Effect of Patient Sex on Intensive Care Unit Survival
Total Page:16
File Type:pdf, Size:1020Kb
ORIGINAL INVESTIGATION Effect of Patient Sex on Intensive Care Unit Survival Hugo Romo, MD; Andre´ Carlos Kajdacsy-Balla Amaral, MD; Jean-Louis Vincent, MD, PhD Background: Human observations have shown differ- female patients older than 50 years (OR, 1.33; 95% CI, ent mortality rates between men and women with vari- 1.12-1.58) but not in the younger age group. The sub- ous pathological conditions, but this issue has not been group of medical admissions had a higher mortality widely studied in a heterogeneous population of criti- (24.4% vs 7.4%, PϽ.001) and a higher female propor- cally ill patients. tion (37.9% vs 34.2%, P=.01) than surgical admissions. In multivariate analysis, female sex remained an impor- Methods: Retrospective analysis of all patients admit- tant predictor of mortality (OR, 1.54; 95% CI, 1.25-1.89). ted to a mixed medical-surgical, 31-bed intensive care Women had a higher mortality than men in the sub- unit (ICU) during 2 different years (1983 and 1995) to group of cardiovascular diseases. The highest mortality evaluate possible differences in mortality between male in female patients was present in the first days after ad- and female patients and between medical and surgical ad- mission and decreased over time, showing a covariance missions and variations in these differences over time. of time and sex. Results: From a total of 4420 admissions (1587 women, Conclusions: In a mixed medical-surgical ICU, older 2833 men), women showed a higher mortality, with an women have a higher mortality rate than men. This dif- odds ratio (OR) of 1.18 (95% confidence interval [CI], ference is not apparent for patients staying longer in the 1.02-1.38). This pattern was the same for the 2 periods, ICU. and all patient data were therefore analyzed together. Af- ter age stratification, the differences were significant for Arch Intern Med. 2004;164:61-65 ECENTLY, THERE HAS BEEN sex hormones, with women and men increased interest in the treated with estrogens developing a more possible influences of sex marked proinflammatory response22-24 and on disease development having improved cardiac and hepatic func- and intensive care unit tion compared with men.25 These changes (ICU) survival. Women have a lower in- are influenced by the female reproduc- R 26 cidence of coronary heart disease (CHD), tive cycle, and the cytokine response is an effect that is blunted with aging.1 Af- reversed when older animals with de- ter an acute myocardial infarction (AMI), creased hormonal levels are compared.27 women have a higher rate of early mor- Indeed, higher male mortality rates have tality,2-5 although long-term mortality af- been observed,28,29 and treatment of sep- ter AMI is not different between the sexes.5 tic male animals with dehydroepiandros- Higher mortality rates in men have been terone resulted in improved outcome.30 observed in some situations, including Whether these experimental findings can congestive heart failure6,7 and trauma,8,9 but be translated to the clinical arena, how- in other areas, including sepsis,10-16 me- ever, remains unclear. The objective of this chanical ventilation,17,18 and burns,19-21 the study was to observe if there were any dif- literature is conflicting. ferences in mortality rates and length of The reasons behind possible sex dif- ICU stay between men and women in a ferences have been explored in several ex- mixed medical-surgical ICU population. From the Department of Intensive Care, Erasme perimental models, mainly focusing on the Hospital, Free University of possible immunological effects of sex hor- METHODS Brussels, Brussels, Belgium. mones. Numerous studies have shown that The authors have no relevant the immune response, including that as- Records for all consecutive admissions to the financial interest in this article. sociated with sepsis, may be modulated by ICU of Erasme University Hospital, Brussels, (REPRINTED) ARCH INTERN MED/ VOL 164, JAN 12, 2004 WWW.ARCHINTERNMED.COM 61 ©2004 American Medical Association. All rights reserved. Downloaded From: https://jamanetwork.com/ on 09/24/2021 Table 1. Demographic Data of 4420 Patients Admitted to the Intensive Care Unit During 2 Study Periods 1983 1995 Variable Women Men Women Men P Value Patients, % 33.5 66.5 38.0 62.0 .004 Age, mean ± SD, y 54.3 ± 20.2 54.5 ± 16.6 56.5 ± 19.8 56.7 ± 18.6 .006* Patients older than 65 y, % 33.9 27.7 42.7 42.4 Ͻ.001† Length of stay, mean ± SD, d 4.1 ± 5.3 3.9 ± 5.5 3.5 ± 7.1 3.7 ± 6.1 .20* Surgical patients, % 58.6 67.4 64.6 64.5 .007† Survivors, % 85.6 87.6 84.7 87.2 .34† *General linear model. †Mantel-Haenszel test. Table 2. Differences Between Medical and Surgical Patients Medical Surgical Variable Women Men Women Men P Value Patients, No. (%) 601 (37.9) 968 (34.2) 986 (62.1) 1865 (65.8) .01 Age, mean ± SD, y 56.5 ± 20.9 55.3 ± 19.8 55 ± 19.4 55.9 ± 16.6 .57* Length of stay, mean ± SD, d 4.0 ± 