Maternal healthcare financing: Gujarat's Chiranjeevi Scheme and its beneficiaries.

Maternal mortality is an important public-health issue in India, specifically in Gujarat. Contributing factors are the Government's inability to operationalize the First Referral Units and to provide an adequate level of skilled birth attendants, especially to the poor. In response, the Gujarat state has developed a unique public-private partnership called the Chiranjeevi Scheme. This scheme focuses on institutional delivery, specifically emergency obstetric care for the poor. The objective of the study was to explore the targeting of the scheme, its coverage, and socioeconomic profile of the beneficiaries and to assess financial protection offered by the scheme, if any, in Dahod, one of the initial pilot districts of Gujarat. A household-level survey of beneficiaries (n=262) and non-users (n=394) indicated that the scheme is well-targeted to the poor but many poor people do not use the services. The beneficiaries saved more than Rs 3000 (US$ 75) in delivery-related expenses and were generally satisfied with the scheme. The study provided insights on how to improve the scheme further. Such a financing scheme could be replicated in other states and countries to address the cost barrier, especially in areas where high numbers of private specialists are available.


INTRODUCTION
It is estimated that about 1.2 million children in the Gujarat state are born each year, including both institutional and domiciliary deliveries. Given that the maternal mortality rate for the state is estimated at 172 per 100,000 livebirths (1), an estimated 2,064 of these mothers die from maternal causes. The primary reason for these maternal deaths is that the majority of deliveries are not attended by skilled persons, women do not have access to emergency obstetric care (EmOC), and there is little postnatal follow-up. It is argued that most of these maternal deaths are avoidable with adequate interventions, such as skilled birth attendant (SBA), referral services, and access to EmOC (2)(3)(4)(5). Families below the poverty-line (BPL) are the most vulnerable since they face significant risk due to their poor socioeconomic status and limited access to healthcare services (6).
The Chiranjeevi Scheme, implemented by the Government of Gujarat, aims at encouraging the BPL families to access institutional delivery at a private hospital. This is done by providing financial protection to these families and covering their out-of-pocket costs incurred on travel to reach the healthcare facility. The scheme also provides financial support to the accompanying person for loss of wages. The scheme uses several mechanisms to target the BPL family, the main mechanism being the BPL card. This card is issued to families earning less than a particular level of income and certain asset ownership criteria. The BPL card helps identify this group of population for provision of various benefits and to target the benefits.
The Chiranjeevi Scheme was launched as a one-year pilot project in December 2005 in five backward districts: Banaskantha, Dahod, Kutch, Panchmahals, and Sabarkantha (7). The scheme has now been extended to the entire state. When the scheme was initiated, the pilot districts were selected based on remoteness and included regions with the highest rate of infant mortality. The private medical practitioners (mainly obstetricians) in these districts were empanelled in the scheme to provide delivery-care services to BPL women. These care providers are re-imbursed on a fixed rate for deliveries carried out by them (8). Details of the financial package for the Chiranjeevi Scheme are given in Table 1. The objective of the study was to explore the targeting of the scheme, its coverage, and socioeconomic profile of the beneficiaries and to assess financial protection offered by the scheme, if any, in Dahod, one of the initial pilot districts of Gujarat.

MATERIALS AND METHODS
The Chiranjeevi Scheme was implemented on a pilot basis in five districts, including Dahod, starting in December 2005.
The scheme involved creating a panel of private care providers who would accept referrals by the families covered under the scheme (9). Identification and empanelment of the private obstetricians were done by the Block Health Officer (BHO). After the private practitioner agrees to join the scheme, a Memorandum of Understanding is signed between him/her and the district health authorities. The District Project Management Unit (DPMU) for the Reproductive and Child Health (RCH) programme handles all documentations for the scheme. The DPMU is also responsible for reporting progress of the scheme to the State Health Directorate and for making payments to the empanelled obstetricians through the District Development Officer (DDO) and Chief District Health Officer (CDHO). A flow diagram of implementation of the scheme is described in Figure 1.
The medical officers and the Auxiliary Nurse Midwife (ANM) of the respective Subcentres undertake the responsibility of motivating the community (BPL families) to take the benefit from the scheme. The ANM visits the communities and registers a pregnant woman for receiving antenatal care (ANC); at that time, she motivates women to avail of the Chiranjeevi Scheme. She encourages the pregnant woman to visit an empanelled obstetrician. Every month, the empanelled providers present their filled-in vouchers for deliveries conducted and claim their reimbursement. The entire document for the reimbursement is submitted at DPMU; the DPMU initiates the process of payment. Payment is made after approval of CDHO and DDO.

