Outbreak characteristics and reproduction numbers of the 2017 measles outbreak in Guinea

Background: Measles is a highly contagious vaccine-preventable disease and a major cause of child mortality in developing countries. Elimination of measles is targeted worldwide, but low levels of vaccination coverage are impeding progress. Vaccine hesitancy and interruptions to routine vaccination in emergency situations create immunity gaps that increase the proportion of susceptible individuals in populations. Guinea is a low-income coastal country in tropical western Africa with a high birth rate that was recently overwhelmed by the Ebola virus epidemic in 2014–2015. During the outbreak, well residents were wary of seeking healthcare, including routine vaccination, and a large cohort of susceptible children entered the population. A large-scale measles outbreak began in 2017 and was ongoing as of June 2018. To target chains of transmission for interruption, the most affected populations should be characterized and context-specific reproduction numbers—the epidemiologic parameters describing the transmissibility of a disease—should be estimated to help determine the feasibility of local elimination efforts. Methods: Likelihood-based estimation was used to infer the temporal pattern of the effective reproduction number R from the observed epidemic curve for cases with illness onset in 2017. The basic reproduction number R0 was inferred from R based on international estimates of vaccination coverage in Guinea and vaccine efficacy of measles-containing vaccines. Results: There were 5,931 measles cases reported in 2017. The cases had illness onsets between January 2 and December 25, 2017, though one-tenth (11.1%) of cases were missing date of illness onset. Among cases for whom age was known, three-quarters (74.1%) were children under 5 years of age. Conakry was the most affected region in Guinea, with 2,142 cases (36.1% of all cases). The casefatality rate as 4.9 per 1,000 cases. All fatalities were among children under 15 years of age. Only 2.8% of cases were over 15 years of age. The proportion of cases with known vaccination history was 88.1% (n = 5,226), of whom only 13.5% were listed as being previously vaccinated. R was estimated at 1.02 (95% CI 0.39–2.03) for all cases with onsets between January 2 and December 13, 2017. The overall weekly smoothed estimate of R for epidemiologic week 1 was 3.01. The value of R0 necessary to generate the outbreak was 5.62 (3.17–8.50) based on international estimates of vaccination coverage or 6.69 (3.78–10.12) based on national estimates. Discussion: As the estimates for R in the first epi-week of 2017 were > 1, it can be assumed that R in Guinea had been > 1 when the outbreak began. R < 1 was not maintained later in the outbreak, even though vaccination control measures were implemented. The fluctuation of R later in the year reflects the fluctuations in the epidemic curve. Standardized methods for calculating reproduction numbers have not been established, however simple methods using existing tools and basic casecount data such as those used in this analysis can be leveraged for comparison of values across epidemics and provide context-specific estimates to inform measles elimination efforts.