Background
India’s slow progress in reducing maternal mortality not only hinders its own goals but also affects the global achievement of MDG 5 as India accounts for about one fourth of the maternal deaths worldwide [
1]. The five most common causes of maternal mortality in India are direct obstetric complications which can be well managed by provision of emergency obstetric care (EmOC) services. The Indian government subscribed to the EmOC provision strategy and in 2005 included access to 24 hour EmOC as one of the ‘service guarantees’ of its public health programme-the National Rural Health Mission (NRHM). The mission launched a safe motherhood programme titled Janani Suraksha Yojana (JSY) which provides cash incentives to mothers conditional to their delivering in health institutions [
2]. The programme logic is increased access to institutional care during delivery would decrease delays in accessing EmOC services if required and thus save lives.
Another challenge in India is health workforce shortages and difficulties in recruiting and retaining them especially in rural and remote areas as in most low and middle income countries (LMICs). The Indian census data shows health worker density about half of the WHO benchmark of 25.4 such workers associated with achieving 80% institutional deliveries attended by skilled personnel in cross country comparisons [
3]. The Indian public health system suffers from a severe shortage of specialists to deliver EmOC- 37% positions of obstetricians are vacant and moreover compared to existing requirements there is a shortfall of 55% obstetricians in rural India [
4]. Contrastingly, the private sector share in services has grown significantly, despite the high expenses involved.
The NRHM also set the Indian Public Health Standards (IPHS) [
5] for service provision. Along with other improvements in the infrastructure, achieving these standards would require a large increase in numbers and skills of human resources [
6]. Thus India sees a phenomenal increase in the number of interventions and strategies towards this, some specific to states while others nationwide. Some alternatives being explored are task shifting for anaesthesia and obstetric services, decentralising recruitment to district level and altered processes like holding walk in interviews, posting junior residents at CHCs for fixed period, providing financial and career prospect incentives [
6]. The extent of success of each initiative to address skilled personnel shortages needs to be viewed against the contexts that influence them. For example some initiatives face legal issues with licensing of providers raised through alternate arrangements, performance rewards, career pathways and job clarity [
7]. However as noted by a recent overview of systematic reviews there are few studies of the potential of the various initiatives and their effect on the size and distribution of EmOC providers and the related service uptake, specifically from LMICs [
8].
With NRHM's strategy to promote public private partnerships, there have been numerous initiatives in various areas in health care; however caution is suggested in execution of these to achieve efficiency and equity [
9]. A recent popular public private partnership scheme focusing on EmOC in India is the Chiranjeevi Yojana [
10] in Gujarat state which is a contracting out model with voucher based financing, whereas other states have contracting in. The private sector though large is concentrated in urban areas. Also feasibility of fruitful partnerships with unregulated private sector is questioned, for instance a study of private maternity homes in Maharashtra state signals possible compromise in service quality with poor standards in private sector [
11].
The NRHM promotes contracting in of specialists for EmOC services and has made financial provisions for the same under two of its prominent initiatives – the JSY and the IPHS. The JSY is the world’s largest conditional cash transfer (CCT) programme in health in terms of number of beneficiaries. While public spending on JSY has been increasing since its launch in 2005-06 [
12] a steep rise in institutional births from 40.7% in 2005-06 [
13] to 68% in 2008-09 [
14] has been concomitantly observed. However studies do not associate this spurt of institutional deliveries with a decrease in maternal mortality [
15].
While the JSY has been studied widely, to the best of our knowledge, there is no report specifically on the contracting in component for EmOC provision in the context of the JSY as also of the IPHS. This paper attempts to fill this gap by reporting on the implementation of contracting in and its effect on EmOC availability in an Indian state. The findings should provide insights to policy makers on the design and implementation of contracting in for EmOC in resource constrained settings.
Methods
This was an exploratory study. A mix of quantitative and qualitative methods was used including a facility based cross sectional survey and provider interviews.
Study settings
This study was conducted in Maharashtra state in western part of India. It has a higher level of economic growth compared to the national average, however in terms of Human Development Index (HDI) that reflects the social determinants of health; it ranked eleventh in 2006 with a HDI value of 0.689. Maharashtra has 35 districts comprising 358 blocks. The state has better health indicators than the national; maternal mortality ratio is 104 and infant mortality rate is 31 against 212 and 50 for India respectively. The public health system in the state is three tiered as in the rest of the country having tertiary care centres at districts, secondary care centres at sub district/block level and primary health care centres within blocks. Maharashtra has highest number of private medical colleges in the country and health care sector is dominated by private providers.