Selection of sample
All five pilot districts were put in three groups based on their geographical proximity to each other. Group 1 was selected as it had the largest number of deliveries among the three groups. Within Group 1, Dahod was selected as it had the higher average number of deliveries per care provider (Table 2). Group 1 is also more backward, has more tribal population, and geographically in the east of the state. Group 2 districts are in the north of the state,

District profile
The population of Dahod was 1,751,000 in 2006 when the study was implemented. The BPL population of the district was about 23% (11). The total number of deliveries in Dahod district was about 41,500 during 2006 assuming a crude birth rate of 23.7 (12) and even distribution of births over all economic groups. Using this as the basis, the estimated number of BPL deliveries was about 9,545 per annum. During 2006, 7,735 deliveries were conducted under the Chiranjeevi Scheme. Hence, the number of deliveries under Chiranjeevi accounts for about 81% of all estimated deliveries by BPL women. This is based on the assumption that the BPL cards were issued only to the targeted population. The percentage of deliveries under Chiranjeevi Scheme was about 18.6 of all deliveries in the district.
The total number of deliveries conducted in the district between December 2005 (when the scheme started) and March 2007 was 9,854, of which 7,584 were normal, 391 by caesarean section (lower segment caesarean section), and 1,879 complicated. Month-wise data are presented in Table 3. On average, about 615 deliveries were conducted under the Chiranjeevi scheme in Dahod district per month.
Within Dahod district, a multi-stage hierarchical cluster-sampling procedure was adopted to select households for data collection (Fig. 2). Data for the number of deliveries under the Chiranjeevi Scheme conducted in all seven talukas (subdistrict) of the district from January to December 2006 were obtained from the DHO of the district. In the first stage of sampling, talukas of the district were classified as low, moderate, and high based on the number of deliveries under the Chiranjeevi Scheme. Three clusters of talukas were, thus, formed, and one taluka from each cluster was randomly picked.
In the second stage, villages falling under these three talukas were further classified into three clusters (low, moderate, and high) based on the number of deliveries conducted under the Chiranjeevi Scheme. Sample households were randomly selected in proportion to deliveries from the three clusters. The list of the households was obtained from the ANM/FHW of the respective areas. Listing of pregnant women with ANM/FHW is quite comprehensive and is the most exhaustive list available for sampling purpose. Two types of samples were selected: (a) beneficiaries of the scheme defined as mothers who delivered in private obstetrician clinics and benefited from the scheme (Chiranjeevi beneficiaries-CB), and (b) non-beneficiary mothers of the Chiranjeevi Scheme (NCM) defined as poor women (BPL) who did not benefit from the scheme. The sample size for the Chiranjeevi Scheme beneficiaries was fixed at around 250, given the available resources. As an estimated 20% of the BPL deliveries are taking place outside the Chiranjeevi Scheme, it was decided to oversample the NCM group. The final NCM sample was 394.

Questionnaire development
Initially, a pilot questionnaire was prepared which was later revised based on the field-testing. The questionnaire was divided into 10 sections, of which the first five, covering background characteristics, were the same for both CB and NCM. Section 6 (obstetric history of recent delivery in the Chiranjeevi Scheme), 7 (details of neonatal care of the recent Chiranjeevi delivers), and 8 (knowledge and practice of Chiranjeevi Scheme) were administered only to CB clients. For CB clients, information was collected for both index and previous deliveries whereas, in the case of NCM, only information on their index (most recent) delivery was collected.
All the field investigators had prior experience in Random selection of taluka from each cluster CYD: Chiranjeevi delivery conducting field interviews in rural settings in the district-level household surveys. They were trained for two days, of which the first day was on the structure of the questionnaire and the second day was for conducting mock interviews. In all, there were four investigators conducting about 14-16 interviews per day. The whole exercise of field survey took about two months. There was one field editor and one supervisor whose jobs were to validate the questionnaire in the field and check for logical inconsistencies.

