International Journal of Gynecology and Obstetrics

Transcription

1 International Journal of Gynecology and Obstetrics 107 (2009) Contents lists available at ScienceDirect International Journal of Gynecology and Obstetrics journal homepage: AVERTING MATERNAL DEATH AND DISABILITY Where there is no obstetrician increasing capacity for emergency obstetric care in rural India: An evaluation of a pilot program to train general doctors Cherrie Lynn Evans a,b,, Deborah Maine a, Lois McCloskey a, Frank G. Feeley a, Harshad Sanghvi c a School of Public Health, Boston University, Boston MA, USA b School of Public Health, Columbia University, New York NY, USA c Jhpiego, Johns Hopkins Program for International Education in Gynecology and Obstetrics, Baltimore MD, USA article info abstract Keywords: Cesarean delivery Emergency obstetric care Human resource delegation India Maternal mortality Task shifting Background: Maternal mortality continues to be high in rural India. Chief among the reasons for this is a severe shortage of obstetricians to perform cesarean delivery and other skills required for emergency obstetric care (EmOC). In 2006, the Government of India and the Federation of Obstetric and Gynecological Societies of India (FOGSI) with technical assistance from Jhpiego, instituted a nationwide, 16-week comprehensive EmOC (CEmOC) training program for general medical officers (MOs). This program is based on an earlier pilot project ( ). Objective: To evaluate the pilot project, and identify lessons learned to inform the nationwide scale-up. Methods: The lead author (CE) visited trainees and their facilities to evaluate the project. Eight data collection tools were created, which included interviews with informants (program/government staff, regional/international experts, trainees and trainers), facility observation, and facility-based data collection of births and maternal/newborn deaths during the study period. Results: More trainees performed each of the basic EmOC skills after the training than before. After training, 10 of 15 facilities to which trainees returned could provide all signal functions for basic EmOC whereas only 2 could do so before. For comprehensive EmOC, 2 facilities with obstetricians were providing all functions before and 2 were doing so after, even though the specialists had left those facilities and services were being provided by CEmOC trainees. Barriers to providing, or continuing to provide, EmOC for some trainees included insufficient training for cesarean delivery, lack of anesthetists, equipment and infrastructure (operating theater, blood services, forceps/vacuum, manual vacuum aspiration syringes). Conclusion: Although MOs can be trained to provide CEmOC (including cesarean delivery), without proper selection of facilities and trainees, adequate training, and support, this strategy will not substantially improve the availability of comprehensive EmOC in India. Recommendations: To implement a successful nationwide scale-up, several steps should be taken. These include, selecting motivated trainees, implementing the training as it was designed, improving support for trainees, and ensuring appropriate staff and infrastructure for trainees at their facilities before they return from training International Federation of Gynecology and Obstetrics. Published by Elsevier Ireland Ltd. All rights reserved. 1. Introduction A severe and expanding shortage of health workers worldwide has created crises in many countries, including India [1]. Globally, the HIV/ AIDS epidemic, emigration, insufficiently trained providers, and poor human resource planning and management converge to create an environment, particularly in rural areas, where there are not enough skilled workers to provide care [2]. With a population of 1.1 billion people, India is experiencing a critical shortage, especially in rural areas where 71% of its people live [1,3]. To mitigate shortages, high- and low-resource countries have used various types of providers to fill human resource gaps [2]. This has Corresponding author. School of Public Health, Boston University, Boston MA, USA. address: cherev65@mac.com (C.L. Evans). been accomplished by expanding the roles of available healthcare staff to include skills previously the purview of more highly trained providers [1,4]. In some cases new cadres have been created, such as the técnicos de cirurgia, who were developed in Mozambique in the 1980s in response to the severe shortage of physicians [5]. Nonphysician clinicians (NPCs) have been successfully trained in orthopedics, trauma, urology, maxillofacial and plastic surgery, gynecology, and obstetrics including cesarean delivery [5,6]. This approach is sometimes referred to as task shifting or task sharing. In India, until very recently, task shifting has not been used as a strategy to improve the availability of emergency obstetric care (EmOC), since only specialist doctors were trained and widely authorized to perform cesarean delivery and other lifesaving obstetric procedures. Fig. 1 shows some of these procedures denoted as signal functions, which serve to identify facilities that provide basic (BEmOC) or comprehensive (CEmOC) emergency obstetric care /$ see front matter 2009 International Federation of Gynecology and Obstetrics. Published by Elsevier Ireland Ltd. All rights reserved. doi: /j.ijgo

