Predictors of Length of Stay, Rehospitalization and Mortality in Community-Acquired Pneumonia Patients: A Retrospective Cohort Study

Abstract

Background: Community-acquired pneumonia (CAP) represents one of the leading causes of hospitalization and has a substantial impact on the financial burden of healthcare. The aim of this study was to identify factors associated with the length of hospital stay (LOHS), rehospitalization and mortality of patients admitted for CAP. Methods: A retrospective cohort study was conducted with patients presenting to a Swiss public hospital between January 2019 and December 2019. Zero-truncated negative binomial and multivariable logistic regression analyses were performed to assess risk factors. Results: A total of 300 patients were analyzed (median 78 years, IQR [67.56, 85.50] and 53% males) with an average LOHS of 7 days (IQR [5.00, 9.00]). Of the 300 patients, 31.6% (97/300) were re-hospitalized within 6 months, 2.7% (8/300) died within 30 days and 11.7% (35/300) died within 1 year. The results showed that sex (IRR = 0.877, 95% CI = 0.776–0.992, p-value = 0.036), age (IRR = 1.007, 95% CI = 1.002–1.012, p-value = 0.003), qSOFA score (IRR = 1.143, 95% CI = 1.049–1.246, p-value = 0.002) and atypical pneumonia (IRR = 1.357, 95% CI = 1.012–1.819, p-value = 0.04) were predictive of LOHS. Diabetes (OR = 2.149, 95% CI = 1.104–4.172, p-value = 0.024), a higher qSOFA score (OR = 1.958, 95% CI = 1.295–3.002, p-value = 0.002) and rehabilitation after discharge (OR = 2.222, 95% CI = 1.017–4.855, p-value = 0.044) were associated with a higher chance of being re-hospitalized within 6 months, whereas mortality within 30 days and within one year were both associated with older age (OR = 1.248, 95% CI = 1.056–1.562, p-value = 0.026 and OR = 1.073, 95% CI = 1.025–1.132, p-value = 0.005, respectively) and the presence of a cancer diagnosis (OR = 32.671, 95% CI = 4.787–369.1, p-value = 0.001 and OR = 4.408, 95% CI = 1.680–11.43, p-value = 0.002, respectively). Conclusion: This study identified routinely available predictors for LOHS, rehospitalization and mortality in patients with CAP, which may further advance our understanding of CAP and thereby improve patient management, discharge planning and hospital costs.

Keywords

• community-acquired pneumonia
• length of hospital stay
• rehospitalization
• mortality
• prediction
• CAP