Health service utilization: In MP, utilization of private allopathic care was highest at 74% with 12% seeking care from government health facilities (Table I). In UP, majority (71%) went to UCPs’ with only 5% not seeking care for their child. Mothers in TN predominantly sought care from private allopathic doctors (75%). Utilization of government care was higher compared to the other two states at 19.6, but no mother reported going to a UCP. Data by district is presented in Web Table I. More than half the sample of mothers from each state (59%-MP, 70%-UP, 80%-TN) sought care for their child within 24 hours of symptom presentation (Table II). As compared to UP where 70% went to UCPs, in MP and TN, private allopathic care was the preferred choice (75% and 74%, respectively). Going to government health facilities was highest in TN at 18.9%. Data by districts is presented in Web Table II).

Factors influencing health service utilization: Cultural practices, either personal, or due to familial pressure, in TN, included exposing the child to incense fumes (Sambrani), feeding a concoction made from Tulsi (Basil), application of Karpuravalli Thaillam (oil extracted from a medicinal herb), seeking the blessings of priests and tying a sacred thread around the babies wrist or waist. However, these mothers simultaneously sought care from qualified allopathic doctors. Beliefs in cultural and traditional practices were more prevalent in MP and UP, with mothers resorting to home remedies like use of mustard oil, Ajwain (carom seeds), hing (Asfoetida), haldi (Turmeric), for treating cough or cold. Oil massages using mustard oil and barasingha (a piece of deer horn that is finely ground) were believed to be effective in treatment of chest in-drawing Jhaad phoonk (a type of exorcism) was seen to protect the baby against the evil eye. These were usually the first steps taken by mothers when their child fell ill. If this failed, care was sought from a care provider.

Decision making: Mothers in TN reported having higher autonomy and decision-making capacity. They had the support of their in-laws and elders in the family who encouraged them to seek appropriate care for their child and would even accompany them to the hospital if required. Instances of joint decision making with the husband were also reported. On the other hand, decisions regarding care seeking in MP and UP were mostly made by family elders or husbands, with mothers usually acquiescing to such decisions. In nuclear families decisions were either made jointly by husband and wife or else only by the husband.

Enabling factors: The health system infrastructure in TN, both government and private allopathic sector, is well developed and fairly equitably distributed across the districts (Web Table III). Added to this, the presence of private care facilities provides rural folk with an alternative choice. In contrast, the numbers of primary health centers (PHCs) and community health centers (CHCs) in MP and UP are distinctly inadequate for their large populations. Besides Bhopal in MP and Faizabad in UP, the number of government hospitals are exceedingly low in these states. Private allopathic healthcare facilities are also few. The presence of UCPs was ubiquitous in these districts, although we do not have any reliable data on their numbers.

Our qualitative interviews with mothers revealed that in TN, both government and private care facilities were equally accessible. Preferences for private care were clearly evident with families switching between doctors depending on how well the child responded to treatment. Doctors in the private sector were believed to be more effective, easily accessible and available till late in the evenings. They also administered injections, believed to bring about rapid cures. Further, doctors and paramedical staff in government hospitals were perceived as unfriendly providing unsatisfactory answers to queries unlike in the private sector. Despite this preference for doctors in the private sector, several mothers gave positive feedback regarding care provided in government hospitals. They described it as being affordable, accessible, of good quality and comparable to that of private facilities. Others spoke of the CHWs who made home visits and provided advice regarding the health of their child. Use of government health facilities for seeking care for children was more evident in TN as compared to the states of MP and UP.

In MP and UP, need for travelling long distances to access care in government health services, coupled with unavailability of doctors in these facilities, acted as major deterrents to care seeking. Connectivity was particularly poor in Shravasti (UP). Seeking care from the jhola chaap (UCP) was common in UP as compared to MP, where they were easily accessible, were cheaper than private doctors, made home visits and usually dispensed allopathic medicines. In MP, preferences for seeking care from private care providers dominated as mothers considered the money well-spent. However, we were unable to ascertain if these private care providers were qualified or unqualified. The belief that government facilities were lacking in cleanliness and competent doctors, involved long waiting time and had inadequate supply of drugs, added to the general negative opinion. Mothers in UP said that even the 24-hour government facilities did not have doctors, thereby defeating the purpose for which they were set up.

Perceived need: Regarding the ‘need factor’, we found that in TN mothers were unfamiliar with the term pneumonia and unaware of its cause or presenting symptoms. Although symptoms of fever, cough and cold were well understood, mothers rarely reported seeing cases of pneumonia. Only those who had sought care for treatment of pneumonia for their child or whose child had died following pneumonia had better awareness about the condition. In MP and UP, some mothers were reasonably aware of pneumonia and described a range of symptoms. Others spoke of the importance of vaccination and cleanliness as protection against pneumonia, indicating satisfactory awareness about the disease. If the child’s cold and cough was perceived to be very heavy, it was referred to as ‘double pneumonia’. Some, while unfamiliar with the term, were nevertheless aware of symptoms like chest indrawing and appreciated the need to seek care for its treatment. Others subscribed to the traditional belief that children were at risk for contracting pneumonia if they had "cold in their bodies" as compared to "heat". Difficulties in recognizing severity of illness were also expressed.

Evaluated need: In TN, the HCPs stated that though awareness about pneumonia was poor in the community, the ability to recognize symptoms of respiratory distress in the child was adequate, which influenced timely and appropriate care seeking. They believed that there was not much delay in care seeking among families and as a back-up, families kept paracetamol syrup and nasal drops at home for use in case of such symptoms occurring. Doctors in the government sector appreciated the sustained health literacy efforts provided by CHWs during antenatal visits. They also credited government run school health programs for increasing health literacy among mothers. Although there was a trend of preferring private over government facilities, especially during an emergency for reasons of faster accessibility, for regular care, families would go to government facilities. According to the CHWs, negative beliefs about the poor quality of health care and long waiting time in the government hospitals influenced many to seek care from the private sector, even at great financial cost. Seeking care for their child from UCPs however, was not reported. In MP, care providers in the government sector, felt that awareness about pneumonia in the community was good perhaps due to its high prevalence. They felt that it was rare for families not to seek care however, the type of care sought was not always appropriate and often delayed due to use of home remedies and magico-religious practices. The CHWs said that many poor families chose to seek care from UCPs who were easily accessible, dispensed allopathic medicines and gave injections. Physicians, both public and private, agreed that there was a preference for private over government care because of the distances involved in accessing these facilities and because doctors in government hospitals were not always available. They also said that decision making concerning care-seeking for children remained with the elders of the household or with the woman’s husband. The HCPs in UP felt that, awareness about pneumonia was poor and. care was sought only when symptoms became serious. Resorting to home remedies was usually the first step. The CHWs said that the easy availability of UCPs combined with the faith people had in them influenced people’s preferences for them. They also spoke of people’s preferences for private care as against government care as it was believed to be better and more prompt. In addition, women’s dependence on their husbands or elder members in their household to make decisions on care-seeking contributed to delays in care- seeking.