Female interviewers administered questionnaires to 25,745 women of a possible 30,918 in the two states; 1.2% declined the interview (345 or 1.8% in Cross River and 37 or 0.3% in Bauchi); a further 15% were not available at the time of the visit (4,213 or 22.2% in Cross River, where more women have formal employment, and 429 or 3.6% in Bauchi). A total of 15,621 women had given birth (7,759 in Cross River and 7,862 in Bauchi) in the last three years.
Table lists the frequency of household characteristics, male knowledge and attitudes, antenatal care, work during pregnancy, IPV and FGM, and female knowledge, attitudes, intentions, and agency. One third lived in urban areas in Cross River, one half of that proportion in Bauchi. Nearly all Cross River women had formal education compared with one in every four Bauchi women.
Reports of pre-eclampsia and eclampsia were comparable in Bauchi (10.3% weighted value of 842/7684) and Cross River (13.0% weighted value of 973/7178). However, post-partum sepsis was much more common in Cross River (30.6% weighted value of 2223/7176), compared with 5.6% (weighted value of 473/7724 in Bauchi). The principal analysis combined pre-eclampsia, sepsis and other complications including excessive bleeding and convulsions as maternal morbidity related to pregnancy, delivery or post delivery: 17.8% of women in Bauchi and 43.9% in Cross River reported one of these.
Table shows the bivariate associations between all potential risk factors and underlying determinants studied and maternal morbidity, indicating a number of promising associations. In addition, in both states, postnatal visits were more common among women who reduced work before the third trimester of pregnancy, who had more antenatal check-ups, who delivered at the health centre, who had healthy attitudes to smoking in pregnancy and who were more likely to know of danger signs in pregnancy. In general, women receiving postnatal visits were better off: they were more likely to have some education, less likely to complain of food insecurity and less likely to live in crowded households.
Bivariate associations between maternal morbidity and potential risk factors
Table shows the final multivariate models for all complications combined. In Bauchi, initial analysis of non-fatal maternal morbidity (pre-eclampsia, sepsis, excessive haemorrhage) showed marked heterogeneity between the minority of women who had a health check up after delivery and the majority who did not. Among those who received a check up, two factors remained in the final model: FGM (ORa 2.10 95%CIca 1.39-3.17) and four or more pregnancies (ORa 1.48, 95%CIca 1.15-1.90). FGM remained in both models in Cross River.
Multivariate analysis of non-fatal maternal morbidity risk factors
Physical IPV during pregnancy showed the strongest association with maternal morbidity in all multivariate models except the small group of Bauchi women who had home visits after delivery. This prominent role remained unchanged when we repeated the analysis using GEE.
Among women who had no home visit after delivery, those who had an unqualified birth attendant (most often to a traditional midwife without government approved training, less often to a neighbour or a family member) were more likely to have complications in both states.
We constructed a compound variable of factors related to the role of a husband or partner in the final model: IPV in pregnancy, IPV in the last year, and report that women had not discussed pregnancy with their husband or partner. Women with all three directly husband-related factors were much more likely to report a pregnancy or birth complication than women who had none, one or two of these factors (ORa 2.39, 95%CIca 1.96-2.92, RD 0.207, 222/432 women with all three and 4,397/14,335 who did not). This association was not explained by any of the factors we could take into account in this study.
Table shows the final models for risk factors for pre-eclampsia and sepsis. Both initial models included the risk factors shown in Table 2. As associations with pre-eclampsia were not significantly different in Bauchi and Cross River, we combined the states for analysis of pre-eclampsia. Four variables showed independent associations after adjusting for the others: IPV in the last year, IPV during the pregnancy in question, rural residence and FGM.
Multivariate analysis of risk factors for pre-eclampsia and sepsis
In the case of sepsis, the variable “state” modified most bivariate measured associations, so we developed a separate multivariate model for Bauchi and Cross River. In Bauchi, sepsis was independently associated with IPV in the last year, IPV in the last pregnancy, perception of being cared for in pregnancy, age of the mother (younger women more likely to suffer sepsis) and FGM (Table ). In Cross River, only two variables remained in the final model, IPV in the last year and perception of being cared for during the pregnancy.
Table shows the low levels of male knowledge of pregnancy and delivery, and the high level of good intentions about maternal risks.
Male knowledge and attitudes about pregnancy and childbirth in Bauchi and Cross River States, Nigeria
Male focus groups discussed what men consider when deciding where a woman should deliver her child. Almost all groups recognized a need for skilled birth attendance, and almost all raised economic considerations in taking advantage of this where it was available. “The man considers the weight of his pocket before deciding where to take the woman for delivery”.
Few of the 180 male focus groups saw men as the cause of IPV; nearly all concluded that IPV could be avoided if women prayed, were obedient and patient, and never refused sex. Asked how IPV could be avoided, several groups suggested increasing women's incomes. The focus groups were uniform in the belief that IPV is a private matter, reporting of IPV bringing shame, disgrace and “greater divisions”. In Cross River, men quoted the Bible (“What God has joined together, let no man put asunder”) as the reason for not reporting IPV. In both states, men gave prominence to community leaders and religious leaders to stop the violence. Despite the strong and uniform belief that IPV is a private matter, many male groups were in favour of locally administered punitive schemes, typically a fine for beating one's wife being a goat, or cash ranging from N500 to N10,000 (US$4-70). Asked what men could do themselves, most groups felt they had the power to stop IPV, “As heads of the households, we can do it”.
A clear theme in the 180 female focus groups was self-blame for the IPV (“strong mouth”, disobedient, demanding or refusing sex). Some concluded that men were “naturally violent so there is nothing you can do”. Others said pregnancy was a cause of violence as it made women irritable and too tired to have sex. They saw marital infidelity as a common cause, whether the woman or man was cheating. Across all regions of both states, women saw money as a major cause. According to women in Cross River, “the Bible says that the wife does not have rights over her body, so we should submit our body to our husbands...” and “the Bible says that God created the woman out of Adam’s rib, the woman should be under the man and should be humble to the man’s relatives to avoid being beaten by the man.” In Cross River, women saw IPV as a family matter, to be resolved at home. In clear contrast, no women's focus group in Bauchi reported this view.