Audit of intensive care unit admissions from the operating room.

An audit of 265 intensive care unit (ICU) admissions from the operating room was performed for the year 1991. In a quality assurance exercise we identified 34 unanticipated ICU admissions (UIAs) by a retrospective peer review of the medical charts. Of these UIAs, 16 were deemed predictable and seven preventable. Five of the seven potentially preventable UIAs were judged to have had inappropriate intravenous fluid management. This has prompted changes in our education programme. In an assessment of our resource management, we evaluated prospectively collected data on the Apache II scores on the day of admission, the incidence of ICU-specific interventions, length of stay in ICU, and outcomes. ICU-specific interventions were not initially required in 36% of admissions and these patients had a low risk (1.1%) of eventually requiring ICU-specific interventions. In comparison with patients requiring ICU-specific interventions, they had lower Apache II scores (10.2 vs 13.1), shorter ICU stays (medians of one vs two days), lower ICU mortality (0 vs 8.2%), P < 0.05, but hospital mortality was not different (7.4 vs 15.3%). This audit has prompted reorganisation of our intensive care services, so that patients not requiring ICU-specific interventions will be managed in an intermediate care area with nurse:patient ratios of 1:3 or 4, in comparison with 1:1 or 2 ratios in the intensive care area.