Vaccination against cholera in Juba

Abstract

In an interesting Personal View, Lucy Parker and colleagues1 reported the difficulties regarding implementation of a reactive oral cholera vaccination (OCV) campaign during the 2015 cholera epidemic in Juba, South Sudan.1 They support the choice to address the global shortage of vaccines by providing just one dose to twice the number of people. However, the epidemic curve provided by Parker and colleagues suggests that the South Sudan epidemic was not hugely affected by this campaign. Indeed, the basic reproductive number (R0), which we calculated as previously described2 using data extracted from this curve with the Plot Digitizer tool and R software, using the R0 package, was not reduced after the campaign was finally launched on July 31, 2015; the R0 was already less than 1 between the first peak on June 28 and the start of the OCV campaign (0·94 [95% CI 0·92–0·95]), only 0·72 (0·66–0·78) between the second peak on July 19 and the start of OCV, and still 0·92 (0·90–0·94) from the start of OCV until the last confirmed case on Sept 12, 2015. Several complementary factors might explain such a disappointing effect. First, vaccine effectiveness of this one-dose campaign could have been lower than the 87·3% (95% CI 70·2 – 100) calculated in a case-cohort observational study by the same group of authors.3 Efficacy of one-dose OCV was estimated to be about 40% (95% CI 11–60) in a double-blind placebo-controlled clinical trial.4 Using the WHO screening method5 with provided data, we calculated that 36% of cholera cases were expected to occur in vaccinated individuals in Juba. The observed proportion was only 6%,3 which suggests biases that the authors could not address despite their efforts to do so. Second, one-dose OCV did not generate any obvious herd immunity, even in the area targeted by mass vaccination, where coverage reached 64%;3 surprisingly, vaccine effectiveness tended to be much higher there (97%) than in the non-mass-vaccinated area (66% with 19% coverage),3 and the calculated cholera attack rate among non-vaccinees was two times higher than in the non-mass-vaccinated area (2·5 vs 1·3 cases per 10 000 inhabitants). 3 Finally, this late campaign probably provided little additional protection to a population in which adaptations to water sanitation and hygiene (WaSH) behaviour—rather than acquired immunity—were probably already reducing cholera transmission. This insightful cholera vaccination field report shows that WaSH activities must remain the cornerstone of cholera control and elimination strategies, even if they are difficult to implement. Reactive vaccination campaigns might help, provided they are promptly rolled out and include two doses as originally recommended.