Factors Associated with In-Hospital Mortality and Transfer in Pneumonia Patients: A Competing Risks Regression Analysis at Bangkhla Hospital

Abstract

Background: Pneumonia remains a major cause of hospitalization and death, particularly among older adults and patients with comorbidities. However, evidence on predictors of inhospital mortality and inter-hospital transfer in community hospitals remains limited. This study aimed to identify factors associated with in-hospital mortality and referral/transfer among hospitalized adults with pneumonia at a community hospital in Thailand using FineGray competing-risks regression. Materials and Methods: This retrospective cohort study included adults aged ≥18 years hospitalized with pneumonia at Bangkhla Hospital, Thailand, between 2015 and 2025. Data were extracted from electronic medical records and ICD diagnostic codes. Variables included demographic characteristics, body mass index (BMI), Charlson comorbidity group, and admission body temperature. The primary outcome was inhospital death, and the secondary outcome was referral/transfer to another hospital. Fine - Gray competing-risks regression was used to estimate subdistribution hazard ratios (SHRs). Discharge alive and referral/transfer were treated as competing events for the mortality model, whereas discharge alive and in-hospital death were treated as competing events for the referral/transfer model. Statistical significance was defined as p <0.05. Results: A total of 1,816 patients were included; 1,815 had complete discharge outcome data. The mean age was 50.53 ± 18.42 years, and 430 patients (23.68%) were aged ≥65 years. Most patients were discharged alive (n = 1,712; 94.33%), whereas 30 patients (1.65%) died in hospital, and 73 (4.02%) were referred/transferred. In the adjusted mortality model, age ≥65 years (adjusted SHR 2.59, 95% CI 1.06-6.34; p = 0.038), BMI <18.5 kg/m^² (adjusted SHR 4.72, 95% CI 1.41- 15.82; p = 0.012), and ≥2 comorbidities (adjusted SHR 6.20, 95% CI 2.37-16.25; p <0.001) were significantly associated with in-hospital death. Admission temperature <36.7°C or >37.8°C was not significantly associated with mortality after adjustment. For referral/transfer, age ≥65 years was the only significant factor (adjusted SHR 2.64, 95% CI 1.53 -4.56; p <0.001). Conclusion: In this community-hospital cohort, older age, underweight status, and multiple comorbidities were associated with in-hospital mortality, while older age was associated with referral/transfer. Early risk assessment may support monitoring, treatment planning, and timely referral in community hospital settings.