6.8 4.2 ± 5.4 3.5 ± 6.2 3.6 ± 6.0 .03* Mortality, % 24.1 24.6 9.3 6.4 Ͻ.001† *General linear model. †Mantel-Haenszel test. Belgium, during 2 different years, 1983 and 1995, were From the total number of patients, 595 (13.5%) retrieved by searching the unit databases. These years were died during their ICU stay. Survivors were younger than chosen because they were readily available from our databases nonsurvivors (55±18 years vs 61±18 years, P=.006), and allowed a comparison of mortality rates over time. Erasme and they had a significantly shorter length of ICU stay University Hospital is a 900-bed, academic, adult hospital with (3.5±5.7 days vs 5.4±7.6 days, PϽ.001). Mortality a multidisciplinary, mixed medical and surgical, 31-bed inten- sive care department. Patients with acute, uncomplicated coro- increased with age from 9.9% in patients younger than nary disease are admitted to a separate coronary care unit. 45 years, 11.8% in patients aged 45 to 65 years, and 2 Patients admitted several times during these years were 18.1% in patients older than 65 years ( trend=40.5, included as new admissions. Demographic data, including sex, PϽ.001). Women had a higher mortality rate than men admission diagnosis, medical or surgical status, age, length of (15% vs 13%; odds ratio [OR], 1.18; PϽ.05). When the ICU stay, and ICU outcome, were collected. Univariate and sample was stratified according to age, women older multivariate analyses were performed to determine predictors than 50 years showed the highest risk of mortality (OR, of mortality in this population using statistical software (SPSS 1.33; PϽ.001). The crude OR and the Mantel-Haenszel version 10.0.5; SPSS Inc, Chicago, Ill). Categorical variables adjusted OR were not significantly different, showing no were assessed with the 2 test. The 2-tailed t test and Mann- Whitney U test were used as appropriate for continuous vari- evidence of confounding. However, the homogeneity ables. Stratified analysis was performed using the Mantel- test across strata confirmed the presence of a significant Haenszel test. General linear model and time-dependent Cox interaction between sex and mortality when age was regression model were also applied. The actuarial method was controlled (PϽ.001). used to assess cumulative differences in survival. PϽ.05 was Medical patients represented 35.9% of the popula- considered statistically significant unless a Bonferroni correc- tion; differences between the medical and surgical tion was applied to evaluate multiple comparisons. Data are groups are depicted in Table 2. There were no differ- presented as mean±SD, excepted where stated otherwise. ences in age between the medical and surgical groups as a whole or when stratified according to diagnosis or sex. RESULTS There were more female than male medical patients (37.9% vs 34.2%, P=.01). Surgical patients had a shorter Demographic data for the 2 years are summarized in mean length of stay than medical patients (3.6±6.1 days Table 1. There were 4420 admissions (1587 women, vs 4.1±5.9 days, P=.04). The mortality rate was higher 2833 men), 1950 admitted in 1983 and 2470 in 1995. in medical vs surgical patients (24.4% vs 7.4%, PϽ.001). The proportion of female patients increased throughout Cardiovascular problems were the most frequent source the 12-year period. There were no significant differ- of admission to the ICU for the medical group, account- ences in the mean ICU length of stay when both years ing for 29% of admissions. The mortality rate of female were compared. Since mortality rates were similar in the patients with cardiovascular disease was higher than that 2 different periods, data from the 2 years were analyzed of their male counterparts (23% vs 12%; OR, 2.07; 95% collectively. CI, 1.50-2.87; PϽ.001) (Table 3). (REPRINTED) ARCH INTERN MED/ VOL 164, JAN 12, 2004 WWW.ARCHINTERNMED.COM 62 ©2004 American Medical Association. All rights reserved. Downloaded From: https://jamanetwork.com/ on 09/24/2021 Table 3. Admission Diagnoses With Corresponding Mortality Rates Mortality Rate, No. (%) Cause of Admission Patients, No. Admissions, % All Men Women Surgical Neurosurgery 891 20.2 158 (17.7) 90 (19.3) 68 (16.0) Cardiac surgery 323 7.3 10 (3.1) 8 (4.4) 2 (1.4) Thoracic surgery 163 3.7 6 (3.7) 5 (4.4) 1 (2.0) Vascular surgery 94 2.1 3 (3.2) 1 (1.4) 2 (8.0) Abdominal surgery 87 2.0 13 (14.9) 4 (8.9) 9 (21.4) Urology 15 0.3 0 0 0 Polytrauma 299 6.8 44 (14.7) 24 (12.1) 20 (20.0) Gynecology-obstetrics 21 0.5 0 0 Medical Cardiovascular 1284 29.0 191 (14.9) 118 (12.1) 73 (22.5)* Respiratory 510 11.5 72 (14.1) 49 (14.3) 23 (13.8) Gastrointestinal 309 7.0 47 (15.2) 28 (13.2) 19 (19.6) Neurology 165 3.7 33 (20.0) 22 (24.4) 11 (14.7) Nephrology 86 1.9 5 (5.8) 4 (7.7) 1 (2.9) Intoxications 64 1.4 5 (7.8) 3 (10.0) 2 (2.9) Internal medicine 48 1.1 5 (10.4) 2 (8.0) 3 (13.0) Psychiatry 19 0.4 1 (5.3) 0 1 (7.7) Endocrinology 16 0.4 0 0 0 Oncology 10 0.2 0 0 0 Others 16 0.4 2 (12.5) 0 2 (28.6) *PϽ.001.