Utilization of intensive care unit days in a Canadian medical-surgical intensive care unit

Objectives: To analyze the utilization of intensive care unit (ICU) days in a Canadian medical-surgical ICU and to identify ICU patients with prolonge d ICU length of stay (LOS). Design:Prospective descriptive study. Setting. A Canadian tertiary care medical-surgical ICU. Patients: Consecutive patients admitted to an adult medical-surgical ICU. N eurosurgical, cardiac surgical, and coronary care unit patients were exclud ed. Measurements: For each ICU admission, patient demographics, diagnosis, Acut e Physiology and Chronic Health Evaluation II (APACHE II) score, ICU LOS, a nd hospital mortality were collected. The patients' risk of death was calcu lated using the APACHE II equation. Admissions were stratified by ICU LOS i nto four groups: 1 to 2, 3 to 6, 7 to 13, and greater than or equal to 14 d ays. Among the four LOS groups, the number of ICU days and observed and pre dicted death rates were compared. Admissions were also stratified by risk o f death into five probability range quintiles. Among the five risk groups, ICU LOS was compared between survivors and nonsurvivors. Results: A total of 1,960 admissions utilized 9,298 ICU days. ICU LOS (mean +/- SEM) was 4.74 +/- 0.2 (median, 2; range, 1 to 178) days. Short-stay pa tients (ICU LOS less than or equal to 2 days) accounted for 60.3% of total admissions but consumed only 16.4% of total ICU days. Long-stay patients (I CU LOS greater than or equal to 14 days) accounted for 7.3% of total admiss ions but consumed 43.5% of total ICU days. Among the long-stay patients, th e most common reasons for admission were pneumonia, multiple trauma, neurom uscular weakness, and septic shock. The mortality for long-stay patients ap proached 50%. When analyzed by patients' mortality risks, those with a risk of death >0.8 (predicted to die) or <0.2 (predicted to live) whose outcome s were opposite to that predicted had twice the ICU LOS compared with patie nts whose outcomes were consistent with prediction. Conclusion: In a Canadian medical-surgical ICU, patients with ICU LOS great er than or equal to 14 days accounted for 7.3% of total admissions but cons umed 43.5% of total ICU days. Identification of patients with prolonged ICU LOS who would ultimately die in the ICU may lead to earlier withdrawal of therapy in these patients, resulting in a substantial reduction in sufferin g and cost savings. In our study population, outcome prediction using the A PACHE II equation did not provide sufficient power to accurately discrimina te between nonsurvivors and survivors.