Selection of study districts
Three districts were selected on the basis of heterogeneity in their socio-demographic profiles as reflected in Human Development Report Maharashtra 2002 [
16] and assuring state wide geographic representation. The study districts thus selected were Nandurbar (low HDI), Amravati (Moderate HDI) and Satara (better HDI) using the latest data available at the time of selection. While Nandurbar is a tribal district with high Schedule Tribe (ST) population, Amravati has a high population of poor people i.e. those living Below the Poverty Line (BPL) and belonging to Scheduled Castes (SC). On the other hand, Satara has a relatively high uptake for institutional births (Table
1).
Table 1
Socio demographic indicators of study districts
| Total | Urban(%) | *SC(%) | *ST (%) | *BPL(%) | Rural (%) | Total (%) | |
Satara | 28,08,994 | 14.2 | 9.5 | 34.7 | 43 | 85.6 | 87.4 | 10 |
Amravati | 26,07,160 | 34.5 | 17.5 | 14.4 | 66 | 50.5 | 63.6 | 15 |
Nandurbar | 13,11,709 | 15.4 | 3.2 | 65.5 | 54.5 | 16.5 | 25.3 | 32 |
Maharashtra | 9,68,78,627 | 42.4 | 10.2 | 8.8 | 30.7 | 54 | 63.6 | |
Selection of health facilities
We studied all the public secondary and tertiary care centres i.e. district hospitals (DH), sub district hospitals (SDH) and community health centres (CHCs) in the selected three districts. Thus 44 health facilities in the study districts- 3 DHs (250 bedded), 8 SDHs (50- 100 bedded), and 33 CHCs (30 bedded) were included. (Table
2) All the study centres are expected to provide Comprehensive EmOC (CEmOC).
Table 2
Public health facilities surveyed
Satara | 1 | 2 | 15 | 18 |
Amravati | 1 | 4 | 9 | 14 |
Nandurbar | 1 | 2 | 9 | 12 |
Total | 3 | 8 | 33 | 44 |
Survey of health facilities
We surveyed each health facility using a modified health facility assessment form based on the facility survey for EmOC designed by the UN agencies and the Averting Maternal Death and Disability programme [
17]. Information on availability of skilled manpower for provisioning of EmOC services was obtained and detailed information was gathered on implementation of contracting in model for private EmOC specialist. Also data on performance of obstetric services by facility in six months prior to survey were retrieved from records at facilities. Study team members with medical training conducted the survey. The data collection at health facilities was undertaken intermittently during December 09 to August 2010. Each facility was visited once without prior notice. In this paper we report data specific to contracting in for EmOC.
Interviews with providers and managers
Interviews were conducted with private EmOC specialists (n=15) either contacted in or located in vicinity of facilities with potential for contacting in and with medical superintendents and public EmOC specialists (n= 20) at selected facilities. Interviews were also conducted with district health officials, and district and block programme managers (n=7).
Mapping of obstetricians
We obtained a list of all practising private obstetricians in the three study districts from the district hospitals. We confirmed this list with practising obstetricians and also used snowballing to complete it. We also approached members of the professional association of obstetricians and gynaecologists at the study districts for validation of the list.
Data collection procedures
Informed written consent was obtained from all study participants. All interviews were audio taped and transcribed verbatim for analysis.
Data analysis
Quantitative data was entered and analysed in Excel. Qualitative data was analysed using thematic content analysis using deductive coding.
State support
We obtained formal bureaucratic support for the study from the State Health Systems Resource Centre, National Rural Health Mission (NRHM) Division, Government of Maharashtra.
Ethics approval
Approval for the study was granted by the Institutional Ethics Committee of the Foundation for Research in Community Health, Pune,India.
Discussion
This being the very first study to our best knowledge that reports on issues of contracting in specialists in the context of the renewed focus on EmOC in India, it is exploratory in nature. The exploratory nature of our study and its focus on rural areas need to be considered while interpreting the findings. Also it needs to be kept in mind that contracting in is an evolving and dynamic process. We discuss the findings in three sub sections below.
