Potential impact of rotavirus vaccine introduction in India's Universal Immunisation Programme on private sector vaccine utilisation: an interrupted time series analysis

BACKGROUND: Despite free immunisation services through the Universal Immunisation Programme (UIP), around 14% of Indian households seek immunisation in the private sector. We examined the potential impact of rotavirus vaccine (RVV) introduction in the Universal Immunisation Programme (UIP) on private-sector rotavirus vaccine utilisation. METHODS: We analysed nationally representative private-sector vaccine sales data. The intervention under consideration is RVV introduction in the UIP in selected Indian states. The outcome is the 'monthly RVV sales volume'-a proxy for vaccine utilisation. We performed a Poisson regression interrupted time series analysis to detect the pre-intervention trend, post-intervention level change and trend change relative to the pre-intervention for monthly rotavirus vaccine utilisation. RESULTS: Poisson segmented regression analysis showed that immediately after RVV introduction in the UIP private-sector RVV sales showed a decline in Rajasthan by 37.4% (Incidence Risk Ratio (IRR): 0.626; 95% CI: 0.504-0.779), in Tamil Nadu by 26% (IRR: 0.740; 95% CI: 0.513-1.068), in Uttar Pradesh-East by 72.2% (IRR: 0.278; 95% CI: 0.178-0.436) and in Kerala by 3% (IRR: 0.970; 95% CI: 0.651-1.447). Rajasthan, Tamil Nadu and Kerala had sustained reduction in the postintervention trend relative to the preintervention trend by 20.1% (IRR: 0.799; 95% CI: 0.763-0.836), 6.4% (IRR: 0.936; 95% CI: 0.906-0.967) and 3.3% (IRR: 0.967; 95% CI: 0.926-0.960) per month, respectively. However, in Haryana and UP-west, in the first-month post-UIP introduction, the private-sector RVV sales increased by 101% and 3.8%, respectively which was followed by a sustained decrease of 14.2% (IRR: 0.858; 95% CI: 0.688-1.070) and 5.8% (IRR: 0.942; 95% CI: 0.926-0.960) per month, respectively. In terms of long-term impact, the private sector RVV sales post-UIP introduction decreased at a monthly rate of 4.4% (IRR: 0.956, 95% CI: 0.939-0.974) in Rajasthan but increased by 5.5% (IRR: 1.055; 95% CI: 1.040-1.070) in UP-east, 0.3% (IRR: 1.003, 95% CI: 0.976-1.031)) in Kerala and 0.2% (IRR: 1.002, 95% CI: 0.993-1.011) in Tamil Nadu whereas Haryana and UP-west had a reduction in RVV utilisation by 2.8% (IRR: 0.972; 95% CI: 0.955-0.990) and 1% (IRR: 0.990; 95% CI: 0.982-0.998), respectively. CONCLUSIONS: The study provides evidence that access to RVV through UIP leads to a reduction in private-sector RVV utilisation. We recommend strengthening UIP to expand the basket of new vaccines.