Save the Children -- Bangladesh MotherCare project : final evaluation

The evaluation covers the period 10/93 through 1993. The project significantly increased knowledge and, perhaps more importantly, practice on the part of mothers and support persons (husbands and mothers-in-law) in regard to such behaviors as attendance at pre-natal care (PNC) sessions, increased food consumption and fluid intake, intake of iron tablets and of Fe, feeding colostrum, exclusive breastfeeding through months 1-4, and use of modern contraception. Also significantly increased was mothers' use of, and perception of support from family members for, such protective health behaviors as use of trained traditional birth attendants (TBAs) and referrals for complicated deliveries. The project improved TBA management of normal deliveries, significantly increasing TBA use of soap and plastic sheets during delivery and of boiled blades and decreasing the placing of materials (e.g., hot compresses, ash) on the umbilical cord stump (a practice significantly associated with neonatal health problems). There was no association between delivery by a trained TBA and the incidence of maternal problems post partum, but infants delivered by trained TBAs were significantly less likely to have health problems. Given the current debate over the impact of TBA training and use of safe birth kits, this association bears more careful investigation, with control for possible confounders. During pregnancy, dizziness was the leading problem, though women usually had to be prompted before admitting any problems during pregnancy. This lack of spontaneity, along with a tendency not to regard problems as "serious", makes it difficult to assess levels of knowledge about potentially serious problems during pregnancy and/or willingness to seek treatment. Prolonged or delayed labor was the most common problem during delivery, while "fever" accounted for the largest proportions of problems in the post-partum period. Acute respiratory infections (ARI) accounted for most problems of the newborn. For problems in pregnancy or delivery, women and their families would first seek care from local providers (SC personnel and village doctors); only in the last resort would they go to the hospital. The training program seems to have improved the status of TBAs to the point where their advice on the need for advanced care was generally heeded. Even after being referred, however, families would seek help from nurse midwives or village doctors. Local government service sites were consulted for only a small proportion of problems; the lack of providers at such sites is a notable problem, despite continuing government efforts to improve maternal child health services and to increase the number of government personnel actually posted in villages. NasirNagar is a particularly difficult area for recruitment. Another major reason for not seeking treatment was lack of money. The services of village doctors, in particular, is prohibitively expensive. The quality of the care they offer is also of concern: their administration of syntocin to women in labor is just one of their many potentially adverse health practices. Proscriptions against leaving the home did not appear to be an important obstacle. Birth and infant death rates decreased during the project. From 1/90 to 12/93, the birth rate dropped from 37 to 32, infant mortality from 116 to 100, and neonatal mortality from 54 to 53, though to what extent these changes were due to project interventions is uncertain. The target population is too small to permit reliable calculation of maternal mortality. A slightly higher proportion of neonatal deaths was attributed to tetanus in 1993 than in 1990, possibly due to the shortage of tetanus toxoid during the first few months of that year. In both 1991 and 1993, the main cause of neonatal death was ARI. A large proportion of women of reproductive age are malnourished, with some groups especially vulnerable. In the final survey, mean mid upper arm circumference (MUAC) for women aged 15-45 was 22.6 cm (up from 22.2), while 5% of the population was severely malnourished and 25% moderately malnourished, down from 8% and 32% respectively. In both surveys, MUACs of poorer and younger (15-20 year old) women were significantly smaller than those of "more affluent" and older (26-35 year old) women.