Les Roberts – Freetown, Sierra Leone – April 19, 2015
I am getting ready to go and I have found that my main point from all of my field work is a hard sell here in Freetown. There are three ways we are finding out about new cases these days:
- People call 117 or walk into a facility because they suspect they are sick (and this has been most new cases in the last couple months)
- People are screened at a clinic or a road check point or a house visit during quarantine and get detected
- As every corpse is supposed to be tested for EVD and probably half are, we detect some from death swabs.
As a humanitarian community, category 2 above consumes about 90% of our surveillance energy and has, over the last couple months, probably produced about 10% of our new cases. Thus, I want to cut back our energy on those things and spend more effort finding and swabbing every corpse… but few are at peace with that suggestion.
In the most Ebola lingering District in recent weeks, Kambia, they had 880 blood samples taken in March and one was positive. There have been a couple of positive samples so far in April, so it is likely the rate of positives in April will be a couple times higher. That contrasts with four percent positive death swabs in Kambia during March. That means per sample taken, a death swab is at least ten times more likely to identify a new case. Finally, about half of all deaths in Sierra Leone, perhaps a little more, are being tested for EVD. If we assume that people experience a fever of 38°C and three of the potential Ebola symptoms (e.g. headache, body ache, exhaustion, diarrhea, vomiting) on average once per year, then not one in a thousand people who should be coming for testing are coming in. Thus, when you combine the rate of positives with the effort needed to get samples with the coverage rate, you would think all the effort would be going into getting more and more deaths sampled and safely buried.
But, others use other calculus. Some people value finding people who are still alive because one can get them into treatment and perhaps save them. Given that 20 or 30% who do not go for treatment survive and that 30% or sometimes 50% survive in the ETU’s, the marginal increase in the chance of survival is perhaps 20 or 25%. (There are folks who will protest this but they usually do not distinguish those detected by others with the self-induced early treatment seekers or include all the chances of death after being tested while waiting for transport, being transported…). There are others, especially Sierra Leoneans, who know that getting more chiefs and remote villages to call in deaths involves huge cultural barriers.
I just repeated a sensitivity exercise in two high incidence Districts that we had assessed in October. The system is no better. We missed two thirds of cases then, we probably have missed two thirds in 2015. The solution would be to put a lot more effort into monitoring all deaths. That is not going to happen…. and it might not be that big a deal. The outbreak is winding down because of social changes and widespread safe burial practices and services. In the grand scheme, it may not be important that the surveillance is poor.
Epidemiology… it might not always matter? Clearly, time for me to go home!