2003 CBD: Evaluation of a Traditional Birth Attendant Programme: Stung Treng Province, Cambodia
Author: Lan, C.; Dunbar, M.
The Traditional Birth Attendant programme began in 1993. The programme's priority was on training as many TBAs as possible to perform clean, safe deliveries; detect complications and refer. Ongoing activities of the TEA program are: follow-up support and on-the-job training, both at meetings and during village visits; three monthly TEA report reviews and re-supply of materials and equipment; two yearly refresher training for TBAs who have already been trained and six-day TEA training courses for TBAs in villages where there is a shortage of trained birth attendants; reimbursement of transport costs for emergency referrals by TBAs and capacity building of the TEA programme managers through formal courses and on-the-job training.
In 2001, UNICEF assumed responsibility for providing the budget to the Provincial Health Department for follow-up activities of TBAs in Sesan District in Stung Treng Province. Youth with a Mission (YWAM) continues to provide money to Sesan TBAs to cover transport costs when they refer women to the provincial hospital. In 2001, the TEA programme was renamed 'The Bridge of Friendship'. The provincial health department assumed complete responsibility for management of the programme in July 2001. YWAMS strategy is to gradually withdraw both financial and technical support, with PHD eventually assuming complete responsibility for the 'Bridge of Friendship'.
The purpose of the evaluation was to provide an objective assessment of progress made in meeting the objectives of the TEA programme, document lessons learned and to make recommendations for the future direction of the programme. The overall objective of the TEA evaluation was to assess the impact the TBA's programme has had on strengthening maternal and child health services in the target areas, and improving the health of mothers and infants. This was to be achieved by:
- Assessing changes during the program period in the knowledge, attitudes and practices of TBAs who participate in the program, and identifying the strengths and weaknesses of their work.
- Assessing the strengths and weaknesses of the various components of the programme, including the training course, reporting system, management system, monitoring and referral system.
- Assessing whether project objectives were achieved at a reasonable cost.
A triangulation of methods was used, including review and analysis of relevant data and several qualitative methods including TBA skills observations, semi-structured interviews with key stakeholders including PHD, UNICEF, YWAM staff, MCH trainers/ health centre managers/ referral hospital and health centre midwives; individual interviews with TBAs, families, mothers, community leaders and village development committee members, and TBA and mother focus discussion group discussions. A total of 101 key stakeholders participated in the evaluation.
Study sites were selected independently by the evaluators based on the geography and location, distance from the nearest health centre/referral hospital and time available for field work. The evaluators decided to spend quality time in fewer villages rather than rushing to cover many villages and districts. Eight villages in three districts were visited.
There was no available baseline information concerning the knowledge, skills and practices of TBAs in the programme area. To address this, the evaluators compared the knowledge, attitudes and skills of Stung Treng TBAs with the results of several other TBA baseline surveys conducted in other Cambodian provinces. Also, during TBA interviews and discussion groups, the evaluators recorded TBA's self-assessment of changes following training. There was a lack of, and inconsistency of, MCH data in Stung Treng. The evaluators focused the analysis of data on the three years preceding the evaluation from 2000-2002. More complete and accurate data was available for this time period.
Findings and Conclusions:
The key question the evaluation sought to answer was: "Has the TBA programme in Stung Treng contributed to the strengthening of the MCH services in the province and improving the health of mothers and infants?" The evaluators found that the provincial health department, health centre staff, YWAM and, more recently, UNICEF, through a strong collaborative effort in supporting the TBA programme, have achieved a significant improvement in the quality of MCH services available to women in remote villages of Stung Treng. With difficult access and often impossible travel conditions, 394 TBAs (with median age 45 yrs. - 55 yrs.) in villages of all five districts have been trained and are active. They are provided regular follow-up support and supervision. Equipment and supplies are re-supplied in a timely manner and information from village level about births, deaths and transfers are reported at regular intervals. The evaluation team found an impressive level of knowledge and skills among TBAs observed and interviewed, concerning how to conduct a clean and safe delivery, and identify high-risk women that need referral. TBAs reported several changes that had taken place in TBA practices, such as the cleaning and sterilising of equipment, wearing of gloves, putting of gentian violet on the cord, waiting for signs of placental separation before attempting delivery of the placentas, and drying the baby and putting to the mother's breast instead of giving an immediate bath. Interviews and discussion with village women who had delivered with a TBA confirmed these practices.