2 278 C.L. Evans et al. / International Journal of Gynecology and Obstetrics 107 (2009) classroom-based instruction followed by 10 weeks of self-directed practicum supervised by an obstetrician at the district hospital. The 17 doctors in the pilot program were trained in 4 groups at 2 training centers in Surat (Gujarat) and Jaipur (Rajasthan) from October 2005 to February The evaluation Fig. 1. Signal functions for basic and comprehensive emergency obstetric care. Source: (2009) WHO, UNFPA, UNICEF, AMDD. The Government of India (GOI) estimates that 6000 doctors are needed to provide 24-hour comprehensive EmOC in the 2000 rural first referral units (FRUs) slated to become operational by 2010 under the current maternal health program [7]. A GOI health information survey in 1999 found only 771 government obstetrician-gynecologists capable of performing cesarean delivery at rural FRUs [8]. This shortfall of CEmOC providers in rural areas is an important factor in the disparity between urban and rural maternal mortality (267 and 619 maternal deaths per live births respectively in 1999 [9]). Choosing to adopt a task shifting strategy to address the scarcity of CEmOC providers, and recognizing the relatively high number of medical officers (MOs; an estimated ) in rural areas, in 2006, the GOI, with the Federation of Obstetric and Gynecological Societies of India (FOGSI) and with technical assistance from Jhpiego (Johns Hopkins Program for International Education in Gynecology and Obstetrics), instituted a program to train MOs to provide CEmOC, including cesarean delivery. This program is based on a pilot project, evaluated here, which was implemented from 2004 to 2006 in two states in India. The pilot project was initiated with funding from Columbia University's Averting Maternal Death and Disability (AMDD) Program and was later funded by the MacArthur Foundation. During the pilot project, 17 MOs were trained over 16 weeks to provide CEmOC. The nationwide scale-up is now being funded by the GOI with input from the states and assistance from the MacArthur Foundation and Jhpiego. Our evaluation of the pilot project was designed primarily to determine whether training the MOs enabled the facilities to provide CEmOC. We also hoped to identify lessons learned to inform the ongoing national scale-up of this program. 2. Materials and methods 2.1. Training intervention Training centers affiliated with medical schools were selected and upgraded to conform to international and local obstetric care standards prior to training [8,10]. The 16-week course was adapted from the CEmOC training curriculum initiated by AMDD and designed by Jhpiego and first used at the regional CEmOC training in Dhaka, Bangladesh, in The course consisted of 2 parts: 6 weeks of Data collection Data were collected from program documents, facility observation, data abstraction at the facilities, and from semi-structured interviews with key informants (program and government staff, regional and international experts, trainees and trainers). Five new interview tools were created and 3 facility data collection tools were adapted from the Quality improvement tools using the Standards Based Management and Recognition Approach developed by FOGSI, MacArthur, and Jhpiego for this program and AMDD needs assessment instruments [11]. All data were collected by the lead author (CE) during November 2007 and February The 5 semi-structured interview tools were tailored for each type of informant based on their involvement with the program. Interviews conducted with key informants were designed to obtain a history of the pilot project and the informant's role and opinions regarding the project. The trainee interview was the most detailed and included 2 parts: the interview portion and a 21-item skills selfassessment survey. The trainee interview contained specific questions on prior experience and training, recruiting and posting, practice site staffing and EmOC service provision before and after training. Because the use of basic EmOC skills was rarely recorded in delivery registers, the skills self-assessment survey was used to measure the trainees self reported change in obstetric practices from the 6 months before training compared with the 6 months preceding the interview. This skills assessment instrument (adapted from a Jhpiego tool) was also used as baseline assessment of the trainees before the course [12]. The 3 facility assessment data collection tools were also adapted from Jhpiego and AMDD instruments. The first instrument, the Facility Observation/Abstraction tool, contains general information (such as facility type, catchment area, population, staffing, facility condition), and delivery statistics. We combined and adapted existing instruments developed by Jhpiego and AMDD and included components on the operation of the facility to rate cleanliness, lighting, electricity, staffing, communications, blood banking, operating theater equipment, event audits, and record