Post mortem examination in the intensive care unit: still useful?

Objective: Post mortem examination rates have decreased worldwide and their usefulness has been challenged. The aim of this study was to compare ante- versus post mortem findings in a multidisciplinary ICU.

Design: Retrospective study.

Setting: Thirty-one-bed, medico-surgical ICU.

Patients: All patients who died on the ICU and underwent an autopsy examination in 1999.

Measurements: Records from autopsies were compared with clinical records. A modified Goldman's criteria was used to categorize the post mortem diagnoses. Unexpected findings were evaluated according to the duration of hospitalization prior to death (fewer than or more than 10 days).

Results: Among 2,984 ICU admissions, there were 489 deaths; 222 autopsies were conducted (45.4% autopsy rate). Post mortem examination revealed unexpected findings in 50 patients (22.5%), including malignancy (22 [9.9%]), fungal infections (9 [4%]), pulmonary embolism (7 [3.2%]), nosocomial infections (3 [1.3%]), Hashimoto's disease (3 [1.3%]), mesenteric infarction (2 [0.9%]), Barrett's esophagus (2 [0.9%]), endocarditis (1 [0.5%]) and myocardial infarction (1 [0.5%]). These unexpected findings were considered as major (Class I/II) in 19 (8.5%), and minor (Class III) in 31 (14%) patients. In patients with a short ICU length of stay (<10 days), there were more major unexpected findings than minor, while after a prolonged stay (>10 days), minor unexpected findings were more common.

Conclusions: After a short ICU stay (<10 days), autopsy revealed discrepancies primarily related to the cause of death associated with diseases whose diagnosis can be difficult. Following more prolonged ICU stays (>10 days), autopsy was more likely to reveal coexisting diseases unrelated to death.