Les Roberts – Freetown, Sierra Leone – November 27, 2014
Day 54: Surveillance at odds with justice and the little people
Finally, after weeks in the works, our MoHS surveillance partners presented our findings from three sensitivity assessments to the National Ebola Response Center… which is the national governing authority on Ebola. In one area (Bo) where the outbreak has peaked and seems under control, our little grab from 12 randomly selected villages suggested that ~54% of likely EVD events were in the national database. In the two districts with the highest incidence, this was ~26% and ~36% and there is reason to believe that if anything, these might be overestimates of sensitivity. Thus, the national numbers on cases and deaths are really two or three times worse than you hear in the press. This is completely expected if you are an epidemiologist used to working on outbreaks in Africa, but somewhat hard to imagine if you are a reporter always hungry for the next scandal. What was interesting in these sensitivity assessments was why the surveillance was so poor. Mostly, surveillance is poor because there are not enough treatment beds in the country.
Above is a picture of the first Community Care Center (CCC) in the country. You can see the two tents in the background (one for dry patients, one for wet patients) behind the triage area. If there are good and rapid lab services, people who only have malaria get identified and get out quickly, reducing the chances they get ebola. If people with ebola can be confirmed quickly, ideally they will be moved to treatment units (ETU’s) before they get really symptomatic and start shedding a lot of the virus, and in doing so will reduce the chances of infecting the malaria patients in the CCCs and holding centers (HC) as well. But what really affects the surveillance, is all those measures designed to get people out of the community and isolated because those are the mechanisms associated with initial case detection.
There is a national ebola emergency alert hotline (117) and a similar back-up number in most districts. In Freetown, only about three percent of alert calls trigger a household visit. Lots of them are about a dead dog in the street or are whacko calls. Often there are many calls about the same suspected case. But, there are also cases not investigated because there are no beds. In fact, in the hottest areas, lots of people reported that when they called 117, no one answered the phone. Thus, the use of 117 calls as data to monitor trends is completely undermined because there is systematic under-calling in the areas with the most Ebola. Likewise, in the most rural chiefdoms where transport cannot easily come, people and chiefs just stop calling because they know that no ambulance will ever arrive.
The most central detection workers for the Viral Hemorrhagic Fever (VHF) surveillance system, case investigators and contact tracers, are likewise doing what is best for the people they encounter in the circumstances in which they exist. Why aggressively investigate a case if there are no facility beds to which one can refer them? In fact, in some areas, a case will be left at home and their family will be forced to stay 24/7 in their home with them… and there is some evidence that this kills family members compared to leaving the index case at home without quarantine. In settings with a household quarantine rule and no empty referral beds, a case investigator doing his/her job could induce ebola deaths in those they encounter… so why on earth would they do their job?
Thus, an odd thing is going to happen when this outbreak peaks (which I expect in the next few weeks) and beds come on line, the surveillance system will get better and better and it will look like the numbers will be going up while they are going down. It may look bad in the newspapers, but at least it will induce less and less strain in all those case-investigators and contact tracers who will finally be able to do their job and serve the public health at the same time.
… and on it goes. Cheers,