Mortality estimates among adult patients with severe acute respiratory infections from two sentinel hospitals in southern Arizona, United States, 2010–2014

BACKGROUND From October 2010 through February 2016, Arizona conducted surveillance for severe acute respiratory infections (SARI) among adults hospitalized in the Arizona-Mexico border region. There are few accurate mortality estimates in SARI patients, particularly in adults ≥ 65 years old. The purpose of this study was to generate mortality estimates among SARI patients that include deaths occurring shortly after hospital discharge and identify risk factors for mortality. METHODS Patients admitted to two sentinel hospitals between 2010 and 2014 who met the SARI case definition were enrolled. Demographic data were used to link SARI patients to Arizona death certificates. Mortality within 30 days after the date of admission was calculated and risk factors were identified using logistic regression models. RESULTS Among 258 SARI patients, 47% were females, 51% were white, non-Hispanic and 39% were Hispanic. The median age was 63 years (range, 19 to 97 years) and 80% had one or more pre-existing health condition; 9% died in hospital. Mortality increased to 12% (30/258, 30% increase) when electronic vital records and a 30-day post-hospitalization time frame were used. Being age ≥ 65 years (OR = 4.0; 95% CI: 1.6-9.9) and having an intensive care unit admission (OR = 7.4; 95% CI: 3.0-17.9) were independently associated with mortality. CONCLUSION The use of electronic vital records increased SARI-associated mortality estimates by 30%. These findings may help guide prevention and treatment measures, particularly in high-risk persons in this highly fluid border population.