RESULTS
A summary of the comparisons of CB and NCM on background variables is provided in Table 4.

Sociodemographic profile
Age at marriage and parity: The mean age at mar-riage for the CB and non-Chiranjeevi mothers was 17.98 years and 18.12 years respectively. The average number of children born to CB was 2.53 compared to 2.84 for the NCM group. This difference is significant at 5%.

Targeting based on income
In our study, 94% of the CB were having annual income below Rs 12,000 (US$ 300). This shows that the large majority of the beneficiaries are poor, and only 6% of the beneficiaries are non-poor. This also 10.30 26.14 -0.538 *Significant at 5%; **Significant at 10% shows a high level of targeting of the scheme to the poor. As the NCM were also selected from the category of the poor, our study showed that 96% of them had annual income below Rs 12,000.

Expenditure on delivery
Chiranjeevi beneficiaries are not supposed to bear any expense relating to delivery, as the Government pays all the costs of obstetrician directly, and they are to pay the women for transport and funds to the accompanying person. To assess this aspect of the scheme, the survey collected information on expenditure incurred by the CB and the NCM group. These expenses were further analyzed in terms of expenditure incurred on self and child's medicines, transportation costs, and other out-ofpocket expenditure.
To understand the financial protection/benefit re-ceived by the CB, we have compared the expenditure by the CB in the previous delivery and the expenditure by the NCM in their present (most recent) deliveries for a normal or complicated delivery at a private facility. This is to make them comparable as deliveries under Chiranjeevi exclusively take place in private facilities ( Table 5). The CB, an average, incurred an expenditure of Rs 3,070 on their previous institutional delivery; this is before the scheme started. During their index delivery, 80% of the Chiranjeevi clients indicated spending, on average, Rs 655 on buying medicines for the neonate and self: Rs 296 for self and Rs 358 for the neonate. The out-of-pocket additional expenditure for CB on medicines is about 36% of the cost of the Chiranjeevi package, i.e. Rs 1,795 paid to the care provider per case basis. The Chiranjeevi beneficiaries are getting Rs 200 for transportation from the service provider as per the scheme benefit. They reported spending, on average, Rs 272, thus incurring an additional out-of-pocket expenditure of Rs 72. After including the transportation cost, the to-tal out-of-pocket expenditure was Rs 727 for CBs which is on top of the government expenditure of Rs 1,795 ( Table 5).
The NCM spent, on average, Rs 2,319 for a normal delivery. This amount includes consultancy and procedure fees of Rs 1,102 paid to the doctor, transportation cost of Rs 276, and medicine cost of Rs 331. In most recent complicated deliveries, the NCM spent Rs 13,524 (detailed break up is given in Table 5). The weighted average for normal and complicated delivery cost for NCM came to be Rs 4,000. Table 5 shows that the amount saved by the CB by availing of the benefit of the scheme is estimated at Rs 3,273 (Rs 4,000 [total expenditure of NCM] minus Rs 727, additional expenditure of CB). As per the design of the Chiranjeevi package, there is a provision of Rs 100 for providing diet to clients during hospitalization. Our survey showed that only 8.8% of the Chiranjeevi clients were provided diet during their stay in the healthcare facility.

