Risk Compensation-Can Improved Safety Degrade Performances?

Abstract

In Brief The management and control of risk are a priority activity in both the design and utilization of medical devices. An important but often overlooked aspect of risk management is how people and organizations respond to changes in risk. Such changes, although they may be small, might have important safety considerations. The hypothesis of risk compensation is that people experiencing a real or perceived change in the riskiness of an activity will deliberately or subconsciously alter their behavior to obtain a new combination of risk and reward. In other words, risk compensation is the achievement of a new balance between behavior and the perceived risk. This is in contrast to the typical albeit tacit assumption that risk will always go down as a result of a technical or system risk-reducing improvement. Instead of the risk actually going down, the behavioral result of risk compensation can be to maintain a level of risk that is relatively constant. This level, known as the target level of risk, is different for everyone because it is based on the perception of the risk and the level at which the individual feels safe. Individuals will compensate for risk reduction with different actions and to different degrees, depending in part on the particular situation. The ultimate risk may become reduced but not as reduced as was theoretically possible. The worst scenario is when the change in behavior is so inappropriate that the risk actually goes up. This can occur in particular when the user's perception of the degree of risk improvement is more than the actual improvement, that is, the systematic change has not reduced the risk as much as the users of the system think is the case. If there is risk compensation, then the compensation can more than offset the improvement. The concept of risk compensation is important because it means that the risk reduction that could theoretically be obtained by a system change will not be fully achieved in the absence of strict controls because the users of the system will alter their behavior in a way that seems favorable to them but that is adverse to overall safety. The management and control of risk are a priority activity in both the design and utilization of medical devices. An important but often overlooked aspect of risk management is how people and organizations respond to changes in risk. Such changes, although they may be small, might have important safety considerations. The hypothesis of risk compensation is that people experiencing a real or perceived change in the riskiness of an activity will deliberately or subconsciously alter their behavior to obtain a new combination of risk and reward. In other words, risk compensation is the achievement of a new balance between behavior and the perceived risk. This is in contrast to the typical albeit tacit assumption that risk will always go down as a result of a technical or system risk-reducing improvement. Instead of the risk actually going down, the behavioral result of risk compensation can be to maintain a level of risk that is relatively constant. This level, known as the target level of risk, is different for everyone because it is based on the perception of the risk and the level at which the individual feels safe. Individuals will compensate for risk reduction with different actions and to different degrees, depending in part on the particular situation. The ultimate risk may become reduced but not as reduced as was theoretically possible. The worst scenario is when the change in behavior is so inappropriate that the risk actually goes up. This can occur in particular when the user's perception of the degree of risk improvement is more than the actual improvement, that is, the systematic change has not reduced the risk as much as the users of the system think is the case. If there is risk compensation, then the compensation can more than offset the improvement. The concept of risk compensation is important because it means that the risk reduction that could theoretically be obtained by a system change will not be fully achieved in the absence of strict controls because the users of the system will alter their behavior in a way that seems favorable to them but that is adverse to overall safety.