An Ebola doctor’s answer to question of who to admit for treatment 

The Bong County Ebola Treatment Unit in Liberia last year. (Michel du Cille/The Washington Post)

The first rule of Ebola treatment was distance, doctor Adam Levine knew. That meant standing back from the woman having a seizure on the floor of the clinic in rural Liberia where he was volunteering. It meant wearing a stifling protective suit that limited his ability to work long hours, even when there was so much to be done. And it meant conducting triage interviews from behind a more than four-foot-wide barrier, hoping that questions and visual assessments would be enough to distinguish those desperately in need of Ebola treatment from those who ought to be sent elsewhere.
Each interview involved weighing two opposing but equally important concerns. On the one hand, ending the epidemic required isolating and treating every Ebola patient so the disease wouldn’t spread. On the other, the Bong Ebola Treatment Unit (ETU) — a modest assemblage of beds and tents set up by Save the Children last September — was a dangerous place for someone not already infected with the virus. Admit someone who might not have Ebola, and Levine risked exposing them to an infection they hadn’t needed treatment for in the first place.
Levine, a professor of emergency medicine at Brown University, spent about a month at the ETU in Liberia’s rural Bong County last fall. During his time there and in the weeks after his return to the United States, the clinic admitted 395 patients, only 160 of whom wound up testing positive for Ebola. He and his colleagues needed a better system for triaging patients — one that could accurately predict who was sick with Ebola given the limitations imposed by safety-mandated distance and the clinic’s strapped resources.
“The Ebola Prediction Score,” published recently in the Annals of Emergency Medicine, is the first proposal for such a system developed according to empirical data from an Ebola outbreak. Based on data from the 395 patients admitted during the first 16 weeks of the Bong ETU’s operations, the score is as good as the existing World Health Organization algorithm at figuring out who might have Ebola and even better at figuring out who might not.
To develop the system, Levine and his team of researchers looked at admission interviews for each of the 382 patients for which the clinic had complete data. They then cross-referenced the 14 symptoms identified during those interviews with the patients’ actual diagnoses to figure out which symptoms were the best predictors of a positive test result. The researchers found six symptoms — contact with another sick person, diarrhea, anorexia, muscle pain, difficulty swallowing and absence of abdominal pain — that corresponded most closely with an Ebola diagnosis and assigned each one a point value. The higher a patient’s score, the more likely they are to have the disease.
These symptoms were useful indicators of Ebola in part because they are less associated with other diseases, Levine said.
“One common sign that everyone thinks about is fever. But fever is not very good at differentiating Ebola from other diseases like malaria or typhoid or even influenza that are really common in West Africa,” he said. 
The Ebola Prediction Score is just as sensitive (meaning that it rarely fails to identify a sick person as such) as the World Health Organization algorithm — the main goal of any predictive tool. And Levine believes his system is easier to use, particularly in a harried treatment center setting where health workers are pressed for time and resources.
“The more complicated the triage algorithm is the more difficult it is to train [staff] and more errors are going to get made,” he said. “… A simple scoring system is a little bit more intuitive. You just have a list of symptoms on a page with the points assigned and people just add up the points and they know whether the person is likely to have Ebola and how likely they are.”
The Ebola Prediction Score is no replacement for definitive laboratory testing, the study’s authors say. But lab tests can take one to three days, and patients without Ebola will spend that time among the sick at the ETU — all the while putting themselves at risk of exposure to the virus.
“Managing an ETU requires balancing the epidemiologic imperative of trying to find every last case in order to stop the epidemic against the ethical imperative to do no harm,” Levine said. “There really perhaps is no greater harm that I think I’ve done in my career than admitting patients without Ebola to our ETU, which I did many, many times, because I put those patients at risk.”
The score is also useful because it ranks patients according to risk. This allows medical personnel to organize patients according to the likelihood that they are ill with Ebola, hopefully isolating the truly sick from those who don’t have the disease. High-risk patients could be assigned a different latrine, Levine suggests, and be treated in a different ward.
What health workers really need now is a cheap, rapid diagnostic test that can effectively determine who does and doesn’t have Ebola. One such test was just approved by the World Health Organization in February — rather than a days-long wait for results from a lab, this test requires only a few drops of blood from a fingerprick and takes about 15 minutes, according to Science Insider.
But the test doesn’t eliminate the need for a confirmation test or for a health worker to dress up in full protective gear just to get a drop of blood, Levine pointed out. An ideal diagnostic would be even faster and far more sensitive. If it works the way rapid tests for other diseases have, such a test could “revolutionize care,” he said.

- Morning Mix