When I start to analyze data for a study, part of me always hopes there is some clear-cut answer that doesn’t have to be qualified in any way, that I can make a blanket statement without any nuance. The truth is always messier and more complicated, as the data I analyze are from the real world, representing real people and real places. As an epidemiologist on the U.S Agency for International Development’s (USAID) Act to End NTDs | East program, I have both the challenge and the opportunity to find some significance and meaning amidst often complex neglected tropical disease (NTD) data.
Two years ago, I was asked to work on an analysis of lymphatic filariasis (LF) survey data. Spread by mosquitoes, lymphatic filariasis is a parasitic disease that causes thread-like worms to develop in a person’s lymph system. Left unchecked, the adult worms can lead to fluid buildup in the limbs and genitals, causing disability and often leading to stigma and economic hardship. Treatment of whole communities for at least five years can lead to such a successful drop in prevalence that treatment campaigns can stop. But sometimes districts show protracted progress, “failing” pre-Transmission Assessment Surveys (pre-TAS), the first step in the process for assessing LF elimination. The question was why?
When we started to see a handful of districts failing their pre-TAS for LF, we knew we wanted to look deeper. When I first started analyzing this data, I was still fairly new to the field of NTDs and learning the many interconnected things that make LF elimination possible. Treatment of whole communities at risk for the disease may sound relatively simple at first glance, but the reality is that it’s actually rather complex. Sometimes people aren’t home, or they have migrated to other areas for a few months, other times they aren’t willing to be treated, fear side effects from taking medicines, or feel like they are healthy so they don’t need to take the medicines. There are other aspects that could impact elimination as well, like vector and parasite dynamics, along with some environments being more suitable to transmission.
The aim of this analysis was to bring together not only data from 13 countries funded through USAID’s NTD programs but also geospatial data from external datasets. There were a lot of potential variables. In fact, I had them all stuck up on my office wall as individual Post-It notes for several months in hopes of creating of conceptual framework that would help explain what was going on. Everyone who came by to visit me commented on them, and I tried to explain, but ultimately ended up saying “It’s complicated.”