keeping. A grid was added to record delivery statistics at each facility by month: total births, cesarean deliveries, maternal and newborn deaths, and stillbirths. To get these data, obstetric delivery and surgical registers were examined during the facility visit. The remaining 2 facility data collection instruments evaluated performance standards for emergency care in labor and delivery. Each used checklists of items to verify the existence of protocols and the presence and functionality of supplies and equipment for the provision of obstetric care including assisted vaginal delivery, cesarean delivery, anesthesia, and blood banking, as well as drugs required to treat infection, pre-eclampsia/eclampsia, and hemorrhage Analysis To prepare interviews for analysis, notes from each interview were transcribed into digital format for ease of content analysis. Interviews were sorted by type (key informant, trainee, etc.) and responses were grouped by instrument question. Specific items from both interviews and facility data collection tools were analyzed using descriptive statistics, and some of these data are reported below Subjects Key informants, other than the CEmOC-trained MOs, were selected based on their involvement with the pilot project. Some of these were identified through literature review and others from contact with

3 C.L. Evans et al. / International Journal of Gynecology and Obstetrics 107 (2009) individuals present at the inception of the pilot project. Using snowball sampling, the initial contacts generated additional subjects with the desired characteristics involvement with the pilot project and/or maternal health in India. Although the plan was to interview and conduct site visits with all 17 trainees who attended the pilot training, 3 trainees had left government practice (all from Rajasthan): 1 was never located, and 2 had left for postgraduate training. One of the trainees in postgraduate training was in residency at the EmOC training center in Jaipur and was interviewed; the other doctor in postgraduate training was interviewed briefly by phone. Thus, complete interviews were obtained from 15 trainees. Site visits to collect birth and facility data were made to 13 of the 14 facilities where trainees were actively practicing; the 14th facility faxed the data. Of the 17 trainees, 6 were from Rajasthan and 11 were from Gujarat. Retention in the public sector from completion of training to the end of the study in February 2008 was 82% overall, with 100% retention in Gujarat and 50% retention in Rajasthan. 3. Results The study was designed to evaluate whether the functioning of facilities was improved in terms of providing comprehensive EmOC after the MOs returned from the training course. Other training programs in India have shown that MOs can be successfully taught to provide BEmOC in significantly less time [13]. The 16-week CEmOC training course used in the pilot project was far more ambitious in scope and thus cost more in time and money. Cesarean delivery is the most complex of the skills taught in this training in breadth of knowledge required, hand skill development, and practice necessary to achieve competency. For this reason, evaluating the success of the training in terms of cesarean delivery was a major focus of this study. Basic EmOC skills require less intensive training, but are no less important in saving women's lives, and the change in trainee use of these basic skills was also captured Comprehensive EmOC After training, 6 CEmOC-trained MOs from Gujarat performed 134 verified cesarean deliveries at their facilities in the months since completion of the training. Four CEmOC-trained MOs reported 91 additional cesarean deliveries (20 in government facilities, 71 in private), but these surgeries could not be verified. By the time of the site assessment, the number of CEmOC-trained MOs doing cesareans had decreased so that only 3 of the 6 trained had performed any in the preceding 6 months. Two trainees who had stopped stated that the anesthetists at their facilities refused to provide services for their patients. The reasoning behind the refusal stated by one of the trainees was that the anesthetist did not believe the training qualified an MO to do cesarean delivery safely and that the anesthetist did not want to be liable. The other trainee believed the anesthetist at his facility was compensated for referring women to private doctors. What emerges from these findings are questions as to why 8 of the 16 trainees never performed cesarean deliveries after the training, and what caused 3 others to stop? Our analysis shows there are 3 underlying themes: training, facility capacity, and candidate recruitment Training The training course was implemented differently in the 2 states, with striking results. Although in Gujarat the training was implemented as designed, in Rajasthan, no trainee was permitted to perform any cesareans during training, 2 were allowed to assist in a few cases, and 1 trainee never even watched a cesarean delivery. Another important difference was that the doctors in Rajasthan were not sent to a district hospital for the practicum; instead they remained at