Mortality reduction benefits and intussusception risks of rotavirus vaccination in 135 low-income and middle-income countries: a modelling analysis of current and alternative schedules

Abstract

Background: Infant rotavirus vaccines have led to substantial reductions in hospital admissions and deaths due to gastroenteritis, but some studies have reported an elevated risk of intussusception, a rare bowel disorder. This analysis aimed to provide evidence on the potential mortality reduction benefits and intussusception risks of current rotavirus vaccination schedules, and to explore whether alternative schedules could have advantages.

Methods: All 135 low-income and middle-income countries, defined by gross national income per capita of less than US$12 236 in the 2018 fiscal year, were included in the model. Mortality reduction benefits and intussusception risks of rotavirus vaccination were modelled by use of an Excel-based static cohort model with a finely disaggregated age structure. Numbers of rotavirus gastroenteritis deaths and intussusception deaths in each week of age were calculated for all infants born in the year 2015 between birth and age 5·0 years, with and without restrictions on age at administration. Benefit-risk ratios (rotavirus gastroenteritis deaths prevented per excess intussusception death) and other indicators were calculated for two vaccination schedules currently recommended by WHO and 16 alternative schedules. Of these schedules, it was assumed that between one and three doses would be given; the first dose of the rotavirus vaccine would be co-administered with either BCG or diphtheria-tetanus-pertussis (DTP)1; and the second or third dose would be co-administered with either DTP1, DTP2, DTP3, or measles (Meas)1.

Findings: A three-dose schedule co-administered with DTP (without age restrictions) could prevent about 74 000 (95% uncertainty interval 59 000-100 000) rotavirus gastroenteritis deaths (38% reduction) and could lead to 201 (77-550) excess intussusception deaths (1·4% increase) compared with no vaccination, resulting in a benefit-risk ratio of 369:1 (160:1-895:1). The benefit-risk ratio was most favourable when the relative risk of intussusception was assumed to decline with the national under-5 mortality rate (2386:1) and least favourable with pessimistic assumptions about access to hospital for intussusception treatment (168:1). Schedules that involve giving the first dose with BCG and the second with DTP1 had the fewest excess intussusception deaths and most favourable benefit-risk ratios.

Interpretation: Rotavirus vaccines have a favourable benefit-risk profile in LMICs. Neonatal schedules have the potential to prevent more rotavirus gastroenteritis deaths and cause fewer excess intussusception deaths than the schedules currently recommended by WHO, but more efficacious rotavirus vaccines would be needed to achieve more substantial mortality reduction benefits.

Funding: Bill & Melinda Gates Foundation.

Figure

RR of intussusception 1–7 days and 8–21 days after one and two doses of rotavirus vaccine in base-case risk scenario and alternative risk scenario The four panels on the top row represent the base-case risk scenario (relative risks [RRs] do not vary with under-5 mortality rate). The four panels on the bottom row represent an alternative risk scenario and assume RRs vary with the national under-5 mortality rate. Solid black lines indicate the median estimate and shaded grey areas represent 95% CIs. Circles and vertical lines represent the RR and 95% CIs reported for each individual self-controlled case-series data point. The size of the circle reflects the inverse of the variance (weight attributed to each study). See appendix (p 12) for more details about the fitting methods used. RR=relative risk.