. An accident investigation will not only recommend corrective actions to remove the sharp surface, it will also address the underlying system weaknesses that caused it.
When we understand that the accident, itself, is actually the final event in a complex series of events, we'll naturally want to know what the initiating events were. When the initiating events occur, they effect, in one way or another, the workplace conditions and actions of others, setting in motion a potentially very complicated process that eventually ends in an injury or illness. The trick is to take the information gathered and arrange it so that we can accurately determine what initial conditions and/or actions transformed the planned work process into an unintended accident process.
In this step, take the information you have gathered to determine the events prior to, during, and after the near miss/injury accident. It is important to note that a serious injury accident can easily be the result of 20 or more events. Events can occur anytime, anywhere, any place, and to anyone. It is possible that pertinent events may have occurred many weeks or months before the accident.
There are four categories of events:
Actual Events: These are events that you are able to determine actually occurred i.e., an event that is witnessed by one or more persons (two or more is best) and they can verify it actually happened. You would want to interview all witnesses to the event.
Example - Bob and Bobbie saw Bill turn off the chipper power switch and then walk over and reach into the chipper in an attempt to remove some jammed wood.
Assumed Events: These are events that must have happened but have not yet been verified. Flag these somehow to remind you that more investigation is needed. Assumed events are harder to establish. In any step-by-step process, you can't get to step 3 without first doing the first two steps. If a worker is injured at step 3, you may assume he accomplished steps 1 and 2 unless, it is established that he bypassed the first two steps. If completing steps 1 and 2 will prevent an injury at step 3, you may assume the worker did not do steps 1 or 2.
Example - If Bill's hand was crushed while clearing a piece of wood that was stuck in a large chipper, we may assume he did not perform lockout/tagout, or we may assume that he performed lockout/tagout incorrectly. Only further investigation and analysis will uncover what actually happened.
Non-Events: If an event was supposed to happen, but did not, that is a non-event. Although non-events describe an event that did not occur, they should be captured because they may help discover conditions and behaviors relevant to the investigation.
Example - Bill did not try to start the chipper to verify lockout/tagout was successfully performed. He failed to perform the verification step of the lockout/tagout procedure.
Simultaneous Events: In some accident scenarios two or more events occur at precisely the same time resulting in a hazardous condition or set of unsafe behaviors that cause an injury.
Example - Ralph wondered why the chipper was off and turned it back on at the same instant in time that Bill reached into the chipper to remove the jammed wood.
Our challenge at this point in the investigation process is to accurately determine the sequence of events leading up to the accident so that we can more effectively understand why the accident event, itself, happened. Once the sequence of events is developed, we can then study each event in the sequence to determine the related causal factors below.
(Hold on... we'll study more about these three elements in the next module.)
In the multiple-cause approach to accident investigation, many events may occur, each somehow contributing to the final event. For instance, if a supervisor ignores an unsafe behavior because doing so is not thought to be his or her responsibility, the failure to enforce safe behavior represents an event in the production process that may contribute to or increase the probability of a future accident.
Each event in the unplanned accident process is composed of two components: an actor and an action.
It's important to understand that when describing an event in writing, first identify the actor and then tell what the actor did. Remember, the actor is the "doer," not the person or object being acted upon or otherwise having something done to them. For instance, take a look at the event statement below:
In this example, "Bob" is the actor and "unhooked" describes the action. First, we describe the actor — Bob. Next, we describe the action — unhooking. The lifeline and harness, although "objects" are not actors because they are not performing an action. Rather, something is being done to them. Also note that the statement is written in active tense.
To get a good idea of what the sequence of events looks like, review the example below that was prepared for an actual fatality investigation conducted by an OSHA accident investigator a few years ago.
Make sure you are constructing only one event
If an event is hard to understand, it may be that the description is too vague or general. The solution to this problem is to increase the detail. We can use two strategies to increase detail:
It's important that the sequence of events clearly describe what occurred so that someone who is unfamiliar with an accident is able to "see it happen" as they read the narrative.
Sample sequence of events
Here is another example that shows how a sequence of events can be developed using cards. Describe each event and then arrange the events on your desk or a wall in the proper sequence.
Read the material in each section to find the correct answer to each quiz question. After answering all the questions, click on the "Check Quiz Answers" button to grade your quiz and see your score. You will receive a message if you forgot to answer one of the questions. After clicking the button, the questions you missed will be listed below. You can correct any missed questions and check your answers again.