the medical college for the 16 weeks. Because of this, the trainees not only encountered significant competition from obstetric residents, but they completed their training in facilities that had support and technical capacity that was not available in the more modest facilities to which they would return. In Gujarat, the number of cesareans performed by the trainees during their district hospital practica varied widely. Despite all 11 doctors being trained in obstetric surgery, 3 never performed cesarean deliveries after training. Two of these 3 performed 3 or fewer cesareans during the training course (Fig. 2) Facility capacity For the pilot project, candidates were supposed to be selected from facilities that had the capacity to provide CEmOC, including the appropriate staff, infrastructure, blood services, drugs, and supplies. Yet this was not the case. Our assessment showed that all trainees came from facilities lacking some or all of these requirements. Anesthesia: The single greatest facility-level barrier to CEmOCtrained MOs performance of cesarean delivery after training was insufficient anesthesia support. Only 1 of the 16 trainees had adequate anesthesia available for their obstetric cases and only 5 facilities even had an authorized position for an anesthetist. Of the 15 facilities without adequate anesthesia, 13 CEmOC-trained MOs had no anesthetist and 2 reported that anesthetists refused to provide anesthesia for their cases. In Gujarat, 5 CEmOC-trained MOs without in-house anesthesia contracted with private sector anesthesia services to perform cesareans. Infrastructure: Asked about the capacity of their facilities to provide CEmOC, many trainees cited problems with inadequate infrastructure and equipment. (These reports were confirmed during facility assessment visits.) In Rajasthan, 2 CEmOC-trained MOs were posted to low-volume primary health centers with the capacity for only basic care and were unable to use much of their training. Two other CEmOC-trained MOs worked in hospitals with no operating theatre, and 5 were posted where the operating theatres were unused because of lack of equipment or disrepair. Blood services: The universal lack of on-site blood services for transfusion was a significant barrier to providing comprehensive EmOC. All CEmOC-trained MOs voiced concern regarding the absence of on-site blood collection or storage and one reported 2 obstetric deaths at his facility in the preceding 12 months owing to lack of blood. Five facilities in Gujarat were reportedly slated for upgrade with blood storage units at the time of the site visits. Four other CEmOC-trained MOs from that state had used private blood services Trainee recruitment and selection As originally conceived, MOs working in rural or high-need areas could apply for the training. However, instead of a voluntary process, state governments are reported to have assigned MOs to the training. Thirteen of the 15 CEmOC-trained MOs interviewed said that after their facility was selected, they were ordered to attend the training. Four of these CEmOC-trained MOs reported displeasure at having been selected; none of the 4 subsequently performed any verifiable cesareans and trainers noted that these doctors performed less well during and after training Basic EmOC After training, more CEmOC-trained MOs reported they were using skills to manage each of the major obstetric complications than were doing so before (Fig. 3). The self-assessment survey was used to compare practice patterns for 21 obstetric skills before and after the training. CEmOC-trained MOs were asked whether they had used each skill in the 6 months prior to the training, whether they had been trained in that skill during the CEmOC course and if they had performed the skill in the 6 months prior to interview. If they had not performed a skill before or after the training, they were asked why.

4 280 C.L. Evans et al. / International Journal of Gynecology and Obstetrics 107 (2009) Fig. 2. India Cesarean deliveries by CEmOC trainees during training and in the months after. At the time of data collection, 10 of the CEmOC-trained MOs facilities were able to provide all signal functions for BEmOC whereas only 2 of those facilities could offer these before (Fig. 4). It is important to note here that the two facilities offering EmOC before the training had obstetricians who left the facilities after the arrival of the CEmOC-trained MOs within 2 and 12 months respectively. All 5 CEmOC-trained MOs who were unable to provide the complete set of 6 basic functions cited lack of equipment for assisted delivery and 3 of these also had no equipment for removal of retained products of conception. Of all the basic functions, the greatest gain was in the performance of assisted vaginal delivery (AVD) (Fig. 3), which is particularly important in facilities where the capacity for cesarean delivery is limited or absent. Before training, only 1 of the CEmOC-trained MOs performed AVD; after training, 10 were performing AVD. The remaining 5 CEmOC-trained MOs, who did no AVD, all expressed Fig. 3. India Number of doctors performing EmOC skills before and after the 16-week training.

