Background The Infectious Disease Society of America/the American Thoracic Society (IDSA/ATS) minor criteria for severe community-acquired pneumonia (CAP) are of unequal weight in predicting mortality. It is unclear whether the patients with non-severe CAP meeting the minor criteria most strongly associated to mortality should have the priority for treatment and intensive care. It is warranted to explore this intriguing hypothesis. Methods A retrospective cohort study of 1230 patients with CAP was performed. This was tested against a prospective 2-center cohort of 1749 adults with CAP. Results The patients with CAP fulfilling the predictive findings most strongly associated to mortality, i.e. PaO2/FiO2 ≤ 250 mm Hg, confusion, and uremia, showed higher mortality rates than those not fulfilling the predictive findings in subgroup analyses of the retrospective cohort. The more the number of predictive findings present, the higher the mortality rates. The prospective cohort confirmed a similar pattern. Interestingly, the patients with non-severe CAP meeting the predictive findings demonstrated unexpectedly higher mortality rates compared with the patients with severe CAP not meeting the predictive findings in the prospective cohort (P = 0.003), although there only existed death of an uptrend in the retrospective cohort. Two similar and intriguing paradigms about sequential organ failure assessment (SOFA) scores and pneumonia severity index (PSI) scores were confirmed in the 2 cohorts. Conclusions The patients with non-severe CAP fulfilling the predictive findings most strongly associated to mortality demonstrated higher SOFA and PSI scores and mortality rates, and might have the priority for treatment and intensive care.
