Do objective estimates of chances for survival influence decisions to withhold or withdraw treatment? The French Multicentric Group of ICU Research.

The authors studied the impact on clinical decision making of providing feedback of objective prognostic information describing the probability of survival for ICU patients with multiple organ system failure (OSF). The prognostic estimates, derived from a control period (1), were to be provided on a daily basis to physicians providing treatment in 25 French ICUs during a subsequent experimental period (2). The types of, frequencies of, and reasons for decisions to limit or stop treatment in the two periods were compared. In the experimental period 2, 17 ICUs participated in the feedback study. Within these 17 units, there was a small but significant (p less than 0.05) increase in decisions to stop active treatment and provide comfort care that was limited to patients with three or more OSFs. There was no change in decision making in the eight units that did not participate in the feedback study. Although these results suggest a direct causal relationship between the provision of objective prognostic data and changes in physician decision making, the small increase in comfort care decisions (n = 14) between period 1 and period 2 and the fact that only 17 of the 25 original units participated in the feedback study make it difficult to eliminate other influences. There was no indication in this study, however, that explicit provision of prognostic data led to a sense of therapeutic futility.