Contracting in for EmOC in rural contexts
Although existing resource constraints make pooling of capabilities and resources necessary, the strategic choice of contracting in specialists seems rational only in selected regions- for instance in regions characterised by shortages of public and concentration of private specialists, as in two of the districts in this study while it proves unfeasible in the other district. This indicates that no single measure could be applicable to all regions and that area specific issues are important influencing factors. There need to be locally relevant strategies to provide EmOC services. A district like Satara with relatively higher concentration and density of private specialists offers more opportunities for contracting in and broad bargaining zone that could be explored by engaging in dialogue with the private sector while another district like Nandurbar in the state has a dearth of specialists even in private sector rendering contracting in unfeasible. For instance of regional variations in strategies the Gujarat experiment of harnessing private sector capacity through the Chiranjeevi scheme which is by contracting out services is greatly divergent from the Tamil Nadu experience with strengthening human resources within the Indian system and by contracting in providers [
18]. However the root cause- the shortages of specialists in public service and the long term sustainability of such arrangements needs to be considered. Our findings suggest that local circumstances will dictate balance between introduction or expansion of contracts with private sector and strengthening public provisions and that neither of these disregard the need to improve public systems. These findings corroborate conclusions from a multi country study by Mills [
19] and additionally suggest that these hold for sub national levels too. A review of literature on contracting with the private sector for primary care in LMICs concludes that theories from new institutional economics and evidence about nature of existing contracts for primary care in the United Kingdom question the viability of a policy of competitive contracting in in the context of LMICs [
20]. Also alternative arrangements like task shifting that involves delegation of tasks to lower level cadre have shown promising results in the developing world [
21]. For instance the post operative outcomes of C sections performed by non physician providers in Malawi were comparable to those of doctors [
22]. However the only available ,though early, evaluations of task shifting for EmOC-for both C section and anaesthesia in India are not encouraging and demand greater support [
23,
24]. The choice revolves around what is the most effective way to improve coverage, reach and quality of EmOC services and reduce its costs in the specific contexts. Also it is important to recognise that contracting in for EmOC is additionally complex than for other clinical and diagnostic services, which is gaining popularity in developing countries like India, as time to response is crucial in saving lives in the former.
Public provisioning for contracting in
The overlapping provision for contracting in under JSY and under the IPHS points to the isolation of policy making from implementation channels. There appear to be multiple provisions for a poor woman eligible for the JSY requiring EmOC- while IPHS spends to provide these services, the JSY allocates to subsidise her costs in private hospitals in addition to routine outlays for functioning of public health facilities. However despite these measures the spending and bankruptcy of poor households to access these services continue.
In planning a public health programme care needs to be taken to ensure resources are not duplicated. This is not to deny the need for interventions but to make the right choice for the larger good. The Accountability for Reasonableness Framework advocated by Mitton and Donaldson et al. [
25] recommends four conditions that could be used to test a public spending decision. These are publicity of the rationale for spending, relevance of the rationale to the context of the decision, appeals to allow feedbacks and change and enforcement of the decision made. The bottom line is whether the public spending decisions are driven by economics and ethics. There is currently no measure of the marginal gains of one scheme over the other. Use of administrative and managerial tools like Programme Budgeting and Marginal Analysis, also termed as options appraisal, which also provides for inclusion of users perspectives in programme management are highly recommended. Also application of concepts of allocative efficiency that addresses the issue of achieving the right mixture of healthcare programmes to maximise the health of society are essential.
Management of contracting in arrangements
Contracting in is within itself a skill demanding job when it comes to implementation at grass roots. The ignorance to developing the required skills in the public system is not a good sign especially when contracting in is seen as panacea to all specialist shortages in the system. Our findings are in agreement with those from authors of other studies on PPP in health in India [
26] in that they emphasise the need to develop capacities in public system personnel in designing and managing contracts. In the absence of experience in developing and administering contract instruments, as our findings show especially with recent decentralisation of this authority to lower levels, there ought to be mechanisms to innovate and evaluate. Also as concerns have been raised of the ability of private providers to serve for public interests, designing mechanisms for overcoming challenges to mutual trust, logistical arrangements and financial transactions like social auditing [
27] could be beneficial.
Prior dissemination
A part of the data presented in this paper was shared at the National Conference on Bringing Evidence to Public Health Policy, organised by Institute of Tropical Medicine, Belgium and Institute of Public Health, Banglore, India, December 2010. The abstract of this conference is published in BMC Proceedings, 6 (Suppl 1):01 (16 January 2012).
Acknowledgements
We gratefully acknowledge our technical advisory group members Dr Dileep Mavalankar, Dr P P Doke and Dr CAK Yesudian for advise with study design. We are grateful to health officials in study districts and study respondents for their cooperation and participation. We are grateful to State Health Systems Resource Centre, Maharashtra for bureaucratic support and to the John D. and Catherine T. Mac Arthur Foundation, USA for financial support to this study. We acknowledge peer reviewers Dr Abhijit Das and Padma Deosthali for critical comments on an earlier version of this paper.
Authors’ contribution
All authors contributed to the conceptualisation and design of the study. BR and SC participated in the data collection and analysis. BR wrote the first draft of this paper. SC and NM contributed to revise the draft and BR wrote the final version. All authors read and approved the final version. The authors declare that they have no competing interests.