Selection and description of participants

All surveyed caregivers had a child between six months and five years of age who had an episode of diarrhea that occurred less than two months prior to the interview and that lasted for more than two days. Table 1 presents a demographic profile of caregivers. We used quota sampling methods to identify appropriate numbers of urban and rural respondents

The survey covered five regions – Centre (including Ouagadougou, the capital), Boucle du Mouhoun, Hauts–Bassins, Nord, and Est – representing the country’s major socio–cultural groups (Table S1 in Online Supplementary Document(Online Supplementary Document) for demographics of each region). We also used quotas for regions, with the sample distributed according to the relative population of each region. Within regions, sampling points were purposively selected, and interviewers used systematic random sampling to identify respondents during recruitment (Table S2 in Online Supplementary Document(Online Supplementary Document) ). Unlike the qualitative research, the quantitative research did not set criteria related to ORS awareness or experience so we could establish a representative measure of population awareness and usage of ORS and other diarrhea treatments. More information about the selection of participants can be found in Appendix S2 in Online Supplementary Document(Online Supplementary Document) .

Survey focus and design

The survey focused on the child’s most recent episode of diarrhea. In the 60–minute interview, topics probed included diarrhea duration, treatments used, sequence/timeframe of administration, and caregivers’ expectations for each treatment (eg, “what did you think [the treatment] would do for your child?”). Caregivers were also asked about the treatment source, spending on treatment, care–seeking behaviors, and dosing of ORS and homemade sugar–salt solution (HSSS). The survey did not probe related costs such as transport or lost work time. Peak dosing estimates were calculated using conservative assumptions to err on the side of overestimating the amounts given. The survey questionnaire is available upon request from the corresponding author.

Other topics covered included awareness and previous use of treatments (ORS sachets, HSSS, other home remedies, herbal remedies, antibiotics, anti–motility drugs, and zinc syrups/tablets). After recording spontaneous recall of treatments used, interviewers used localized illustration cards to prompt or to assist in recall of treatment types. They also assessed perceptions of ORS by using positive–negative statement pairs and evaluated willingness to pay for a diarrhea treatment. Attribute association was carried out based on the four main treatments used (established previously in qualitative research): ORS, antibiotics, anti–motility drugs, and HSSS. This involved caregivers selecting which treatments they felt fulfilled each attribute (such as “easy for children to take,” “stops the diarrhea,” and “not expensive”). Interviewers also asked caregivers to rank the four treatments on effectiveness and value (HSSS was not included for the perception of value).

The survey was pretested with a small number of caregivers (n = 20) in Ouagadougou. This ensured that survey questions were appropriate and refined before widespread data collection.

Analysis

Quantitative data were analyzed for all respondents as a whole as well as for key groups, such as ORS users vs ORS non–users (based on usage at last episode of diarrhea) and urban vs rural respondents. We also evaluated data according to demographic and regional splits.

Data from open–ended questions were analyzed through a similar procedure. The process began with review of verbatim responses for each question. Key common themes were identified for each question, as well as factors associated with each theme. This represented a code frame. Each verbatim response was then analyzed and assigned to its appropriate code.

Duration of illness, treatments used and ORS dosing

Caregivers reported that the average diarrhea episode lasted 4 days (SD ± 1.96 days), with 15% of cases lasting 6 to 10 days. The vast majority of caregivers (84%) took some form of action when they realized their child had diarrhea. Notably, a greater number of caregivers in the Est region decided to “wait and see” before acting (42% vs 16% overall). Among caregivers who reported taking action after realizing their child had diarrhea (n = 385), 46% said they gave their child a special kind of food that would make him or her feel better or adjusted the child’s diet in some way. Among these caregivers, 33% reported that they stopped or avoided certain food types, and 27% said they increased liquid given to the child.

When asked about treatments, 328 caregivers (82%) said they were aware of ORS as a treatment option, and 385 (96%) administered some form of treatment. Among those who administered treatment during the last episode, less than half reported using ORS, even when illustration cards were used to prompt responses ( Table 2 ). However, this is twice the coverage reported in a meta–analysis by Wilson et al. in 2012 [7]. Antibiotics were the second most popular primary treatment, followed by zinc syrup or tablets. Among caregivers who used more than one treatment (n = 166; 43% of those using treatment), there were many secondary treatments: 33% used zinc remedies (19% syrup; 14% tablets), 20% used antibiotics, 15% used anti–motilities, 14% used herbal remedies, and 10% used ORS ( Table 3 ).