5 C.L. Evans et al. / International Journal of Gynecology and Obstetrics 107 (2009) the original training in Dhaka. Using this as a model, 2 willing MOs at a given facility would be sent for training: 1 to the CEmOC course and 1 to the anesthesia course. In the interim, coordination of the placement of generalist doctors trained in comprehensive EmOC and anesthesia is needed Select candidates for training based on experience, level of interest, and willingness to commit to a predetermined number of years in a comprehensive EmOC facility Fig. 4. India EmOC capacity at 15 trainees' facilities before training and at assessment. confidence in their ability to use vacuum for delivery if they had access to the equipment. 4. Recommendations 4.1. Select or upgrade facilities as required to support comprehensive EmOC Ideally, each state should have a coordinated plan to select and upgrade facilities to be designated CEmOC centers based on demographic and geographic needs. After selection, several key activities are required: A complete facility assessment Upgrades in equipment, infrastructure, supplies, and drugs prior to placement of the trainee to prevent the loss of newly acquired skills. Where there is no blood banking at the facility, the availability of private, local blood services should be determined. The presence of a cooperative anesthetist should be assured. Anesthesia coverage warrants special discussion. Where the anesthetist is absent or uncooperative, cesareans will not be performed. Problems with anesthesia were the single greatest obstacle for the doctors who were appropriately trained in cesarean delivery. Anesthetists refusing to work with the trainees at government facilities signal a potentially widespread problem and could hamper the nationwide expansion of the program. To address the shortfall of specialist anesthetists in rural areas, the GOI has developed a 17-week training program for MOs to provide anesthesia for obstetric emergencies [14]. In Gujarat, the state government is working to place these newly trained anesthetists at the same facilities as the EmOC trainees. Where there is no anesthetist at a facility, several options can ensure coverage as befits the state context: Post one of the new doctors trained by the government in anesthesia or recruit a willing MO from the facility for anesthesia training. Transfer an anesthetist from a different facility if available. Contract with an anesthetist in the private sector as was done by 5 trainees. Because of continued limited anesthesia capacity, consideration should also be given to instituting team training similar to that used in The importance of appropriate candidate selection cannot be overstated. Priority should be given to selecting trainees with prior surgical training or experience in sterilization or abortion care as we learned such candidates were far more likely to be performing cesarean deliveries after training (Table 1). Length of time remaining in the applicant's government contract should also be considered, as should their willingness to make a commitment to practice in a high-need facility for a specified period. Finally, candidates must be willing to remain in public service without going into postgraduate training within the specified period. Using an application process as is now being done in Gujarat (as opposed to mandatory assignment to training) may help with selection of willing candidates Implement the training as designed As plans for scale-up move forward in other states, it is important that stakeholders agree to the basic principle of training MOs to provide CEmOC. Stakeholders must also agree with the practicum taking place in district hospitals that have been assessed for appropriate obstetric volume and teaching capacity. It is clear from this assessment that the trainees in Rajasthan were never trained to do cesarean delivery, and thus were never able to provide CEmOC. It has already been shown that MOs can be successfully trained in basic EmOC in India in 2 weeks, therefore committing to the more extensive comprehensive EmOC training is an inefficient use of resources if doctors will not be taught to perform cesarean delivery. Table 1 India 2007 CEmOC trainee characteristics. Trainee a Facility type Training batch Obstetric experience prior to training, y Surgical training prior to CEmOC training Trainees who did not provide cesarean delivery after training 2R PHC 12/05 R 3 3R PHC 4/06 R 1 4R? 2/06 R??? 5R CHC 2/06 R 13 6R CHC 12/05 R 20 7R PHC 12/05 R 0 8G CHC 12/05 S 11 13G CHC 4/06 S 7 Yes 14G CHC 12/05 S 7 Trainees who did provide cesarean delivery after training 1G CHC 4/06 S 9 9G CHC 4/06 S 9 10G CHC 12/05 S 10 Yes 11G CHC 12/05 S 9 Yes 12G CHC 4/06 S 9 Yes 15G CHC 4/06 S 9 Yes 16G CHC 4/06 S 8 Yes 17G CHC 4/06 S 15 Yes Obstetric training prior to CEmOC training Abbreviations: PHC, primary health center; CHC, community health center. a Trainee numbers are followed by R for doctors trained and posted in Rajasthan and G for those trained and posted in Gujarat. 4R was never located.

