When the British medical journal The Lancet called it “potentially the most important medical advance this century” in 1978, its editors weren’t referring to penicillin or the heart-lung machine.
The publication had heaped this praise on a mixture of glucose, sodium, and potassium stirred into water. Oral rehydration solution (ORS) can mitigate what has been a scourge of humanity for millennia: diarrhea. Although mostly just a nuisance for adults, diarrhea can cause fatal levels of dehydration in children.
Yet, according to data from UNICEF, nearly a half century after The Lancet‘s declaration about ORS, diarrhea still kills 444,000 children worldwide every year. More than 30,000 children younger than 5 succumb to the condition in India annually despite the nation’s rapid economic growth and growing middle class.
Ironically, recent research suggests that India’s economic advances could pose one of the biggest barriers to making ORS the standard of care for pediatric diarrhea. Medical providers have the perception that parents want seemingly more sophisticated treatments, and India’s burgeoning market for pharmaceuticals such as antibiotics also plays a role.
These trends were highlighted in a study by researchers at RAND, the California-based think tank, and published by Science in February 2024. It was unconventional in that its authors hired actors to pose as parents of children with diarrhea to obtain medications—and data.
“This is the gold standard for measuring provider behavior,” said Zachary Wagner, the study’s lead author and now an economist with the Center for Economic and Social Research at the University of Southern California.
Other experts concurred, but noted that such studies need to rigorously address certain ethical issues. These include whether the secret shoppers are being revealed to the study subject and whether they take up too much of a clinician’s time that they actually affect the quality of care being delivered to other patients.
Secret shopper studies need to rigorously address certain ethical issues
David Levine, an economics professor at the University of California at Berkeley who has studied ORS distribution in developing nations, observed that some patients have an implicit bias against it. Most ORS in India is either distributed in small paper sachets that resemble tea bags or in juice boxes accompanied by a piercing straw.
“It’s just assumed you go to the doctor to get treatment, and so being sent home for rest and fluids is a breaking of the [implied] contract,” Levine said. “ORS is just sugar and salt. It doesn’t look like a medicine, and it’s not packaged like a medicine.”
Unconventional Measurements
The intent of the RAND study was to determine why ORS was not being prescribed more regularly and whether parent preferences play a role in how it is dispensed. The study was massive: The dispensing habits of more than 2,200 medical providers—including physicians, traditional Indian medicine providers and osteopaths, rural medicine practitioners and pharmacies—were observed. The study took place in the Indian states of Karnataka and Bihar, chosen for their economic and demographic diversity. Karnataka is significantly wealthier than Bihar and its population is better educated.
Wagner noted that simply observing typical encounters between clinicians and parents were not likely to yield the most reliable data.
One reason is that observers might wait hours, if not days, for a parent to come into a clinic to seek treatment for a child with diarrhea. The second is the Hawthorne effect: study subjects behaving differently because they are aware of being observed. In the case of prescribing patterns, this issue was enormous.
“We know there’s this big gap in what providers say they do versus what they actually do,” Wagner said. Although the study noted that nearly 90% of providers said they would prescribe ORS for child diarrhea without prompting, the actual rate was a fraction of that.
Wagner said that 25 data collectors—a fairly common job in India he equated to census workers in the United States—were trained for two weeks to pose as a parent of a 2-year-old child who had had diarrhea for two days. The child’s symptoms as described lacked features such as fever or bloody stools, which would lean toward a bacterial infection diagnosis requiring antibiotics.
All of the secret shoppers were men in their 20s and 30s and sought treatments unaccompanied. Wagner said that that is a customary clinical encounter because mothers usually stay home with ill children. Along with being coached for their roles, the secret shoppers were also trained to recall and record the specifics of each encounter.
The data is sobering. Despite 86% of providers saying that they would prescribe ORS without prompting, only 28% actually did so. More than 70% prescribed antibiotics instead. If the secret shopper stated a preference for antibiotics, the prescription rate for that medication rose to nearly 80%. However, when the secret shopper stated a preference for ORS, that prescription rate nearly doubled, to more than 55%.
Wagner has concluded many providers are self-conscious about prescribing ORS, which they believe their patients do not consider a real medication. That in turn prods them toward antibiotics.
“They know they’re not generally supposed to be prescribing antibiotics all the time for child diarrhea, but they do it because they think that’s what patients really, really want,” he said. “And if they don’t do it, it makes them look bad.” That trend is further exacerbated by the fact that two-thirds of the providers secret shoppers visited had had a visit from a pharmaceutical sales representative in the previous four months, Wagner noted.
Arthur Caplan, a professor of bioethics at the New York University Grossman School of Medicine, noted that pharmaceutical sales representatives in India often offer ethically dubious inducements to stock and sell products such as antibiotics.
“It’s free food, free tickets, and maybe even under-the-table payments—that sort of thing,” Caplan said.
Wagner acknowledged that that was also an issue. “One of my Indian colleagues was telling me a saying that ‘if five patients are in a waiting room, at least one of them is a pharmaceutical rep,'” he said.
Ethical Considerations
Using secret shoppers in studies raises a few ethical concerns. Kelsey Rankin, MD, was lead author of a 2022 study that praised the use of secret shoppers for providing “insight into the challenges of access to health care that may be difficult to measure through more standard investigative techniques.” But it also warned that secret shopper studies can introduce complications such as taking time away from providers to treat patients who have genuine health issues. However, in the RAND study, the average encounter was only four minutes.
They were told that a secret shopper would be visiting to obtain care, but not specifically when they would
Caplan also recommended secret shopper studies be reviewed by outside parties for ethical issues, such as dealing with providers who may take offense if they later learn they had an encounter with an ersatz patient. The RAND study’s construction was vetted by both that institution’s Human Subjects Protections Committee and the Karesa Independent Ethics Committee in India.
The biggest concern that both Levine and Wagner raised in secret shopper medical studies is the deception of using a fabricated patient to obtain data on provider behavior. However, the RAND study obtained the consent of all participating providers in advance. They were told that a secret shopper would be visiting to obtain care, but not specifically when they would.
“That’s a much higher level of transparency than usual” in such studies, Levine observed.
Potential Solutions
Levine noted that Uganda has had a successful program distributing free ORS at clinics. India also once had a free distribution program in place, but it relied on community health-care workers known as ashas to provide it, according to Levine.
“Ashas are very poorly paid and have a very long list of duties, and were given a very small amount of bonus pay to distribute ORS,” he said. “We didn’t see much evidence that in most parts of the country they were actually distributing it.”
Instead, he suggested that India develop a program to get free ORS not only into the hands of medical providers and pharmacies, but into nonclinical institutions as well. Levine suggested that India’s approximately 1.4 million government-supported childcare centers called angawandis would be a good place to stockpile supplies.
“They should just be distributing ORS to parents before there’s an illness,” Levine said.
