Aggregate risk measures

Section 4.2.2 discussed the use of methods other than F-N curves for determining aggregate risk. An aggregate risk measure can be calculated for Example 10 by multiplying each incident frequency by the expected number of serious injuries or fatalities. This results in a parameter of "fatalities per year."... 

This paper presents a framework to evaluate risk in multiple dimensions and provides an aggregated risk measure using one-dimensional utilities. 

Aggregate risk is used to measure the collective risk to people in a facility who could be exposed to an event or events. It indicates the frequency that a specified number of people will suffer a specific level of harm (e.g., death). Just as is the case with societal risk, aggregate risk can be expressed in terms of the frequency distribution of multiple-casualty events. 

Aggregate risk A measure of the risk to personnel within a building or buildings or within a facility who are impacted by the same events, taking into account the total time spent in the building(s) or facility. (Aggregate risk is "societal risk" applied to a specific group of people within a facility.)... 

Bi represents basic initiating events in the fault tree analysis, leading to an undesired event, and Cy represents different possible end events resulting from the event tree analysis. C-z are the aggregation of the consequences of all end events into a common risk measure. 

I 10 Tantry US, Bliden KR Gurbel PA. What is the best measure of thrombotic risks—pretreatment platelet aggregation, clopidogrel responsiveness, or posttreatment platelet aggregation Catheter Cardiovasc Interv 2005 66 597. 

Post-MI management calls for strict adherence to a program of secondary prevention. Cardiac risk factors have to be excluded or modified, for instance, by reduction of overweight, cessation of smoking, optimal control of diabetes mellitus, and physical exercise (a dog that loves to run is an ideal training partner). Supportive pharmacother-apeutic measures include administration of platelet aggregation inhibitors, p-blockers, and ACE inhibitors. 

It is more sensible either to classify trials by some measure of the baseline risk (that is to say, in terms of measures taken before treatment) of all patients or, even better, to classify patients individually in this way. Note that even if a statistically sound way could be found of estimating the treatment effect in a series of trials as a function of the response in the placebo group, this information would be of no direct use to prescribing doctors. Not only is such information related to a trial aggregate, while what is important is individuals, but to make use of it inorder to prescribe effectively a prescribing physician would have to know the response a patient would have under placebo were he or she to be entered into a trial. There is no such problem if the relationship between the treatment effect and an individual baseline measurement can be established. 

---