Perinatal mortality and anaemia in pregnancy in rural northern Tanzania

Abstract

The aim was to estimate the perinatal mortality rate, to study potential determinants and causes of perinatal and neonatal deaths, and their avoidability in a rural area of Tanzania. Further, we studied the prevalence of anaemia in pregnancy and the determinants of anaemia in pregnancy. Attendants of the existing Mother- and child-health (MCH) clinics in the area run by the Haydom Lutheran Hospital were registered on their first antenatal visit and followed up one month after birth. Those who disappeared were traced to their home. In the cohort of 3618 women, some women were not possible to find (106) and 3512 were followed up and identified. Spontaneous abortion was reported in 42 cases, 3359 had a living baby, 53 had a still birth, 42 had an early neonatal death and 16 had a late neonatal death. We studied the causes of stillbirths and neonatal deaths and their avoidability in the cohort. To secure a more complete picture, we also included deaths identified in a household study of 1259 homes in the same area: 6 stillbirths, 8 early neonatal deaths and 7 late neonatal deaths. The MCH attendants (n=3836) were studied cross-sectionally for prevalence and determinants of anaemia in pregnancy. Also, based on their hemoglobin (Hb) value, we selected 153 controls and 159 cases of anaemia with varying severity “nested” within the cohort, for a study to evaluate micronutrients and infections as determinants of anaemia. The perinatal mortality rate (PMR) in the cohort was 27/1000 births, 56% were stillborn and 44% were early neonatal deaths. There was increased risk of perinatal death among babies with low birth weight and among women with a positive s-VDRL (Venereal Disease Research Laboratory). Women who had previously lost a baby and very lean women were also at higher risk. Still births and neonatal deaths were often related to infection (39%), asphyxia (24%) or immaturity (15%). Among these deaths, 15% were estimated to be probably avoidable and a further 10% possibly avoidable, under the prevailing circumstances. Among the potentially avoidable stillborns and neonatal deaths, a patient-oriented avoidable factor was present in 51%, and a provider-oriented factor in 65%. Very few of the women were aware of their risk factors. The mean Hb of the pregnant women was 12.1 g/dl, and 4.5% of them had Hb below 9.0 g/dl. The mean Hb was higher among persons living at higher altitude and at higher maternal age, and was lower during the malaria-season, and among women with malaria parasitaemia.Anaemia among pregnant women was associated with iron deficiency, folate deficiency, and vitamin A deficiency. It was also associated with elevated C-reactive protein (likely infection) and elevated Lactate dehydrogenase (hemolysis, likely due to malaria infection). It was also associated with general signs of undernourishment (a small arm circumference). The estimated PMR in this setting was lower than what has been found for other areas of Tanzania. This might be attributed to a well functioning health system in the area. The MCH had a high attendance of pregnant women, and although the referral to higher level was low, the existing MCH familiarized the women with the health system and made it more natural to be admitted to hospital if needed for delivery. Still, at least a quarter of the deaths in the study could have been avoided under the prevailing circumstances. The MCH should make sure that messages are understood and referrals accomplished. Anaemia in pregnancy was less prevalent than in other areas of Tanzania. Common risk factors of anaemia were deficiencies of iron, folate, vitamin A, and infections, including malaria.