Of the 295 private sector providers included in the sample, 232 (78.6%) agreed to participate in the assessment. Of these, 97 were included in both the interview and the observation and 135 provided information in the interview alone ( Figure 1 ). In Table 1 , we present an overview of the demographics and training of the providers included in our assessment stratified by the informal and formal sector. We also present an overview of the demographics and training of the providers included in our assessment stratified by informal vs formal sector. Among all interviewed providers, 30.6% had a formal medical degree (including MBBS/MD, pharmacy, and nursing) and 69.4% were practicing as part of the informal sector with either Ayurvedic/homeopathic training (21.1%) or no recognized degree or training program (48.3%).

Figure 1.  Study sample. Asterisk indicates that the information collected from village administrative heads, other providers, and local shops.

Ninety–five percent of providers reported prescribing drugs/medications but only 89% reported providing diagnosis or patient consultation; the remaining 11% only sold drugs/medications. In the previous 6 months, 155 providers (67%) reported they had been visited by someone who provided training/information about diarrhea treatment. In addition, 84% of private providers reported seeing posters advertising zinc for diarrhea treatment. There were no differences between formal and informal sector providers with regard to having been visited by a drug seller or having seen a video about diarrhea treatment in the last 6 months, nor were there any difference between the two groups in the proportion who had seen a billboard advertising zinc and ORS. Because the survey was not designed to compare provider types and given there were no differences in these programmatic indicators, we did not stratify by formal/informal in subsequent analyses.

In Table 2 , we summarize provider responses on the diarrhea treatments typically recommended during five case vignettes. For simple acute diarrhea (5 loose, watery stools/d for 3 days), providers reported recommending ORS (68.1%) and antibiotics (65.9%) most often. However, the proportion reportedly prescribing ORS decreased for the remaining four scenarios, each of which represented more severe diarrhea. For simple acute diarrhea, 35.8% of providers reported prescribing zinc. In all five scenarios, providers reported advising antibiotics more frequently than zinc. As the severity of the case vignette increased with either signs of dehydration, duration of illness, or accompanying signs and symptoms, providers more frequently reported referral in lieu of any type of treatment, including life–saving ORS.

When providers were asked specifically about willingness and frequency of prescribing zinc and ORS, a small percentage reported prescribing zinc to all patients (14%), and an additional 36% reported prescribing zinc to at least 50% of cases [data not shown]. When asked what circumstance would lead a provider to NOT recommend zinc, the most common responses were not fully understanding zinc treatment (57.9%) and not having zinc in stock (30.8%). Providers were asked to recall the duration of zinc treatment; 15.6% reported duration of at least 10 days, in adherence to the current WHO guidelines [2]. Eighty–six percent of private providers reported routinely recommending ORS as part of their practice; yet 42.7% of those could not correctly describe how to prepare ORS. Zinc and ORS were in–stock (ie, providers able to show the stocked products) for 38% and 69% of providers at the time of the interview, respectively.

A direct observation of the provider treating a child 2–59 months of age with diarrhea was conducted for 97 (42%) private providers ( Table 3 ). Of the children treated during the observation session, 71% were male children and the median age was 24 months. Among the 97 providers who participated in the observations, 98% asked at least one question with regard to the history of the diarrheal episode. Zinc was sold in 16.5% of cases and caregivers were recommended to go elsewhere for zinc in an additional 13.4% of cases. Among the 28 providers who recommended zinc, 50% gave no reason as to why zinc would benefit the child and only three administered the first dose during the treatment session. Providers prescribing zinc gave variable instructions on the duration of zinc treatment; 50% recommended zinc for 10–14 days; 18% for 7 days, and 32% either recommended zinc for <7 days or did not mention duration of treatment. ORS was sold as a treatment in 56.7% of cases with an additional 20.6% instructed to purchase ORS elsewhere. Of the providers who recommended ORS, 50% gave correct preparation instructions. Providers commonly prescribed treatments other than zinc and ORS, including antibiotics (61.9%) and antidiarrheals (17.5%). Twenty–six providers (26.8%) were in compliance with current WHO/IAP guidelines and either sold or prescribed both zinc and ORS. Only 2 providers sold or prescribed zinc without ORS.

We conducted a provider assessment of 232 providers from both the formal and informal private sectors in UP, India. Private providers are the mostly widely sought sources of careseeking for pediatric diarrhea in UP; yet there is a dearth of information on private, especially informal, providers and thus it is critical to understand their knowledge and willingness to provide appropriate diarrhea treatment, including zinc and ORS. We carried out this study using short vignettes and direct observations one–year after roll–out of program implementation, by which time a pharmaceutical or NGO representative promoting and selling zinc and ORS should have visited all providers [12].