Among ORS non–users, more than half used antibiotics, with especially high rates of use in the Nord and Centre regions ( Table 2 ). Overall, most caregivers (57%) used only one treatment (ie, monotherapy), and only 10% used three or more treatments. Among caregivers who used only one treatment, 33% exclusively used ORS, and 30% exclusively used antibiotics. Among caregivers who did something to treat their child’s diarrhea (n = 385), 8% exclusively used a combination of ORS and zinc remedies and 5% used a combination of ORS, zinc, and other treatments.

Of the 177 children treated with ORS, 101 (57%) were less than 2 years old ( Table 4 ). On average, caregivers who used ORS started doing so 1.8 days after the onset of diarrhea, administering it for an average of 2.8 days and using an average of 2.5 sachets during the episode. The average volume of ORS given per day to children less than 2 years old was approximately 600 ml, and children aged 2 to 5 years received an average of 740 ml.

Treatment goals and reasons for using treatments

Caregivers were asked to rank four key diarrhea treatment goals. They ranked “prevent child’s condition from getting worse” as the most important treatment goal, followed by “restore child’s energy and appetite.” “Reduce diarrhea motions” was the third most important, and “replace fluid lost due to diarrhea” was the lowest priority ( Table 5 ).

When caregivers who used ORS were asked why they decided to use this treatment, half said they were “instructed to do so by a nurse” ( Table 6 ). Similarly, about half of those who used antibiotics reported that they had been given a prescription or a health care provider recommended it. Almost all those who used zinc products reported doing so because of a recommendation by a health care professional or community health worker.

Caregiver perceptions of ORS

Overall, caregivers who were aware of ORS had positive perceptions of the product ( Table 7 ). About two–thirds perceived ORS to be easy to prepare, and more than half said that clean water was easy to obtain and that instructions to prepare ORS were clear. About a third, however, reported that it was difficult to get the child to drink ORS, and many expressed uncertainty about whether use of ORS requires a young child to drink too much liquid or whether the frequency of giving ORS is acceptable.

Most caregivers said they believed that ORS reduces the frequency of bowel movements, stops diarrhea, increases the child’s energy and appetite, and replaces lost fluid/water and minerals ( Table 7 ). Nearly two–thirds of caregivers indicated that ORS is not an expensive treatment.

Caregiver perceptions of ORS varied substantially across regions (Table S3 in Online Supplementary Document(Online Supplementary Document) ). For example, those in the Nord region especially perceived ORS as too much liquid for a young child to drink (56% agree).

Caregivers who did not use ORS at the last episode were prompted with a list of reasons for not using ORS and were asked to state whether the reason applied to them. The main reasons why ORS was not used pertained to administration and palatability: 24% said it was too difficult to get their child to take ORS, 24% said their child did not like the taste of ORS; and 20% reported having ORS left over that was then wasted ( Table 8 ). Fewer than 10% of caregivers said that ORS was too difficult to obtain or too expensive.

Asked to associate treatment types with a series of product attributes read aloud, ORS users associated ORS with positive attributes describing efficacy, availability, convenience, and trust ( Table 9 ). Noticeably, ORS users, compared with non–users, more strongly associated ORS with being recommended by health care professionals, easily available, easy to prepare, good for restoring the child's energy and appetite, safe for children under 5 years old, and able to “significantly” reduce diarrhea motions. Additionally, ORS users were less likely to associate antibiotics with the same attributes, whereas non–users tended to associate both ORS and antibiotics with similar attributes.

Among all caregivers, ORS was ranked most effective at treating diarrhea (40% vs 24% for antibiotics) and the best value for money (53% vs 16% for antibiotics) (Table S4 in Online Supplementary Document(Online Supplementary Document) ). However, ORS non–users felt that antibiotics were more effective than ORS for treating diarrhea. Interestingly, ORS non–users said ORS and antibiotics had similar value for money. Also, there were substantial differences in rankings across regions.

When prompted to select a single preferred treatment (regardless of cost), caregivers generally preferred ORS. However, some regions showed a preference for antibiotics (Table S5 in Online Supplementary Document(Online Supplementary Document) ). In addition, herbal remedies were especially popular in the Boucle du Mouhoun region, and zinc syrup was preferred by many in Hauts–Bassins.

Treatment sourcing

Community health workers were the primary source of treatment recommendations for users of ORS or antibiotics (Table S6 in Online Supplementary Document(Online Supplementary Document) ). Recommendations by doctors, nurses, or pharmacists were fairly infrequent. Most caregivers acquired ORS or antibiotics from pharmacists or public clinics/hospitals.