Antenatal care
Of all the clients who took benefit of the Chiranjeevi Scheme, 96% had gone for ANC services. The average number of ANC visits made by the clients was 2.84. The ANMs provided ANC services to 61% of the clients-16% were provided services by private doctors, and 2% were provided services by government doctors. About 17% of the clients received ANC services from more than one source.
Almost 49% of the CB reported antenatal problems compared to 53% in the NCM group. Of those who had antenatal problems, 71% went to private and 16% to a government hospital. Only 2% of the clients received intervention for ANC complications at home while 3% did not seek intervention for

Delivery services
Although all deliveries under the Chiranjeevi Scheme are supposed to be in the hospital of private empanelled doctors, only one delivery (0.38%) of the CB group was conducted in the home, and 2.7% of deliveries were conducted in a government institution. All other (97%) deliveries were conducted in a private health facility. As far as the NCM is concerned, 21% of deliveries were conducted in the home, 1.8% in a government institution, and about 77% in private institutions.
About 94% of deliveries conducted in the CB group were normal; in the case of the NCM group, this figure was 97%. There was some indication that more complicated deliveries were going to the Chiranjeevi Scheme.

Service providers
The percentage of deliveries conducted by private doctors was 41 in the case of CB and 32 in the case of NCM deliveries. Nurses and other trained attendants at a private facility conducted deliveries in 48% of the CB cases. This percentage for the NCM group was 43. The key difference here lies in the deliveries conducted by trained and untrained attendants. In the case of CB, TBAs conducted 1% of deliveries whereas, in the case of NCM, TBAs conducted 20% of deliveries.

Postnatal care
Data showed that about 28% of the CB went for postnatal care (PNC); the corresponding percentage for the NCM group was around 31%.

Awareness about the Chiranjeevi Scheme
Awareness generation: Of those clients who took advantage of the scheme CCB group), 55% were informed about the Chiranjeevi Scheme by the ANM/ FHW and 17% by the Anganwadi workers (AWWcommunity-level nutrition worker). Public-health facilities, including the Subcentre, Primary Health Centres (PHCs), Community Health Centres (CHCs), and the district hospital, were the sources of information for only 6% of clients. Friends/ neighbours were the sources of information for 4% in the CB group. Printed material and pamphlets were the sources of information for only 1% of clients. TBAs informed only 1% of the beneficiaries about the scheme. Others, such as panchayats (governing bodies at the village level) members, balwadi teachers, doctors, and nurses provided information to 6% of the CB. The questionnaire did not seek this information from the NCMs.
BPL cards: All clients who took the benefit of the scheme were aware about the requirement of the BPL card/certificate for availing of services. All beneficiaries mentioned that they possessed a BPL card. Also, 98% of the beneficiaries reported that their BPL cards were inspected at the healthcare facility prior to availing of services under the scheme.

Satisfaction of clients
About 89% of the CB and 87% of the NCMs were satisfied with services provided at the health facilities. The main reason for satisfaction of 66% of the Chiranjeevi beneficiaries was 'good quality of services' while the main reason for satisfaction of about 18% of the CB was 'good facilities.' The reasons for satisfaction of 7% and 5% of the CB were, respectively, for 'good behaviour of staff' and 'prompt services'.
Eighty-six percent of the Chiranjeevi beneficiaries reported that the doctor was available when they reached the facility; only 1% reported that the doctor was not available; and the remaining beneficiaries stated they did not remember. In the case of the NCM, the doctor was available for 86% of the cases when they reached the facility.
Sixty percent of the CB group found that medicines were always available, and 30% reported that medicines were available most of the time. In both CB and NCM cases, 87% of the respondents expressed that the staff was courteous.
The respondents who used the Chiranjeevi Scheme were asked to provide suggestions for improving the scheme. The availability of medicines was identified as one important factor to improve services. Around a quarter of the clients suggested providing medicines to the beneficiaries under the Chiranjeevi Scheme to improve it. About 12% of the clients suggested payment of increased compensation for transportation (Rs 200). Around 5% of the beneficiaries also reported that the transportation expenses are not being provided to them.
Only about 4% of the beneficiaries reported that the nurses at the health facility asked for money, and this should be addressed. Another suggestion was to improve proximity of the community to the healthcare facility.

