ObjectiveTo determine prognostic factors of virus-associated pneumonia other than coronavirus disease 2019. MethodsWe retrospectively studied patients suffering from virus-associated community-acquired pneumonia, and who were admitted to Saitama Cardiovascular and Respiratory Center from 2002 to 2020. Prognostic factors were analyzed by univariable and multivariable regression analysis of patient demographics, laboratory data, chest imaging, severity on admission, and initial treatment. PatientsHIV-positive patients, those with non-resected lung cancer or receiving chemotherapy, and those with COVID-19 were excluded. Included were 363 patients diagnosed by nucleic acid amplification method, paired sera, and rapid diagnostic tests. ResultsA CURB-65 score of ≥3 was significant by univariable analysis for 60-day mortality but was nonsignificant by multivariable analysis. The poor prognostic factors that were significant by multivariable analysis (p < 0.05) included immunosuppressive state due to systemic corticosteroid or immunosuppressant administration, acute kidney injury on admission, and corticosteroid administration initiated within 5 days or 5 days to 2 weeks from onset. ConclusionA CURB-65 score of ≥3, which is considered to indicate severe pneumonia, was of limited value for predicting mortality of virus-associated pneumonia. We showed patients’ underlying diseases and complications to be independent factors of poor prognosis for 60-day mortality. Timing of the initiation of corticosteroid administration remains to be elucidated.
Read full abstract