Statistics and follow-up investigation show that neonatal tetanus is probably not a problem in Stung Treng. This is an important indicator of success for the TBA program, which has very likely contributed to this achievement by increasing the number of clean deliveries conducted by TBAs and achieving appropriate care of the cord. Tetanus toxoid vaccination rates remain too low to contribute towards this achievement (15% T2 in 2002).
Stillbirths and neonatal deaths remain high but there are some signs of improvement. As there is no record of the age at death of the neonate, it is impossible to estimate perinatal mortality rates. Neonatal mortality statistics show a decrease from 4.0% of total reported deliveries in 2000, to 3.7% in 2001, and 3.2% in 2002. There is very little reporting or recording of infant or under-five mortality so it's impossible to assess if the health of under-five children has improved. From comments during village-level discussions, it appears that, in both groups, deaths rates are high.
Immunisation coverage rates for under-one year old children have improved in the last three years (66% of children receiving DTC 3 and Polio 3 in 2002 as compared to 53% in 2000). These improvement have occurred due to increased outreach activities by health centre staff. Antenatal care with government midwives has increased dramatically in 2002 (from 28% in 2001 to 58% in 2002) and this is attributed to increased outreach activities by health centre midwives and the participation of TBAs during outreach.
The TBAs are to be commended for their ever-increasing involvement in transferring high-risk women. The number of transfers of emergency obstetric cases to the referral hospital is increasing yearly (49 in 2000, 57 in 2001 and 74 in 2002). A review of the cases transferred in 2002 showed that TBAs competently, and sometimes under very difficult conditions, accompanied women to the hospital. TBAs often stayed with the woman several days until her condition stabilised. From the review of transferred cases, it is also clear that, without the prompt action of the TBA concerned, some of these women may have been added to the maternal mortality statistics.
Observed difference in maternal death statistics in a small population province such as Stung Treng is not a good indicator by which to measure the success of an MCH programme. The maternal mortality statistics are too few in number and differences observed may be due to changes in the reporting system or may be subject to a wide random variation resulting from a small number of events. There may be large changes in the burden of morbidities before this is reflected in the MMR. It is positive that maternal death audits have been conducted since 2002 and that TBAs in Stung Treng are active in the reporting of maternal deaths. In 2002, TBAs reported the majority of maternal deaths that were investigated. In a review of the cases that TBAs were involved in, it appears that the TBAs responded appropriately according to the situation. It is likely that all maternal deaths are reported.
Recommendations to PHD, UNICEF and YWAM
1. A decision should be made as soon as possible concerning the future of the presently divided Stung Treng TBA programme. PHD in consultation with YWAM and UNICEF should decide which agency is in the best position to offer long term (at least two years) continuing technical and financial support to the TBA programme in all five districts. The evaluator's assessment is that YWAM because of its history with the TBA programme in the province, its experienced staff, its present support to four of the five districts and its long term commitment to the Stung Treng health programmes is in the best position to do this. If all parties involved are in agreement and YWAM is willing and able to agree to this, they should re-assume responsibility for the support of TBAs in Sesan district. If funds are not immediately available to do this UNICEF should be requested to provide funds to YWAM as an interim measure until further funds can be identified.
2. PHD should play the lead in organising and conducting a workshop involving key stakeholders in the TBA programme to develop a comprehensive TBA policy for the province, addressing issues such as management structure at each level of the programme, a standard TBA job description, relationships, role and responsibility of health centre and provincial level staff concerning the TBA programme, transfer of emergency cases, TBA perdiem levels and the resupply of TBA supplies and equipment.
3. An extra full time midwife should be assigned to work at provincial MCH level. The present MCH staffing level is too low to adequately cover the amount of activities currently being implemented.
Recommendation to TBA programme managers at provincial level.
