The global burden of influenza is increasingly recognized, but data from India remain sparse. We conducted a multi-site population-based surveillance study to estimate and compare rates of influenza-associated hospitalization at two rural Indian health and demographic surveillance system (HDSS) sites at Ballabgarh and Vadu during 2010–2012.

Prospective facility-based surveillance for all hospitalizations (excluding those for trauma, elective surgery and obstetric, ophthalmic or psychiatric reasons) was conducted at 72 health facilities. After collection of clinical details, patients had nasopharyngeal swabs taken and tested by reverse transcription polymerase chain reaction for influenza viruses. Annual healthcare utilization surveys (HUS) were conducted in HDSS households to identify proportion of hospitalizations occurring at non-study facilities to adjust for hospitalizations missed through facility-based surveillance.

HUS showed that 69% and 67% of hospitalizations occurred at study facilities at Ballabgarh and Vadu, respectively. Overall, 6004 patients hospitalized with acute medical illness at participating facilities were enrolled (1717 from Ballabgarh; 4287 from Vadu). The proportion of patients with influenza was higher at Vadu than Ballabgarh annually (2010: 21% vs. 5%, p < 0.05; 2011: 18% vs. 5%, p < 0.05; 2012: 23% vs. 5%, p < 0.05). Annual adjusted influenza-associated hospitalization rates were 5–11 fold higher in Vadu (20.3–51.6 per 10,000) vs Ballabgarh (4.4–6.3 per 10,000). At both sites, influenza A/H1N1pdm09 and B predominated during 2010, A/H3N2 and B during 2011, and A/H1N1pdm09 and B during 2012.

The markedly different influenza hospitalization rates by season and across communities in India highlight the need for sustained multi-site surveillance system for estimating national influenza disease burden. That would be the first step for initiating discussions around Influenza prevention and control strategies in the country.