Diarrhea remains a leading cause of death among infants and young children [1–4]. The major pathogenic mechanism of diarrhea mortality is dehydration, which is responsible for more than half of diarrheal deaths in developing countries [5]. Oral rehydration salts (ORS) dissolved in water to form oral rehydration solution can be absorbed in small intestine, replacing water and electrolytes lost in faeces, and is likely to produce fast recovery and fewer side effects [6, 7]. So it is a safe and effective treatment administered at home or at medical centers [7]. ORS can prevent 93% of childhood diarrhea mortality [4], and the use of anti-diarrhea drugs and antibiotics, which have no clinical benefits, cannot be allowed in the treatment of acute watery diarrhea [8]. However, even up to 80% of the children with diarrhea in some areas were reported to have received no ORS but only anti-diarrheal drugs [8].

Successful reduction of diarrhea mortality in the 1970s and 1980s can be attributed largely to the scaling-up use of oral rehydration therapy and programs to educate caregivers on its appropriate use. Regional data shows that since 2000 there has been slight decline in the proportion of children receiving ORS therapy during episodes of diarrhea [9, 10]. So, the simple and effective tool of child survival has fallen off the priority lists for global and national policy leaders and program managers [9]. How to sustain and achieve a high level of oral rehydration therapy continues to be a challenge [10].

Households and village clinics play an important role in the management of childhood diarrhea [1, 4]. A previous report from Mongolia has showed that more than one-fifth of infant and child deaths occurred at home [11]. In rural China, village clinics located at the first tier of the rural healthcare system supply the population living in the village with many services of disease control and prevention [12]. But village doctors, who were both peasants and healthcare workers, were inclined to use medicine inappropriately [12, 13]. So, it is of critical importance for childhood diarrhea to find new strategies of promoting and sustaining a high level of ORS use in home or village-level care against the diarrhea mortality. This study has assessed the utilization rate of ORS in the care of diarrhea among children under 36 months old and explored the determinants of ORS use in home-based and village-level care. Such a study can provide some insights of promoting ORS use in home-based and village-level care and reducing the deaths due to diarrhea among children under 36 months old in rural Western China.

Oral rehydration therapy is the cornerstone of fluid replacement and national programs to promote ORS have been strongly supported by WHO, UNICEF and USAID in the treatment of diarrhea [1, 16]. Our study found that the overall therapy rate of ORS in all care of diarrhea among children under 36 months old is 34.62%. In the 4 care groups, the utilization proportion of ORS in HBC was only 15.03%, significantly lower than that in VLC, TLC or CLC. It was clear that these caretakers did not use ORS against childhood diarrhea in HBC more often. In our study, overwhelming majority of the mothers of these children completed only a primary education. It is possible that this may contribute to less education about when to appropriately use ORS. The previous study also showed that the devalued status of ORS in the eyes of caretakers had become a major problem [17]. It should be acknowledged that these children with diarrhea in HBC usually had symptoms not as severe as those who were taken to a formal healthcare facility. These may lead to the low utilization of ORS in HBC. So programs of promoting ORS use should give a significant priority to the households. Meanwhile, communication strategies are needed to ensure that families understand and accept ORS as a key treatment component in HBC [1]. Also, there should be an urgent need for caregivers to be educated to use ORS packets at home as early as possible when a diarrheal symptom appeared in their children. GEE model analysis of ORS use in HBC found that the caretakers who cared for more than one child seemed less likely to use ORS in recent diarrheal episode. In the multi-child families in rural China, it is common for the caretakers to delegate some care burdens to their older children. However, other children in the household may not have the appropriate knowledge or skills to care for a younger sibling with diarrhea. The demands of caring for multiple children negatively impacted caretakers’ ability to provide appropriate and timely diarrheal treatment for the ill child. This may contribute to low use of ORS packets for sick child in families with multiple children. Our study also found that the younger children were less likely to use ORS in HBC. Due to the fact that younger children were more likely to be breastfed or fed with more liquid food, caretakers would not like to think it necessary to use ORS frequently in their children with diarrhea in such a feeding period [18]. In addition, a similar study of ORS therapy in rural Bangladesh demonstrated that the mothers generally had the perception that infants should not drink any fluids other than breast milk before this age, and the infants were introduced to water and other clear fluids after this age [19]. In rural China, such a perception among the caretakers in home care may also be a factor, explaining why children less than 12 months old were offered ORS less frequently in HBC than those aged 25–36 months. So health communications should specifically inform caregivers that ORS can be used in sick children who are currently being breastfed.

The utilization rate of ORS in VLC (41.89%) was only lower than that in TLC (45.77%). Recent study of prescriptions of village doctors in these areas has shown that in the village clinics more than one-third of the doctors had no full-time medical education and village doctors were inclined to adopt inappropriate drug utilizations in the treatment of diarrhea [12, 13]. More educational or training projects about appropriate and early ORS use by the government should be carried out in village-level medical sectors urgently. GEE model analysis of ORS use in VLC showed that ORS use was positively associated with the habit of drinking boiled water often in the families. Families that drank boiled water often at home may have more faith in ORS as a treatment or consider it standard treatment of childhood diarrhea. When they took their children to village clinics, they also agreed that village doctors could use ORS for their children with diarrhea. Our study also showed that receiving educational materials about childhood diseases was more likely to increase the likelihood of ORS use in VLC. The prescription of ORS in village clinics seemed to meet with the profile of educational materials about the treatment of childhood diarrhea and thus made the caretakers more likely to believe in the decision of village doctors to use ORS. Our study also found that ORS available in village clinics was more likely to increase the use of ORS, as found by a study of ORS use at home [20]. Thus, when ORS packets were not available in village clinics, self-made ORS based on the WHO formulation could be used to treat children with diarrhea.

In conclusion, the utility rate of ORS in HBC was the lowest in all sought care. So, ORS promoting program should give the highest priority to home care. The number of pre-school-aged children in a family and the children’s ages affected ORS use in HBC. The families’ habit of drinking boiled water, educational materials about childhood diseases and the availability of ORS in village clinics affected ORS use in VLC. ORS promoting strategies for low-level care could be strengthened based on children characteristics, the habit of drinking water and the situation of receiving educational material in the families and on the availability of ORS packages in village clinics in rural Western China.