The NCSS demonstrated several limitations. Although suspected cholera cases were retrospectively documented by the health ministry as early as October 14, 2010, official surveillance data covered the period beginning October 20, 2010, the date on which stool specimens showing toxigenic V. cholerae O1 were first collected. As such, the system was not set up to trace back to the origin of the epidemic.
Second, because the cholera surveillance system is largely facility-based, the true burden of cholera morbidity and mortality is likely to have been underestimated, particularly in remote areas with poor access to health facilities.
Third, although the number of community deaths provided a useful indicator for deaths that occurred outside health facilities at different times and places, the enumeration was not performed by trained medical personnel. Anecdotal evidence from Haiti and elsewhere suggests that community deaths may be underreported.
Fourth, although it was difficult to accurately measure the day-to-day completeness of reporting on a national scale, reports were regularly submitted by facilities in the majority of communes. However, completeness was sometimes sacrificed for timeliness, and when necessary, antecedent data were updated during the preparation of daily reports, making historical data dynamic in nature and difficult to analyze. Although reports were disseminated in a timely fashion early in the epidemic, as peak periods waned, timeliness dropped off, requiring periodic interventions to reinvigorate reporting.
Fifth, as with all surveillance systems established for epidemic cholera, the majority of cases in the NCSS were not laboratory-confirmed, and inevitably some cases of acute watery diarrhea caused by pathogens other than V. cholerae were misclassified as cholera. Surveillance data on the two age groups suggest that this misclassification was more apparent during epidemic lulls, when background rates of noncholera diarrheal disease among children under the age of 5 years probably represented a higher proportion of reported cases.
Using the stricter WHO case definition of acute watery diarrhea in persons 5 years of age or older, the health ministry reported 525,696 cases of infection, 295,303 hospitalizations, and 6856 deaths from cholera through October 20, 2012.So some suspicions are confirmed: cases and deaths are likely underreported, especially from rural areas far from health centres. "Reinvigorating" reporting from the departments is an ongoing problem (especially, in recent weeks, in Grande Anse). Casual updating of cases and deaths has resulted in difficulties in analyzing what's been going on. And despite more than two years of experience, this process doesn't seem to have improved much.
It's striking that NCSS tracked cholera into all ten departments by late November 2010, which means that cholera itself moved really fast—even into very remote communities that healthcare workers and NGOs found almost impossible to reach.
It's also striking that this is the first real overview of the response to cholera that I've seen. MSPP and its NGO partners may have done a reasonable job of surveillance, but the reports on that surveillance have been erratically published with virtually no explanation or analysis.
I would have thought a public information campaign to fight cholera would have benefited from such explanation. Instead, MSPP and the Haitian media continue to say very little about what the report describes as "57% of all cholera cases and 45% of all deaths from cholera reported to the WHO in 2010 and 2011."
Surely WHO and the NGOs must have urged the Haitian government to develop improved public information for both its own people and for other countries whose support could help stop the outbreak. And yet the government's efforts seemed to consist of little more than posters on the virtues of pooping in a proper latrine and remembering to wash your hands. The silence of governments, NGOs and media remains one of the most baffling aspects of cholera in Haiti.