A More Definite Risk Management Model

Our risk analysis framework is incomplete. The idea of risking a lot to save a lot is obviously subjective; this is not really a problem and cannot be overcome, and we probably would not like the results if it could. We need the flexibility of subjectivity to ensure our service does not become riddled with bureaucracy on the fireground. The real limitation is the lack of a conceptual framework to define “a lot” on the benefit side. Too often, we are faced with arguments and recriminations after a fire that went wrong because everybody has a different idea about what justifies actions despite everyone’s knowing that we risk a lot to save a lot. We need a way to express our agency’s philosophy on risk and translate that into policy while maintaining flexibility for the incident commander (IC) to operate effectively. This failure to understand, define, and communicate the benefit side of the equation is at the root of the very common misapplication of strategies and tactics.

In the judicial system, the idea of burden of proof gives us a framework for levels of confidence. In a criminal trial, the defendant has to be found guilty beyond a reasonable doubt. A search warrant can be issued if there is a reasonable suspicion. A civil suit is won with a preponderance of evidence. These levels lack concrete definitions and also allow some subjectivity, but they point the user (in this case a judge or a juror) toward choosing a point in a continuum of confidence. For the purposes of our discussion, we can say risk management is analogous. If we are to risk a lot to save a lot, we need to know how to broadly define “a lot.”

Table 1. Savable Life Continuum Range

No reasonable chance

The building is on the ground when you get there or the fire has been going from the basement to the weather vane for 30 minutes.

Possible

This is the range of the majority of our fires; think of it as the standard residential fire. The victims are people such as the teenager who came home from school early without the parents knowing or the visitor forgotten by a highly stressed resident. This is also the abandoned building where we’ve seen vagrants camping.

Probable

This is the example we hear all the time when discussing size-up: the house with the cars in the driveway and toys in the yard at 0300 hours with no one to meet you. It is also the mercantile occupancy around lunchtime or the nursing home.

Highly likely

This is as sure as you can be. A wife says her husband went back inside to get the kid, or the nursing director says they could not get to a patient in time. We cannot be certain the person did not jump out of a back window, but we have to operate as if the occupant is trapped and imperiled.

Defining “A Lot”

On the risk side of that equation, we can easily define “a lot,” and fire service texts frequently do. The ultimate is a firefighter fatality; a lot, but not the most, is a serious firefighter injury; a lot, but still less, is something like destruction of equipment or minor injuries. We can readily define these terms because they are personal values that affect us directly. What is not so easy is the right side of the equation. In one sense “a lot” is obvious—a life. But there are actually two parts to this, and it can be more fully and usefully expressed as a probability. Can we save a life? Is there life left to save? How confident are we in that assessment? It is that confidence that we need to name to make the risk management equation meaningful on the street.

I propose we use something akin to the burden-of-proof framework when determining how much we should risk. We will never be able to know anything inside a burning building with certainty until we go inside and lay hands on it. Our subjective judgment, borne of experience, information gathered on the spot, and training, has to guide us. We must reach some level of confidence in the idea that someone savable is or is not in any given structure fire. Note that I am ignoring questions of how much our response can accomplish given its resources, experience, and training. This is simply a question of placing a name on how confident we are that there is a savable life. This will also be subjective and broad. It has to be to be useful.

Note that I say a “savable life.” There is an element here of what some have termed victim survivability profiling, but that systematic approach is too imprecise and depends too heavily on information we cannot possibly know. For our purposes, we can very broadly assume that some fires are too severe and some buildings too far gone for anyone inside to be alive. These are the “I know it when I see it” incidents. In other cases, we should assume that any potential victim is savable. If we are to retain the public trust and fulfill our duty, we must operate in a risk forward fashion that tends to consider victims savable. In other words, we should not and cannot be defeatists.

The tough questions we are paid (or funded) to answer are those about whether someone is inside the burning structure. Let’s think of the probability of a savable life inside a structure fire as lying along a continuum. On the far left, we have zero percent, and on the far right 100 percent. We cannot reach the ends of the continuum in the real world. In fact, any particular number value cannot be determined. The best we can do is to make our estimate and place it within a range on that continuum. Each range should allow us to more rationally evaluate the chances that someone is inside and make our decisions about whether to extend ourselves and place our members at varying levels of risk.

The ranges are simple and, again, analogous to the burden of proof. We can have a reasonable suspicion someone is in there, believe by a preponderance of the evidence that someone is in there, or believe beyond a reasonable doubt that someone is in there. This is nothing more than an index of suspicion. Let’s lay that out graphically and use our own terminology.

We already have these distinctions in our heads. The familiarity of the examples indicates that evaluating this index of suspicion is nothing new. Naming these ranges allows us to turn these size-up cues into policy tools. We have to operate with standard operating guidelines (SOGs) if we are to improve. Standard actions applied to standard incidents should result in standard outcomes. Nonstandard outcomes are reasons to reevaluate our procedures. By naming these ranges, we can lay down standard procedures for taking a risk.

Table 2. Range-Action Policy

No reasonable chance

We will not deviate from our SOG.

Possible

Our SOGs are written with this benefit level in mind.

Probable

We will suspend our initial rapid intervention team requirement and allow the first-in company to deploy all members in the immediately dangerous to life or health area.

Highly likely

Incident commanders may deploy all reserves on scene and conduct targeted actions, such as vent-enter-isolate-search or searching ahead of a hoseline or before a charged hoseline is in place and allow expanded calculated risks for short duration in conditions that in another fire might indicate withdrawal.

Standard Operating Guidelines

SOGs tell us certain things we cannot do for each range. If we incorporate these ranges into our policy manuals, we can use these terms as shorthand and help our ICs identify how much the organization is willing to risk. For example, we might have an SOG on lightweight construction that stipulates that we do not operate under or above a fire in a truss loft unless there is a high likelihood of a savable victim in the structure. An SOG on fire attack might require a possible savable victim as a precondition for operating without a secured water supply. Alternatively, we could give examples of actions we are willing to countenance corresponding to each range.

The important thing to remember is that this is an analytical tool for writing policies and expressing the agency’s philosophy on risk. It is not to be used as a checklist or a new way to conduct size-up; it is merely a way to flesh out what a good fire department already has in place. The terms can be whatever your agency wants them to be, and the examples of actions and conditions will necessarily be decided by your agency. This has the additional benefit of allowing fire chiefs in underresourced departments to illustrate the consequences of budgets.

Decades of investigations have identified inadequate size-up, improper strategic modes, and strategic/tactical mismatch as major threats to firefighter safety. The fire service must recognize that the risk management model as usually formulated is not particularly useful as a policy instrument. We need a better, more definite method for specifying when and why we would allow a particular level of risk.

Patrick S. Mahoney has been in the fire service since 1998 and is a battalion chief with the Baytown (TX) Fire Department. He has a master’s degree in fire and emergency management administration from Oklahoma State University, is a graduate of the National Fire Academy’s Managing Officer, and is a member of the board of directors of the Company Officers Section of the International Association of Fire Chiefs.

No posts to display