6 282 C.L. Evans et al. / International Journal of Gynecology and Obstetrics 107 (2009) Conclusion Task shifting of the lifesaving skills required for CEmOC is one important tool to strengthen health systems and save lives. Recognizing this, India has made the bold decision to train MOs in CEmOC including cesarean delivery. For this to be successful, lessons learned from the pilot project need to be heeded as the training is scaled-up nationwide. What we know from looking at those MOs who were successful in providing CEmOC is that they were appropriately selected, trained, and supported and placed in adequately equipped facilities with access to required staff. From those CEmOC-trained MOs who did not provide CEmOC, we learned that training must be voluntary and include all aspects of CEmOC. To provide good quality obstetric care, training cannot occur in a vacuum. A sufficiently functioning health system is required and training MOs in CEmOC is only one piece of the puzzle. Other efforts, such as the anesthesia training of MOs, Chiranjeevi Yojana (reimbursement of participating private specialists who provide obstetric care to poor women), and JSY (a voucher scheme that pays women to have institutional births) need to be evaluated for effectiveness and then coordinated so these programs can work together. Finally, adequate monitoring and evaluation of maternal health outcomes needs to be institutionalized and simplified within the health system. Acknowledgements Columbia University's AMDD Program provided a planning grant of US $ to FOGSI in 2002, which was followed by a grant of US $ from the MacArthur Foundation to Jhpiego and FOGSI to implement the pilot. The lead author was funded by AMDD to evaluate this project. Dr Dileep Mavalankar at the Indian Institute of Management in Ahmedabad provided assistance and support. References [1] Chen L, Evans D, Evans T, Sadana R, Stilwell B, Travis P, et al. Working together for health: The World Health Report. Geneva: WHO; p. 6. [2] Lehmann U. Mid-level health workers: The state of the evidence on programmes, activities, costs and impact on health outcomes. A literature review. Geneva: WHO; p [3] Haub C. Population Reference Bureau: 2007 World population data sheet. Washington, DC: Population Reference Bureau; p [4] Dovlo D. Filling the gaps: Introducing substitute health workers in Africa. id21 Research Highlight; p. 1. [5] Vaz F, Bergstrom S, Vaz Mda L, Langa J, Bugalho A. Training medical assistants for surgery: Policy and practice. Bull World Health Org 1999;77(8): [6] Pereira C, Bugalho A, Bergstrom S, Vaz F, Cotiro M. A comparative study of caesarean deliveries by assistant medical officers and obstetricians in Mozambique. Br J Obstet Gynecol 1996;103(6): [7] Desai S. Expanding emergency obstetric care: Innovative role by Federation of Obstetric and Gynecological Societies of India, Delhi; [8] Desai S, Bhatt P, Bhardwaj A. Operational plan - EmOC FOGSI, MOHFW; p [9] Ministry of Health and Family Welfare India. RCH Phase II: National program implementation plan. New Delhi: Government of India; [10] Jhpiego. Strengthening the provision of high-quality emergency obstetric care services in India in collaboration with the Federation of Obstetric and Gynecological Societies of India and the Indian College of Obstetrics and Gynecology. Jhpiego; p [11] Jhpiego. Emergency obstetric care (EmOC): Quality improvement tools - Using the Standards Based Management and Recognition approach. Jhpiego; p [12] Jhpiego. Individual assessment of emergency obstetric care. Jhpiego; p [13] Srinivasan V, Dwivedi H, Mavalankar D. Increasing access to Emergency Obstetric Services: Experiences from Rajasthan, India. UNFPA; [14] Mavalankar D, Callahan K, Sriram V, Singh P, Desai A. Where there is no anesthesiologist - increasing capacity for emergency obstetric care in rural India. Int J Obstet Gynecol 2009;107(3).