December 1996

Influence of Multiple Organ Dysfunction Syndrome on Duration of Critical Illness and Hospitalization

Author Affiliations

From the Section of Acute Care Surgery, Department of Surgery, Cornell University Medical College and the Anne and Max A. Cohen Surgical Intensive Care Unit, The New York Hospital—Cornell Medical Center, New York, NY.

Arch Surg. 1996;131(12):1318-1324. doi:10.1001/archsurg.1996.01430240072010

Background:  Multiple organ dysfunction syndrome (MODS) is the leading cause of death in the general surgery intensive care unit (SICU). The development of MODS is a powerful predictor of prolonged SICU stay in survivors and nonsurvivors of critical illness, but its relation to less severe illness and briefer duration of care is unknown.

Objectives:  To determine the relation between modest degrees of MODS and length of stay in the SICU and hospital and whether daily MOD score calculations can distinguish survivors from nonsurvivors before the SICU stay becomes prolonged.

Setting:  An SICU of a university tertiary care medical center.

Design:  Prospective inception-cohort study. Illness severity data were collected in retrospect only for the calendar year 1991.

Patients:  Of 2646 consecutive patients studied, 115 stayed in the SICU more than 21 days.

Methods:  Acute Physiology and Chronic Health Evaluation (APACHE) II and III scores were calculated after 24 hours, with daily and cumulative MOD scores (0-4 points for 6 organs, 24 points maximum). Patients were followed up until hospital discharge or death. Data analysis was performed by unpaired 2-tailed t test, exact contingency analysis for multiple groups, univariate 1- or 2-way analysis of variance with repeated measures, or linear or polynomial regression tests as appropriate, α=.05.

Results:  The mean (±SEM) age of the patients was 65±1 years; mean (±SEM) APACHE II score, 13.8±0.2; APACHE III score, 44.2±0.7; incidence of MODS, 1173 of 2646 patients, 44.3%; and hospital mortality rate, 9.2%. Cumulative MOD scores correlated closely with SICU length of stay in survivors, especially for SICU stays of less than 10 days (R2=0.99, P<.001). Similar correlations existed between the prevalence of MODS related to the increasing length of the SICU stay (R2=0.98, P<.001) and between the length of hospital stay and the cumulative MOD score (R2=0.79, P<.05). Daily MOD scores in patients whose SICU stay was more than 21 days distinguished survivors from nonsurvivors by day 2 of the SICU stay (P<.05) and thereafter.

Conclusions:  Modest degrees of MODS correlate closely with the duration of care in less severely ill patients. Early identification and daily quantitation of MODS may help identify patients at risk for prolonged illness and death. Prevention of outcomes that contribute to organ dysfunction is critical for reduction of length of stay and cost of careArch Surg. 1996;131:1318-1324