Ebola diaries: Creating ways to understand an outbreak

Mikiko Senga

Mikiko Senga, a WHO epidemiologist specializing in emerging diseases was sent to Kenema, Sierra Leone in early June 2014 to gather data about the Ebola outbreak. There she found herself trying to make sense of information coming in a variety of ways, from bits of paper, blood samples, hospital records, and soon realised she was facing an outbreak about to catch fire. She called for help and, with colleagues who came to support her, set about developing ways to document and understand the size and nature of the Ebola outbreak racing through the district. Here is her story.

Mikiko Senga, Ebola data specialist, Sierra Leone
WHO

"A few days after I got to Kenema, we had the first 3 infections in healthcare workers in a hospital setting. The outbreak was centred in neighbouring district of Kailahun at the time, so we were getting a lot of patients from that area in Kenema. But the moment you find a case within a hospital setting – and these were healthcare workers who did not work in an Ebola treatment centre – then you know something else is going on. We didn’t know how they got infected as there was no established chain of transmission. Later on we figured out that one of the three healthcare workers was an ambulance driver who was involved in burial and funeral activities.

A call for help

Looking at the data and conceptualizing what may be going on in the communities, I called for help. I notified the country office and I alerted colleagues at Headquarters in Geneva. We had a team in Kailahun as well, but they were hours away and also low on resources. Less than one week after my call for help, through the Global Outbreak Alert and Response Network (GOARN), which responded very quickly, an experienced epidemiologist from Public Health England, Chris Lane, was sent to Kenema. Chris had previous experience in Ebola outbreaks in Gulu, Uganda, and knew exactly what to do in terms of case investigation and contact tracing.

Dr David Brett-Major was also deployed by WHO to help with care of the increasing number of Ebola patients at the treatment centre. A team from Tulane University and a laboratory-based researcher from the company Metabiota were already in the area because they had been doing research on Lassa fever there for many years. The fact that they were already there helped a lot, especially since they were known to the communities. This made our case investigations and contact tracing work a lot easier.

Needing more than just a name

One of the major challenges I faced was the lack of standardized data collection forms. The number of cases was increasing, and yet we didn’t have appropriate forms to record the data we wanted. We would have a piece of paper that would come in with the name of a patient and that’s about it. We didn’t have a good surveillance mechanism in place.

"The solution to the data problem, especially from the surveillance side, was to develop and implement a standardized case investigation form. Data management at the beginning was a nightmare, with data coming in in all sorts of forms."

Mikiko Senga, WHO Sierra Leone

Ideally, you would develop a data collection form and know in advance which variables you want to record. Whoever completes the form – be it a case investigation or surveillance officer – needs to complete the entire form. That wasn’t happening, partly because of an overwhelming number of cases, so we had to deal with training and making sure work was thorough, in addition to not having good forms to begin with. Ideally you would have information on what we call patient identifiers: who they are and where do they come from, so that we could go back and trace their contacts if they were deemed to be a suspected case. Correctly identifying the symptoms is crucial, because you want to know if the patient actually had the symptoms that meet the case definition.

And lastly, you need to know all other possible exposures and risk factors so that you can go back and retrospectively investigate individuals who may have come in contact with Ebola virus. But we often had none of those things, only a patient name. In addition, there needs to be a blood sample to complement that piece of paper, right? Sometimes we would have a blood sample without knowing whose it was. Or we had a piece of paper, but no blood sample to go with it. If you can’t match the two, there will be reporting errors. Missing key information impacted subsequent response activities, including our ability to properly perform contact tracing.

Solving the data problem

The solution to the data problem, especially from the surveillance side, was to develop and implement a standardized case investigation form. Data management at the beginning was a nightmare, with data coming in in all sorts of forms. In addition, we also created a contact tracing form that is still being used in Sierra Leone. If the contact tracers don’t know what kind of symptoms they need to look for, then you are not really doing contact tracing. So, there was an improvement. I assisted in the development of standardized data collection methods. Updating the daily situation reports for Sierra Leone happened manually, but I automated that procedure in an Excel spreadsheet.

Should I have done things differently? I think I should have screamed more. I should have been louder, asking for more help. At one point we knew that there were some cases in Freetown. They were brought to Kenema because we didn’t have any Ebola treatment centre or capacity to treat any patients in Freetown at the time. There was great fear of having cases in the capital city, and that was very alarming. We got five cases that came from Freetown and none of them had clear linkages to others—very concerning. If this is happening in the capital, you really should be screaming for help. But as one person you only can do so much, screaming for help and hope to be heard."