Prognostic Value of Clinical and Laboratory Markers in Hospitalized Influenza‐Positive Patients: A Retrospective Cohort Study With ROC‐Based Risk Stratification
Mohsen Daneshmand, Arefeh Ebrahimian Shiadeh, Farzane Sadeghi, Hemmat Gholinia, Farzin Sadeghi, Masomeh BayaniABSTRACT
Background and Aims
Seasonal influenza continues to cause substantial morbidity and mortality worldwide. Identifying reliable clinical and paraclinical predictors of adverse outcomes may aid in early triage and management. This study aimed to evaluate clinical and paraclinical factors associated with prolonged hospitalization, ICU admission, and in‐hospital mortality in influenza‑positive patients.
Methods
This retrospective study analyzed 294 RT‐PCR–confirmed influenza cases admitted to Ayatollah Rouhani Hospital (Babol, Iran) during fall–winter 2022–2023. Demographic, clinical, and laboratory data were extracted. Key outcomes were prolonged hospital stay (> 5 days), ICU admission, and in‐hospital death. Logistic regression and ROC analysis were used to identify independent predictors and cutoff values.
Results
Prolonged hospitalization occurred in 43.5% of patients, ICU admission in 8.8%, and mortality in 5.0%. In multivariate analysis, opium use (OR: 5.4), renal disease (OR: 9.27), and LDH > 1000 (OR: 8.74) were significantly associated with longer hospital stay. LDH > 1000 also predicted ICU admission (OR: 46.9) and mortality (OR: 21.7). Oxygen saturation < 93% increased mortality risk (OR: 7.98). ROC analysis identified LDH ≥ 633 and CRP ≥ 88.5 as optimal cutoffs for predicting mortality.
Conclusion
Readily available markers such as LDH, CRP, liver enzymes, and oxygen saturation, which showed significant associations with prolonged hospitalization, ICU admission, or mortality in our analyses, may assist in early risk stratification of hospitalized influenza patients.