Presentation (CTI)

OBJECTIVES: Pneumonia is the most common fatal infection in the European Union, causing over 100,000 deaths annually, yet regional, seasonal and health-system determinants of in-hospital mortality remain unclear. This study aimed to identify determinants of in-hospital pneumonia mortality across Portuguese health regions using hierarchical mixed-effects modelling.
METHODS: A mixed-effects linear model was fitted to a panel dataset of five Portuguese health regions (Norte, Centro, Lisboa e Vale do Tejo, Alentejo, Algarve) from 2019-2025 (28 quarters per region totalizing 140 region-quarter observations). Data was obtained by a linkage of different national database sources. The outcome was in-hospital pneumonia mortality rate from the National Health System. Region was modelled as a random intercept to capture unobserved structural heterogeneity. Fixed-effect predictors spanned demographic (% population aged ≥65), seasonal (summer quarter), environmental (temperature range) and health-system domains (primary care consultation rate, emergency room waiting time, non-urgent emergency attendance rate). Continuous predictors were standardised.
RESULTS: The model converged successfully, with region explaining 21.8% of total variance in mortality. Summer quarters were associated with a 4.12 p.p. increase in mortality (β=4.12, 95%CI 2.32-5.92, p<0.001), and primary care consultation rate showed a strong inverse association with mortality (β=−4.56, 95%CI −6.32 to −2.80, p<0.001). Among the remaining predictors, population aged ≥65 (β=1.94, p=0.043) and non-urgent emergency attendance (β=0.82, p=0.033) were positively associated with mortality. Temperature range was inversely associated with mortality (β=−1.94, p<0.001) while emergency room waiting time showed a non-significant negative trend (β=−0.74, p=0.099).
CONCLUSIONS: In-hospital pneumonia mortality in Portugal is shaped by regional structural factors, demographic vulnerability, seasonality and health system characteristics. Geographic and seasonal heterogeneity points to opportunities for targeted public health action - in particular, pneumococcal and influenza vaccination programmes - especially in regions with ageing populations. Further research integrating clinical severity data and causal modelling is needed to inform effective policy.