Decisions to forgo life-sustaining therapy in ICU patients independently predict hospital death.

Intensive Care Med. 2003 Nov;29(11):1895-901. Epub 2003 Oct 7

Azoulay E, Pochard F, Garrouste-Orgeas M, Moreau D, Montesino L, Adrie C, de Lassence A, Cohen Y, Timsit JF; Outcomerea Study Group

Saint-Louis Teaching Hospital, Medical ICU Department, 1 Avenue Claude Vellefaux, 75010, Paris, France. Cette adresse e-mail est protégée contre les robots spammeurs. Vous devez activer le JavaScript pour la visualiser.

OBJECTIVE: More than one-half the deaths of patients admitted to intensive care units (ICUs) occur after a decision to forgo life-sustaining therapy (DFLST). Although DFLSTs typically occur in patients with severe comorbidities and intractable acute medical disorders, other factors may influence the likelihood of DFLSTs. The objectives of this study were to describe the factors and mortality associated with DFLSTs and to evaluate the potential independent impact of DFLSTs on hospital mortality.

DESIGN AND SETTING: Prospective multicenter 2-year study in six ICUs in France. PATIENTS: The 1,698 patients admitted to the participating ICUs during the study period, including 295 (17.4%) with DFLSTs.

MEASUREMENTS AND RESULTS: The impact of DFLSTs on hospital mortality was evaluated using a model that incorporates changes in daily logistic organ dysfunction scores during the first ICU week. Univariate predictors of death included demographic factors (age, gender), comorbidities, reasons for ICU admission, severity scores at ICU admission, and DFLSTs. In a stepwise Cox model five variables independently predicted mortality: good chronic health status (hazard ratio, 0.479), SAPS II score higher than 39 (2.05), chronic liver disease (1.463), daily logistic organ dysfunction score (1.357 per point), and DFLSTs (1.887).

CONCLUSIONS: DFLSTs remain independently associated with death after adjusting on comorbidities and severity at ICU admission and within the first ICU week. This highlights the need for further clarifying the many determinants of DFLSTs and for routinely collecting DFLSTs in studies with survival as the outcome variable of interest.