Aim Severe pneumonia management in the hospital setting often relies heavily on established clinical practice and physician experience. This approach has the purpose of enabling early identification of risk factors most strongly associated with severe pneumonia at the time of hospital admission. Methods This retrospective study analyzed inpatients with pneumonia treated in a Respiratory disease unit, stratifying them into two groups—severe and non-severe pneumonia — according to the 2007 IDSA/ATS criteria, identifying differences in demographic profiles, clinical features, treatment strategies, and prognostic outcomes. Results Out of a cohort of 302 patients, 26 (8.6%) met the criteria for severe pneumonia. A statistically significant difference was observed in the Pneumonia Severity Index (PSI 90), recorded in 61.53% of patients with severe pneumonia compared to 41.31% in non-severe cases. The Charlson Comorbidity Index (CCI ≥ 4), indicative of lower 10-year survival due to comorbidities, was significantly more frequent in the severe group (84.61% vs. 61.23%). Microbiological analysis of bronchoalveolar lavage (BAL) showed a positivity rate of 75% in the severe group versus 35.48% in the non-severe group ( p 0.05). Significant differences were also found in the use of respiratory support: high-flow nasal cannula (HFNC) was used in 69.23% of severe cases versus 32.97% in non-severe cases, while CPAP was administered in 23.07% versus 5.43%, respectively. Finally, antibiotic therapy was significantly longer in the severe group, with a mean duration of 14.69 days compared to 11.77 days in non-severe patients, involving both intravenous and oral regimens as part of initial or sequential treatment. Conclusion Timely recognition of these factors is essential to ensure optimal patient care, facilitate close monitoring of critically ill individuals, allow for prompt therapeutic escalation, and support ICU admission when needed. This analysis highlights the need for a critical reassessment of existing guidelines and underscores the value of integrating them with real-world clinical experience.
Pisanu et al. (Thu,) studied this question.