[Assessment of intensive therapy: 4-year experience using the Apache II severity score]

To determine the outcome of critically ill patients and to study the characteristics of the population admitted to the intensive care unit.

Retrospective chart review.

Intensive Care Unit in a civic hospital.

The medical records of 1156 patients admitted to the intensive care unit over a period of 4 years were analysed according to the Apache II system on day one of admission. Demographic information, severity of illness, major interventions; mortality rate were examined.

Of 1156 patients 888 (76.8%) survived and 268 (23.2%) died. The mean Apache II score was 17 (SD 6), 15 (SD6) and 22 (SD7) for survivors and non-survivors respectively, the mean risk of death was 29 (SD 14), 25 (SD14) for survivors and 43 (SD 21) for nonsurvivors. mortality in emergency surgical admission was 28.3%, in booked elective surgical admission 3.8% and 14.3% in unbooked admission after elective surgery, against a predicted mortality of 34%, 16% and 27% respectively. No significant variations of the number of admissions were observed for each month of the year, day of the week and daily duty. Mortality ratio (actual death/predicted death) was 0.79. Apache II score showed in this series a specificity of 95.72% and a sensitivity of 25.37%.

These findings suggest that Apache II score is able to investigate critical ill patients, to stratify patients into disease groups and to quantify correctly the overall risk of death but cannot be used in decision making or to govern admissions in intensive care unit.