Via his excellent blog Random Analytics, Shane Granger looks at incongruities in the numbers for Ebola in Sierra Leone (to 14 Aug 2014) as tallied by WHO and the Sierra Leone government Excerpt:
Although there is certainly issues around ‘boots on the ground’ capacity (see Jennifer Yang’s excellent article from Sierra Leone) the answer to the question as to why there was a significant difference between the Sierra Leone MoH and the WHO numbers turned out to be blindingly obvious.
Most of the reports coming out directly from Sierra Leone are the Ministerial statements which only include confirmed numbers whereas if you dig a little further you can find the full updates including suspected and probable cases in the Ebola Situation Reports.
Here is the look at the situation in Sierra Leone by District based on the most updated Ebola Situation Report (Vol. 78 dated 14 August 2014).
The above infographic details the cases and fatalities from Ebola in Sierra Leone. Cases include all suspected, probably and confirmed cases and the provisional CFR is based on those total numbers.
As you can see the Kailahun and Kenema Districts are the most impacted regions with Sierra Leone accounting for 84.7% of all cases and 95.8% of all fatalities.
One striking fact emerges from Shane's infographic, which Sierra Leone official reports mention but hadn't sunk in with me yet: the absence of cases in the province of Koinadugu. Is it genuine, or an artifact of incomplete surveillance?
I can appreciate the government's sticking to confirmed Ebola cases only; "probable" and "suspected" cases could be due to other infections altogether. But unless we also have evidence of Marburg or Lassa fever or some other disease with similar symptoms, Ebola does seem like the best guess. (And if an epidemic of some other disease is going on simultaneously with Ebola, West Africa needs to know it.)