Targeting
Possession of the BPL cards is the criteria for selection of beneficiaries for the Chiranjeevi Scheme, and all Chiranjeevi beneficiaries had a BPL card. Ninety-four of the CB have an annual income below Rs 12,000 which comes to Rs 32.90 a day, less than the World Bank poverty-level of a dollar a day (equivalent to Rs 40). This suggests that the scheme achieved its objective of targeting the poor. Leakages or benefits to the non-poor were limited (about 6%). The NCMs were also equally poor but did not avail of the benefits of the scheme. In this study, the non-beneficiaries were also selected from the population eligible for the scheme, which are the people living below the poverty-line.

Role of health workers
It can be inferred from the data that the ANM/ FHW and the AWW were the most common source of providing information and building awareness about the Chiranjeevi Scheme among the beneficiaries. It should also be noted that all these health workers have acted as an essential 'link' between the healthcare-delivery system and the beneficiaries. The health workers have not only made the services available to the beneficiaries but also guided them on how to access these services.
The decision for choosing the place of delivery for the majority of the Chiranjeevi clients has also been taken by the ANM. The health workers are either from the community itself or are well-known to the community, and therefore, the community places a lot of faith in them. The health workers can be developed as a more important link in the healthcare-delivery system even if actual services are provided by private doctors.

Quality of care and satisfaction of clients
It is heartening to note that most clients of the Chiranjeevi Scheme and non-clients were quite satisfied with delivery-related services. They also reported positive behaviour from the service provider and the staff. This shows that the scheme is able to provide client-pleasing services at almost half of the cost of the regular private-sector charges. Further research should address the technical quality of care.

Improved healthcare-seeking behaviour
This scheme encourages poor women to deliver in a healthcare facility; for many, it is likely that they have accessed health services at an institution for the first time. Given their high level of satisfaction, they are likely to use services in the future for themselves and their children as well as recommend to others.

Summary and conclusion
The Chiranjeevi Scheme has provided financial protection against the cost of delivery and EmOC to the marginalized section of the population. In the study, it was seen that a Chiranjeevi client saves around Rs 3,273 (about US$ 82) in delivery compared to those who did not avail of the benefits of the scheme. However, the scheme is not 100% cost-free to the BPL families as they had to pay outof-pocket expenses for medicines and transportation. The Government spends Rs 1,795 (US$ 45) for each delivery. Thus, by buying the services in bulk from private care providers, the Government is getting delivery services at a much lower price than the market rates being paid by non-beneficiaries (Rs 4,000 or US$ 100). The findings suggest that the scheme needs to be strengthened by improving some aspects, including more funds for medicines, transportation, etc. and offering at least two antenatal and two postnatal visits.
The monitoring of the scheme needs to be improved in several ways. The client should not be made to pay any extra money; the doctor should maintain proper records and follow standard evidence-based protocols. Deaths of mothers and children need to be systematically documented and analyzed. Referrals made by the private care providers should be analyzed to ensure that they are not doing so to avoid more expensive treatments. Future studies are required to assess the technical quality of care and mortality impact of the programme. Women not using the services yet should be motivated to take benefit of the scheme. The private care providers should collaborate with institutions to develop an appropriate costing framework to implement this scheme and develop an understanding of whether all components of the programme are incorporated adequately. Unfortunately, private care providers are often reluctant to share their cost information which is vital to develop adequate understanding of the pricing issues in such a scheme.
The findings of the study suggest that, given the demographic characteristics of this district and the economic profile of the clients using this scheme, wrong targeting is not a major issue. About 94% of the Chiranjeevi clients earn much less than a dollar a day. Further study is needed to see as to why many of the poor who should be covered are left out of the scheme.
The government health employees, such as ANMs/ FHWs, have been found to be effective in building awareness and guiding clients to use private services. Their role in the process has been found to be quite important and needs to be strengthened.
Overall, the study has shown that the scheme reaches the poor and provides substantial benefits to them. Based on the learning from this scheme, the Government of India and other developing countries can think of replicating this scheme on larger scale. The Government of Gujarat should consider expanding the scheme and its coverage to people who are just above the poverty-line.