Additionally, 70% of caregivers said they had previously received information about ORS, and more than half of those in the Nord region (53%) and Est region (62%) recalled hearing about ORS within the past 3 months. Among the 282 caregivers who recalled hearing about ORS, information was most commonly heard at a health center (38%), at a hospital (22%), from neighbors or relations (20%), or from television/radio advertisements (18%).

Access to health services

Among all caregivers, the average travel time to a pharmacy, a community health center, and a general public hospital was reported to be 23.2 minutes, 29.3 minutes, and 35.9 minutes, respectively (Table S7 in Online Supplementary Document(Online Supplementary Document) ). The average travel time to a hospital varied across regions, from 25.5 minutes in Hauts–Bassins to 47.6 minutes in Nord.

Nearly half of all caregivers reported sometimes visiting a traditional healer when their child was ill (with any disease, not just with diarrhea) (Table S8 in Online Supplementary Document(Online Supplementary Document) ). The greatest use of traditional healers occurred in the Est (81%) and Boucle du Mouhoun regions (59%).

Financial considerations and preferred product formats

Nearly all caregivers who reported using ORS or antibiotics paid for the product (97% and 96%, respectively), with a median product cost of US$ 1.00 (CFA 1 = US$ 0.0020181 as of September 2014) for antibiotics and US$ 0.20 for a single ORS sachet ( Table 10 ). In addition, nearly half (46%) of ORS users also purchased water to make the ORS.

On average, caregivers spent a total of US$ 0.80 for both the ORS and the water. Rural caregivers tended to spend less (median = US$ 0.40 for ORS sachets and US$ 0.60 for water) than urban caregivers (median = US$ 0.60 for ORS sachets and US$ 1.00 for water). In addition to treatment costs, 51% of caregivers paid for other health–related expenses, such as consultation fees (92%; median = US$ 0.30) or a medical card fee (12%; median = US$ 0.30).

Perception: ORS treats symptoms of diarrhea

Prior research has found that a primary contributor to nonuse of ORS in some countries is the perception that ORS is ineffective for treating the symptoms of diarrhea. This perception, often held by providers as well as caregivers, influences both whether ORS is prescribed/recommended and whether it is requested by caregivers [9]. A number of studies have shown that a sense of unmet expectations underlies this perception. Caregivers expecting ORS to reduce or stop the diarrhea may be disappointed by the product [10].

Our findings, however, suggest that the perception that ORS is ineffective may not be the largest barrier to uptake. Among ORS non–users in this study, only 11% mentioned lack of efficacy in stopping diarrhea as a reason for not using ORS. Our data suggest that most caregivers, in fact, believe that ORS stops diarrhea, and only a small proportion of caregivers (12%) ranked rehydration as the most important or second most important treatment goal. This suggests a fundamental lack of understanding of the critical contribution of dehydration to diarrheal–related deaths and the role of ORS in rehydration. Among ORS users, only 7% chose ORS because of its role in treating dehydration, whereas 68% reported using it because it was recommended by a nurse or was well known by caregivers and nurses. However, 56% of ORS users perceived that ORS helps to replace fluid compared to only 30% of non–users. These findings suggest that ORS usage is mostly due to recommendations by providers and not to an understanding of the core function of ORS. However, it does appear that more ORS users than non–users associate ORS with treating dehydration, suggesting that personal experience improves perception and understanding.

To set appropriate expectations among caregivers, communication messaging should focus on the seriousness of dehydration due to diarrhea and the ability of ORS to prevent dehydration. Our findings highlight the potential to increase uptake of ORS through clear, consistent communication around dehydration and the role of ORS. As suggested by Coreil and Genece, caregivers who understand the importance of ORS in treating dehydration are significantly more likely to use it [11]. Additionally, marketing ORS as a medicine to increase strength, rather than as an antidiarrheal, may improve usage given the concerns around unmet expectations as noted by Green et al. [12]. Surveyed caregivers in Burkina Faso identified increasing strength and energy as an important treatment goal, and ORS users already associated ORS with this goal.

Implications of our findings for future program development in Burkina Faso include a shift in messaging from use of ORS as an antidiarrheal to use of ORS to address dehydration and increase energy. Awareness–raising around the central role of dehydration in child mortality would be beneficial. Additionally, given the importance of recommendations by health workers, some communication/education efforts should focus on providers to increase their desire to recommend ORS.

Perception: antibiotics and drugs offer a better alternative

Prior research has documented widespread, inappropriate use of antibiotics and other drugs for treatment of diarrhea. This is likely due to a preference by caregivers for treatments they see as “powerful” [13], especially for treating the symptoms of diarrhea. Because of the self–limiting nature of most cases of diarrhea, the point at which caregivers begin ORS administration, and the role of ORS in rehydration and not treating the symptoms, it is not surprising that caregivers seek a treatment that that they believe will give more immediate results in halting symptoms. In addition, there is a well–documented history of provider misinformation leading to recommendations of inappropriate treatments [14]. In a study in India, for example, 59% of prescriptions for children with diarrhea were for antibiotics [15].

