Deworming pills may be simple and cheap, but they’ve got some big muscle behind them. Nobel laureates, the World Health Organization, and even Pacific Standard have endorsed programs to distribute pills in developing countries where worms are endemic. Such medicines are supposed to improve the nutritional status of kids—having worms can prevent the body from fully absorbing nutrients—and to improve school attendance, because kids aren’t spending as much time home sick with diarrhea, nor spreading the worms to their friends and classmates. Now, however, two new studies question whether deworming campaigns do all they’re purported to. It seems, according to the authors, that ridding kids of worms might not actually affect how often they go to school.
Published yesterday in the International Journal of Epidemiology, the new papers are re-analyses of an influential study originally published in 2004, which found that deworming treatments reduced absenteeism in kids by 25 percent. The fact that the new studies got different results from the original underscores how important it is to re-analyze and repeat research. Even experiments conducted in good faith—and it appears the 2004 one was—can come up with flawed answers. But so-called replication studies aren’t popular with researchers, for a number of reasons. For one, researchers generally get a bigger career boost from conducting new studies, not replicating old ones. Plus, research teams may be reluctant to hand over their data to strangers for replication.
The fact that the new studies got different results from the original underscores how important it is to re-analyze and repeat research.
But what’s small change to researchers can be a big deal to other organizations. Governments, charities, and other groups spend billions in aid each year. Studies influence what they choose to fund. Experts once ranked deworming as the fourth most effective public-health intervention in the world, based on the science. What will groups who fund deworming campaigns do now? They may still find deworming worthwhile—it’s cheap and works to relieve some uncomfortable symptoms—but surely their calculations will now change. Will researchers start demanding re-analysis for other studies, too? If they indeed do, there’s a good chance they discover something new. Replication studies often find something different from the originals they cross-examine. In fact, one recent study of replicated randomized controlled trials found “one-third of them were so different that the take-home message of the trial shifted,” as BuzzFeed News reports.
In this case, a study published in the journal Econometrica in 2004 found that a deworming campaign in Kenya upped kids’ school attendance. However, when an independent team of researchers combed through that old data, they found errors that obviated the school-attendance result. In a second study, the re-analysis team, from the London School of Hygiene and Tropical Medicine, tried running the data through a different statistical tool than the one the original researchers used. They found deworming to have just a small effect on school attendance. Taken together, the two new studies indicate that the evidence for deworming helping with school is … iffy.
“Our findings suggest that on the basis of this study alone, we should be cautious about concluding that there are educational benefits from deworming children,” Calum Davey, a public health researcher who worked on the repeat studies, said in a statement.
More replications like these may lie ahead. The deworming replication was commissioned and funded by the International Initiative for Impact Evaluation, which, for the last several years, has been trying to encourage teams to re-do some of the biggest studies in international aid. After this deworming result, policymakers may start watching the initiative more closely.