4. With the support of the YWAM TBA programme technical advisor, and in collaboration with health centre staff a plan should be developed to ensure the smooth and gradual decentralisation of TBA programme activities to health centre level. It is recommended that ultimately health centre staff should be responsible for all TBA training, follow-up meetings and monitoring of TBA activities at village level. Provincial level staff should increase their involvement in on the job training of health centre staff and co-ordination and monitoring of MCH activities at health centre level.
Recommendation to PHD and UNICEF
5. PHD with the support of UNICEF and possibly an external consultant should conduct an assessment at the referral hospital, addressing the urgent need for improvement of provision of quality emergency obstetric care. Funds should be identified to follow through on subsequent recommendations.
Recommendation to TEA programme managers at provincial level.
6. The provincial TEA programme managers should work with health centre staff, health centre management committees, village development committees, TBA's and other organisations and groups working at village level to develop and strengthen the referral system at village level.
The TBAs presently lack support concerning organising emergency transfers. They require strong backup support from village leaders or committees when the decision is made to make an emergency transfer. The development of a realistic achievable village emergency transfer fund is also important. This emergency fund scheme should be started on a pilot basis in a few villages who have active motivated development committees. Following a review of lesson learned the programme could be adapted and introduced to other villages.
7. The accuracy of the TEA reporting system would be improved by more frequent collection of TEA statistics. Statistics may be more reliable if they were collected monthly, as the TBAs would then be encouraged to record their activities closer to the time they were conducted. The logistics involved in doing this will vary according to the area and the human resources avail able. It might be possible that VDC members could collect monthly statistics from TBAs in their respective villages and send them to the health centre level.
8. Future refresher training for TBAs should concentrate on improving the skills of TBA in the areas identified during the evaluation. These are: the promotion and support to mothers in the establishment of exclusive breast feeding, identification and referral for treatment of anaemic women both antenatally and post natally; the prompt identification of problems in infants and mother in the post natal period; improved health education skills to promote antenatal care, tetanus vaccine, exclusive breast feeding and the introduction of appropriate timely complementary feeding. If funding is available it would be helpful to develop and pre-test a TBA flip chart with colourful pictures, key messages and minimal script that TBAs could use as a tool when discussing health topics with village women.
9. The system of documentation and recording of TBA programme activities and statistics should be reviewed. If possible a user friendly centralised accessible data base should be designed that standardises how important TBA programme indicators are recorded. The TBA programme provides a lot of important information. If the data is easy to access, it can help identify and respond to problems promptly and facilitate easier monitoring of the programme.
10. There is a need for effective IEC materials and health education sessions for village people concerning important health topics and prevention common diseases. Important health education topics for women and their partners are antenatal care, danger signs during pregnancy labour and delivery, family spacing, breast feeding, nutrition, immunisations and introduction of appropriate timely complementary feeding. The TBA cannot be expected to take complete responsibility for village health education, human resources at village level should be identified that can participate in conducting these activities.
11. Despite the reluctance of TBAs to promote or use the individual home birth kit, the kit has some important benefits including the active participation of women in planning for their delivery and ensuring it is as clean as possible. The kit also guarantees that each mother will have her own disposable gloves and plastic sheets. The kit should be continued to be promoted at village level. Women might be more enthusiastic about buying it if they understood better the advantages the kit offers. A flexible approach should be used with TBAs who have been proficiently and safely using scissors and clamps for many years. They should be encouraged to use as many of the disposable items in the kits as possible.
12. The TBA programme managers should strongly advocate that the National MCH Centre review the adequacy of the materials supplied in the individual home birth kit and make recommendations for adding additional materials according to field experiences in Stung Treng.
13. The possibility of 'waiting homes' at provincial level for high-risk pregnant women should be explored. They have found to be successful in several other countries, especially in countries that developed simple realistic plans with active involvement of communities. Stung Treng transfer statistics show that very few high-risk women are electively transferred during pregnancy. Most are transferred when they are already an obstetric emergency. This report has mentioned some of the reasons for this situation (WHO, 2002).
Full report in PDF
PDF files require Acrobat Reader.