Decisions for Critically Ill Patients

Standard analytical tools employed for quality assessment such as root cause analysis are simply inadequate. Probabilistic risk assessment (PRA) is a systematic and comprehensive methodology widely applied in other industries to evaluate undesirable outcomes such as risk, and PRA may be a promising tool for assessment of adverse events in the ICU.2–5 PRA entails constructing the chronology of a scenario and a “fault tree” of occurrences that could lead to a specific adverse outcome. By analysis of the likelihood of each contributor to the adverse event, one could quantify the likelihood of that outcome occurring and focus efforts on modifying each of these events. Results from PRA would also provide a quantitative basis for specific event rates and for overall system performance over time. The need for serious discussions of ICU performance metrics is overdue. We gauge the “health” of financial or business entities by measurements like price-to-earnings ratio, equity, debt load, or cash on hand. Similarly, we monitor the integrity of engineering systems by their failure rates, reliability, recovery rate, and level of redundancy. The performance measures described by Dr. Garland are good first steps toward arriving at community consensus; others may propose alternatives. The ultimate test would depend on the usefulness of these measurements in improving care in the ICU.