It is not surprising that we found providers were more likely to report their intention to prescribe ORS and zinc in the short vignette portion of the interview than in the observation portion of the study. This same discrepancy has been observed in Bihar, India in an evaluation of a similar program promoting zinc and ORS. Mohanan et al. found poor overall knowledge of diarrhea treatment among the 340 included providers, and none were found to treat diarrhea correctly when tested by an unannounced standard patient [8].

Despite IAP Guidelines that do not recommend antibiotics for the treatment of acute diarrhea, antibiotics remain the first line treatment for most private providers [5]. We observed consistently higher rates of antibiotic prescribing compared to zinc. In previous effectiveness studies, zinc has been shown to displace the unnecessary overuse of antibiotics [6,12], but this was not observed in our evaluations. The pharmaceutical training provided, as part of the DAZT program did not address the unnecessary use of antibiotics nor suggest that zinc could and should replace antibiotics for simple acute diarrhea. Pharmaceutical representatives may have been promoting other treatments for diarrhea in addition to zinc and ORS, such as antibiotics and antidiarrheals. Providers who prescribed zinc typically did so in addition to antibiotics and ORS. Only two providers recommended zinc in the absence of ORS, lessening concerns that NGO and pharmaceutical representatives might place more attention on zinc than life–saving ORS during drug detailing visits. Program planners employed the drug detailing approach to build rapport with providers before challenging their current practices. The visits were designed to increase confidence among informal providers, the majority of whom do not receive routine or formal training. Our results illustrate that the tendency to overprescribe antibiotics will not be changed quickly, even with the introduction of zinc. More research is warranted to determine how to best influence lasting change in prescribing practices; tactics might include group training sessions, incentives, and/or community promotion.

There were several limitations to our study, including the failure to conduct observations of all providers. It is possible that the prescribing knowledge and practices of observed providers are not representative of providers for whom an observation was not possible due to logistical constraints. We conducted the provider assessment during the dry season to facilitate the logistics of data collection, but as a consequence of this timing, diarrhea prevalence was low in some villages, resulting in less diarrhea careseeking. Our data collectors waited at each provider’s practice for one day but proceeded with the interview alone if no child eligible for observation sought care during that time frame. We also recognize that there is no gold standard method for observing providers in practice. Direct observation may bias providers to give higher quality care—a phenomenon known as the Hawthorne effect. As such, we expect the observation results to represent the best–case scenario, such that the care given to a child during an observed treatment interaction is the highest quality care the provider is able to give [10]. Lastly, our assessment is also limited by its cross–sectional nature and the failure to draw upon repeated measurements across the duration of implementation. The evaluation was originally designed to assess the knowledge and practices of providers early in the intervention and then again at the end of the program, but the plan was changed mid–course by the broader project steering committee. We maintain, however, that the results of this one–time assessment are critical because they shed light on the challenges of changing the behavior of rural providers even early in the program implementation period when training is fresh and stocks are largely in place.

The DAZT private sector model mimics the reality of how drug information is currently delivered to private providers through drug detailing visits. The expansion of drug detailing to include zinc and ORS thus represents a sustainable method, especially for reaching the informal sector with critical information and products for adequately treating childhood diarrhea. The DAZT model has also highlighted the challenges in identifying and working with informal private providers who often operate underground to avoid government penalties for practicing medicine without recognized credentials. However, despite these challenges, informal private providers are often the first choice of caregivers and thus a worthwhile target of diarrhea management programming. If substantive improvements in diarrhea treatment are to be made in rural India, improvements in the treatment practices of the informal private sector will need to be addressed by increasing access to current guidelines and providing formal and informal training opportunities. Policymakers often resist developing programs targeting the informal sector, but denying the role of informal private providers in treating the rural poor is detrimental to the overall goal of improved diarrhea treatment for young children. Programs like DAZT aim to demonstrate that informal providers, including those with little–to–no education, can be taught to adequately treat childhood diarrheal episodes with ORS and zinc and to recognize the signs and symptoms requiring referral. The low–level training provided to public sector community health workers in many countries could be expanded to cover the private sector in areas where the latter is already providing the majority of care and treatment, such as rural UP. Diarrhea remains an important cause of morbidity and mortality among young children, especially the poorest and least–served, in low– and middle–income countries. Policymakers and program planners may not be able to influence the source through which children receive care, but collaborative and targeted efforts can improve the quality of care children receive through the most frequently utilized sources, in turn improving outcomes for all young children with diarrhea.