The prevalence of influenza bacterial co-infection and its role in disease severity: A systematic review and meta-analysis.

Abstract

Evidence suggests that influenza bacterial co-infection is associated with severe diseases, but this association has not been systematically assessed. We aimed to assess the prevalence of influenza bacterial co-infection and its role in disease severity. We searched PubMed and Web of Science for studies published between 1 January 2010 and 31 December 2021. We performed a generalised linear mixed effects model to estimate the prevalence of bacterial co-infection in influenza patients, and the odds ratios (OR) of death, intensive care unit (ICU) admission, and requirement of mechanical ventilation (MV) for influenza bacterial co-infection compared to influenza single-infection. Using the estimates of OR and prevalence, we estimated the proportion of influenza deaths attributable to bacterial co-infection. We included 63 articles. The pooled prevalence of influenza bacterial co-infection was 20.3% (95% confidence interval (CI) = 16.0-25.4). Compared with influenza single-infection, bacterial co-infection increased the risk of death (OR = 2.55; 95% CI = 1.88-3.44), ICU admission (OR = 1.87; 95% CI = 1.04-3.38), and requirement for MV (OR = 1.78; 95% CI = 1.26-2.51). In the sensitivity analyses, we found broadly similar estimates between age groups, time periods, and health care settings. Likewise, while including studies with a low risk in confounding adjustment, the OR of death was 2.08 (95% CI = 1.44-3.00) for influenza bacterial co-infection. Based on these estimates, we found that approximately 23.8% (95% uncertainty range = 14.5-35.2) of influenza deaths were attributable to bacterial co-infection. We found that bacterial co-infection is associated with a higher risk of severe illnesses compared to influenza single-infection. Approximately one in four influenza deaths could be attributable to bacterial co-infection. These results should inform prevention, identification, and treatment for suspected bacterial co-infection in influenza patients. PROSPERO CRD42022314436.