Although we found ORS usage to be quite high in Burkina Faso (44% of caregivers used ORS at last episode), usage of antibiotics was also high (34%). Usage of both appears to be driven by professional recommendations, indicating a lack of consistent messaging among health care providers. ORS usage in Burkina Faso may be relatively high compared to antibiotic use, despite comparable views of efficacy, because ORS is perceived to be relatively inexpensive, easily available, and better value for money than antibiotics. Caregivers generally preferred ORS as their one preferred treatment, except in the Centre region, where antibiotics were preferred.

Zinc usage appears to be much higher in Burkina Faso than in India and Kenya [6]. Overall, 27% of caregivers used some form of zinc (either syrup or tablet) at the last episode of diarrhea. The relatively high use of zinc in Burkina Faso is likely due to the work of the Ministry of Health, supported by UNICEF and the Micronutrient Initiative. In a recent report, UNICEF noted that the proportion of community health centers using the combination of ORS and zinc to manage diarrhea went from 0% in 2010 to 42% by September 2012 [16]. Caregivers confirm this dramatic change, with 88% of those using zinc products noting a recommendation by a health care provider as the reason why they used the product.

The role of providers in decision–making and the fairly positive opinion of ORS efficacy among caregivers suggest that the elimination of inaccurate or mixed messaging regarding alternative treatments by providers will likely lead to improved use and outcomes. It is crucial for messaging to indicate that diarrhea will generally resolve without sophisticated or expensive medicines and that lost fluids must be replaced. Raising awareness about the rise of antibiotic resistance is a broader concern. Although pharmacist recommendation may be partly motivated by profit, providers are generally motivated to recommend the best available prescription despite caregiver requests for specific treatments. Well–informed providers and experienced mothers are the natural best advocates for successful dissemination of key messages.

Perception: ORS is not user friendly

Prior research indicated that challenges with preparation and administration are the main reasons that ORS is perceived to not be user friendly. Measuring and mixing the proper amounts of ORS and water has been a challenge in household settings without proper measuring utensils [17]. In some settings, people describe duration and frequency of preparation and administration as a challenge [10]. Historically, adequate ORS dosing has been a hurdle, with most children consuming only minute quantities [18,19]. Palatability has also been investigated as an obstacle to administration of ORS. Findings from Burkina Faso echo findings from previous research studies in other countries in terms of the perception of ORS as not being user friendly.

Among caregivers aware of ORS, most do not have issues with administration aside from there being leftover liquid. However, 32% of caregivers in our study reported that it is difficult to get the child to drink ORS, and only 41% of children received a 1000 mL or more of ORS per day. Children under two and children in rural settings were at higher risk of receiving only a lower dose.

Our findings point to the potential for increasing ORS use through interventions to address ease of administration. The issue of leftover liquid may be addressed by offering a smaller ORS packet (200 mL, for example). Offering child–preferred flavors of ORS may be one way to improve upon the current difficulty of getting children to drink ORS. Unique product presentations such as a juice–box or pre–mix may provide one option to address administration and dosing challenges (though cost would need to be taken into consideration). In addition, a more sophisticated product presentation might simplify administration and enhance caregiver confidence that ORS is a “powerful” medicine. This is supported in the research by findings that the pre–mix was well received by caregivers. Flavor improvements might also improve uptake.

Study limitations

Limitations of this study include:

Caregivers’ recall of reactions to a diarrhea event that happened up to 2 months before the interview may be inaccurate. Recall of measurements of treatments used may be particularly inaccurate. We tried to counter this by providing illustrations of bottles/cups and using the illustrations as an aid when asking respondents how much liquid was used.

There was some confusion among caregivers about which treatments fall into which categories. Caregivers often do not know what type of treatments providers advise/prescribe. Illustration cards with pictures of typical treatments that fall into each category (eg, antibiotics, anti–motility drugs, etc.) were used to aid recall.

Caregivers may have been influenced by illustration cards and just chosen a treatment shown that they didn’t actually use. However, we used spontaneous recall before using illustration cards, and the spontaneous and prompted recall of treatments was fairly consistent, indicating that prompted recall is reasonably reliable.

We did not probe into use of zinc (eg, doses used, timing, etc.). Higher–than–expected use indicates that motivations for use, sources of information, and usage practices, including use with ORS, should be explored further.