Date Published: May 05, 2018
Publisher: The American Society of Tropical Medicine and Hygiene
Author(s): Andreas Rieckmann, Charlotte C. Tamason, Emily S. Gurley, Naja Hulvej Rod, Peter Kjær Mackie Jensen.
Cholera outbreaks in Africa have been attributed to both droughts and floods, but whether the risk of a cholera outbreak is elevated during droughts is unknown. We estimated the risk of cholera outbreaks during droughts and floods compared with drought- and flood-free periods in 40 sub-Saharan African countries during 1990–2010 based on data from Emergency Events Database: the Office of Foreign Disaster Assistance /Centre for Research on the Epidemiology of Disasters International Disaster Database (www.emdat.be). A cholera outbreak was registered in one of every three droughts and one of every 15 floods. We observed an increased incidence rate of cholera outbreaks during drought periods (incidence rate ratio [IRR] = 4.3, 95% confidence interval [CI] = 2.9–7.2) and during flood periods (IRR = 144, 95% CI = 101–208) when compared with drought/flood-free periods. Floods are more strongly associated with cholera outbreaks, yet the prevalence of cholera outbreaks is higher during droughts because of droughts’ long durations. The results suggest that droughts in addition to floods call for increased cholera preparedness.
Half of the world’s reported cholera cases occur in sub-Saharan Africa, which is also the region with the highest case-fatality rates of cholera.1 Single-country studies and one review of cholera outbreaks in Africa have linked cholera outbreaks to both droughts and floods.2–4 Whereas floods are a recognized risk factor of cholera transmission by the World Health Organization (WHO), droughts are not explicitly regarded as a risk factor in guidelines.5,6 If the risk of a cholera outbreak is increased during droughts in addition to floods in Africa, established cholera preparedness procedures should apply to droughts and floods.
This is a register-based country-level ecological study from 1990 to 2010.
Two hundred and forty-eight outbreaks were registered as cholera outbreaks. Thirty-nine outbreaks were registered as “acute diarrheal syndrome” and “acute watery diarrhea,” of which 28 were confirmed via location and time period by reputable sources (see Supplemental Table 1) to be cholera outbreaks; thus, a total of 276 cholera outbreaks were registered during the study period. Four of 276 of the cholera outbreaks were missing start dates. We were able to confirm the start month of four of the outbreaks through online epidemiological records (WHO and Reliefweb) and the remaining three were excluded. We imputed the missing start month of the 15% of registered droughts (8/118) and the missing end month for 72% of droughts (85/118). The mean length for droughts with no missing versus all droughts including imputed values was 9 and 14 months, respectively. No floods were missing data for the start month. With regard to the number of affected individuals from a drought or flood, we imputed the median value being 750,000 people for 14 (11%) of the droughts and the median value being 10,000 people for 33 (6%) of floods.
Floods were—as expected—associated with an extraordinarily high incidence rate of cholera outbreaks during floods compared with flood/drought-free periods (IRR = 144 [95% CI = 101–208]). However, droughts too were associated with a higher IRR of 4.3 (95% CI = 2.9–7.2) compared with flood/drought-free periods. As droughts naturally last longer than floods and affect a larger area, a higher proportion of droughts than floods may be expected to overlap with cholera outbreaks; we found that a cholera outbreak began during 30% of all registered droughts; this is more than four times higher than the proportion of cholera outbreaks that began during registered floods. This understanding that cholera outbreaks are often seen during droughts could potentially improve preparedness during droughts and prevent cholera deaths; however, it is also important to emphasize that the vast majority of cholera outbreaks (217/276) began during drought/flood-free periods, whereas 35 began during droughts and 34 began during floods.
Our analysis suggests that countries in sub-Saharan Africa can expect to see a cholera outbreak in one out of every three droughts. This is more than four times as often as they would occur during floods. Our findings indicate that droughts in addition to floods should be considered as periods of increased risk for cholera outbreaks. We recommend that cholera control guidelines from local to international levels recognize drought periods as higher risk times, warranting increased vigilance to prevent and control cholera outbreaks, as well as floods. Research on how a lack of water, hygiene, and crowding may affect cholera transmission during droughts and floods is needed to better understand the drivers that are perpetuating this deadly disease.