Background: Understanding factors affecting length of hospital stay (LOS) in patients with severe influenza may improve their management.
Methods: A retrospective cohort study on laboratory-confirmed, adult influenza patients hospitalized in 2004 and 2005 was conducted. For all influenza cases during that period, immunofluorescence assay on nasopharyngeal aspirate was used for rapid diagnosis, and oseltamivir (75 mg twice daily for 5 days) prescribed if the patient presented within 2 days of symptom onset. Independent factors associated with time to discharge were identified using Cox proportional hazards models. An adjusted hazard ratio (aHR) >1 signifies a higher chance of early discharge. Viral shedding and influenza vaccination history were studied during one 'flu' season.
Results: A total of 356 patients (influenza A 93.5%) were studied. The majority of patients were old (70.2 +/- 8.4 years), had > or = 1 comorbid illness (69.1%) and developed respiratory or cardiovascular complications (69.4%). Oseltamivir initiated within 2 days of illness was associated with shorter total LOS (Kaplan-Meier estimated median 4 versus 6 days [-33%]; aHR for discharge 1.54, 95% confidence intervals [95% CI] 1.23-1.92, P < 0.0001). Older age (> or = 70 years), comorbidities and complications were associated with prolonged LOS. Prolonged viral RNA detection >day 4 of illness (23 out of 99 consecutive patients) was also independently associated with longer LOS (aHR 0.36 [95% CI 0.19-0.71], P = 0.003), whereas influenza vaccination within 6 months was associated with shorter LOS (aHR 2.14 [95% CI 1.18-3.85], P = 0.012).
Conclusion: Our analyses suggest that timely oseltamivir treatment is independently associated with shorter LOS in patients hospitalized for severe influenza. Efforts to ensure early diagnosis and therapeutic intervention are warranted.