Maternal referrals to specialized care in Mtwara Tanzania The impact of a modified referral criterion on compliance

Abstract

Background: Maternal mortality in Tanzania which is estimated at 5781100000 live births is unacceptably high. Available data show that about two thirds of maternal deaths are caused by direct obstetric complications which can be avoided if women who develop complications can receive emergency obstetric care timely. The risk screening approach during ANC is used with the aim of increasing access to emergency care for women who are likely to have complications. However, it has been reported that referral compliance among women identified to be at a high risk is low and one of the reasons reported to lead to low referral compliance is having many unjustified risk factors. The ANC risk catalogue used in Tanzania has been found to be too wide and unspecific. The new WHO ANC model package identifies the list of risk factors for referral which are evidence based but are not used in Tanzania. The main objective of this study was to determine the barriers to referral compliance and compare the effectiveness of the New WHO ANC model referral criteria and the Tanzania MoHSW ANC referral criteria in improving referral compliance and their validity in predicting adverse pregnancy outcomes. The study tested the hypothesis that the modified ANC referral criteria is more effective than the Tanzania MoHWS ANC referral criteria with regard to prediction of adverse pregnancy outcomes and improves referral compliance for women identified to be at risk of an obstetric complication. Methods and Materials: The study was conducted in Mtwara rural district in south eastern Tanzania. The design of the study was an intervention which was done in three phases. In the first phase, baseline studies were conducted at the community and in health facilities. These were cross sectional in design and employed qualitative and quantitative research' methods. The second phase was an intervention which involved training of maternal health care providers on Focussed ANC. Also a Cluster Randomized Controlled Trial to compare the effectiveness of the modified ANC referral criteria and the MoHSW ANC referral criteria on referral compliance and their validity in predicting adverse pregnancy outcomes. The third phase of the study was a post intervention assessment which was cross sectional in design. Data was entered into the computer using EPI INFO for windows version 3.1.1 and analysed using SPSS and STATA softwares. Bivariate analysis was done using Chi square tests and multivariate analysis was done using Factor analysis and Multiple logistic regression. Results: A total of 974 women were interviewed and eighty six indepth interviews were conducted during the baseline assessment. Findings revealed universal use of antenatal care (99.8%), low use of skilled delivery care services (45.8%), a large proportion of women with risk factors (52%), high referral rate (40%), and low referral compliance rate (28%). Significant determinants of use of skilled care at delivery included distance to a health facility providing delivery care (OR=4.35 (2.76-6.89), knowledge of risk factors (OR=1.96 (1.05-3.68), discussion during pregnancy with partner on place of delivery (OR=2.48 (1.76-3.48) and advice of a health worker during ANC on health facility delivery (OR=1.61 (1.06-2.43). Similarly, compliance to a referral advice was seen to be significantly influenced by distance to the facility providing emergency obstetric care (OR= 4.89 (2.66-8.96), knowledge of risk factors (OR=.2.86 (1.26-6.5) and parity (OR=0.25 (0.12-0.59). Knowledge of pregnancy and delivery risk factors was found to be low with only 32.6% of women able to mention at least one pregnancy or delivery risk factor. Only few risk factors were perceived by women to require hospital delivery and they include previous operative delivery, excessive bleeding after delivery, eclampsia, prolonged labour and low HB. Facility survey showed a serious shortage of health care personnel of 49% but the state of infrasturture, availability of essential equipment and supplies in the surveyed facilities was good with all facilities scoring above the minimum score of 80% on service availability. Barriers to skilled care at delivery and referral compliance included distance to facilities providing delivery care and lack of reliable means of transport, unnecessary referral to hospital, presence of TBAs in the community and a belief that if a woman went through pregnancy safely she will also have a safe delivery. In the cluster randomized controlled trial, a total of 7104 women had completed followup of which 3455 were in the control arm and 3649 in the intervention arm. There was almost a four fold reduction in ANC referrals in the intervetion arm (13%) compared to the control arm (52%). Referral compliance was significantly higher in the intervention arm compared to the control arm (OR=1.88 (1.32-2.68). Use of skilled care at delivery was much higher in the intervention arm (48.9%) than in the control arm (41.7%) and the difference was statistically significant (P<0.05). The modified ANC referral criteria was found to be a better screening tool with regard to prediction of obstetric complications (PPV= 27% Vs 7% in the MoHSW ANC referral criteria), mode of delivery (PPV=8% Vs 2% in the MoHSW referral criteria), and pregnancy adverse outcomes (PPV=lO.1 % Vs 2.3% in the MoHSW ANC referral criteria) while fetal and maternal outcomes were similar in the two groups. Conclusion and Recommendations: Findings of the study lead to a conclusion that utilization of skilled assistance during delivery and the referral compliance rate for women at risk of obstetric complications are both very low in the study area. Findings also showed that the MoHSW ANC referral criteria lead to too many unneccessary referrals with low referral compliance rate where as the modified ANC referral criteria which employed the referral criteria in the new WHO ANC model leads to improved referral compliance and also increased use of skilled care at delivery without jeopardising pregnancy outcomes. The recommendations include adopting the modified ANC referral criteria, improving counselling on individual birth planning during pregnancy and improving knowledge of pregnancy and delivery danger signs at the community level.