Factors associated with inpatient mortality in a field hospital following the Haiti earthquake, January-May 2010.

To describe factors associated with inpatient mortality in a field hospital established following the 2010 Haiti earthquake. Data were abstracted from medical records of patients admitted to the University of Miami Global Institute/Project Medishare hospital. Decedents were compared to survivors in terms of age, sex, length of stay, admission ward, diagnosis, and where relevant, injury mechanism and surgical procedure. Three multivariate logistic regression models were constructed to determine predictors of death among all patients, injured patients, and noninjured patients. During the study period, 1,339 patients were admitted to the hospital with 100 inpatient deaths (7.5 percent). The highest proportion of deaths occurred among patients aged < or = 15 years. Among all patients, adult intensive care unit (ICU) admission (adjusted odds ratio [AOR] = 7.6 and 95% confidence interval [CI] = 3.4-16.8), neonatal ICU/pediatric ICU (NICU/PICU) admission (AOR = 7.8 and 95% CI = 2.7-22.9), and cardiac/respiratory diagnoses (AOR = 8.5 and 95% CI = 4.9-14.8) were significantly associated with death. Among injured patients, adult ICU admission (AOR = 7.4 and 95% CI = 1.7-33.3) and penetrating injury (AOR = 3.3 and 95% CI = 1.004-11.1) were significantly associated with death. Among noninjured patients, adult ICU admission (AOR = 6.6 and 95% CI = 2.7-16.4), NICU/PICU admission (AOR = 8.2 and 95% CI = 2.1-31.8), and cardiac/ respiratory diagnoses (AOR = 6.5 and 95% CI = 3.6-12.0) were significantly associated with death. Following earthquakes in resource-limited settings, survivors may require care in field hospitals for injuries or exacerbation of chronic medical conditions. Planning for sustained post-earthquake response should address these needs and include pediatric-specific preparation and long-term critical care requirements.