Maternal and newborn healthcare practices: assessment of the uptake of lifesaving services in Hoima District, Uganda

The study was carried out in Hoima District of Uganda, in Bunyoro sub-region [12]. In 2012, the mid-year district population was estimated at 548,800 [13]. It is made up of 13 sub-counties, most of which are rural, with the urban centre based in the Hoima municipality, which is also the main trading centre of the district. A total of 31.9% of adults over 18 years old are illiterate, with 63% of the working population involved in agriculture and fishing.

The study was a cross-sectional quantitative study. Women who had a child aged 0–59 months were included in the study. Once the woman gave an affirmative response about having a child aged 0–59 months, only those who consented participated in the study. The study enrolled 691 women, and women with multiple children within the selected age range were included only once into the study.

Through simple random sampling, households were visited, and household heads were asked for permission to interview the woman. The data was collected using a questionnaire adapted from the Emergency Nutrition Assessment (ENA) household [14] tool with aspects adapted from the WHO signal functions for maternal and child health care tool [15]. The tools were translated into the Runyoro Language, and counter translated back into English to ensure that meaning is not lost. These tools were then pretested in one community in Hoima during the training process of the data collectors to ensure all data collectors can interpret the questions the same way. Any error in the tools were corrected before the actual field data collection.

A team of 10 experienced research assistants recruited from Hoima and who were degree holders, fluent in spoken and written Runyoro and English languages collected the data. They were trained by the investigator for 3 days in ethical data collection, aspects of pregnancy and childbirth, seeking consent from households and the entire methodology for the study.

Following the sampling frame, all mothers with a child presently 0–59 months were included in the study. Those who were currently pregnant, lost their child or only had a child older than 59 months, were excluded in the study. At the household level, a team of two assistants facilitated the data collection process, with one acting as translator. The tool was uploaded on a mobile platform, the Open Data Kit (ODK) [16], and data was collected directly online. A list of the villages in the selected sub-counties was generated and randomized using ENA for SMART software. Households in each cluster (village) were selected using the EPI methodology [17], which is a survey in 30 systematically selected clusters of seven or more children to estimate the immunization coverage.

Five teams with two supervisors collected and checked the quality of the data. To ensure quality of data, the statistician on the team monitored the data on the platform as it was being uploaded. Any errors noticed would quickly be relayed to the field team through the supervisor for correction. Table 1 shows the study variable.

Data was downloaded from the ODK into Excel. As the data was being collected, each variable was coded independently with a yes or no response. Therefore, where multiple variables contributed to a given indicator, cleaning and alignment were done in Excel. Since the study focused on cross-tabulation of quantitative data through frequencies to deduce percentages, the data was imported into SPSS for analysis. Any missing data meant elimination of that variable in the data going to be analysed since the information followed a continuum of care approach for Maternal Newborn and Child Health [3].

Income sources of the household were categorised as permanent or temporary job, or private business with other forms also noted. The availability of the child health and ANC cards served as home-based evidence that particular services had been given to the woman or the baby. Sex and age of the child were also assessed.

ANC attendance focussed on the percentage of pregnant women attending four or more visits and having had their initial visit in the first trimester [18]. These parameters were considered sufficient for the uptake of ANC [19].

The quality of Goal-Oriented ANC services [20] variables selected included those that can be acquired from the ANC card or the mother’s passport and those that assesses the mothers’ knowledge. This therefore focussed on the percentages of women who received 5 quality services for the most recent pregnancy. This meant that the women: 1) received iron-folic acid supplements at least once; 2) could recite the 7 danger signs during pregnancy (analysis placed emphasis on seven categories of danger signs [21, 22]: vaginal bleeding, fever, swollen face or extremities, tiredness or breathlessness, headache or blurred vision, convulsions, and reduced or absent foetal movement); 3) was given Fansidar as intermittent presumptive treatment of malaria (IPT) [23]; 4) claimed to have been able to acquire and store the 6 essential commodities of the birth preparedness and complication readiness; and 5) had been tested for HIV [16]. To be able to qualify for this study, a woman had to have a birth preparedness and complication readiness, emergency transport, money saved, sterile blade/scissors for the cord, gloves, cotton wool, and clothes for the baby [24].

The essential newborn care practices concerned whether the women received all 7 essential newborn care services as stipulated by the Ministry of Health (MoH) during the birth and immediate newborn period [18] for the most recent birth including if: 1) childbirth was performed at the health facility; 2) childbirth was using a clean birthing kit; 3) childbirth was attended by a doctor, midwife or nurse [25]; 4) placing a baby on skin to skin contact with the women immediately after birth; 5) initiating breasting while still in the labour suite; 6) women and baby stayed at the health facility for 24 h after childbirth; and 7) women claimed to know all six danger signs of the newborn. These newborn danger signs included: difficulty in breathing, not suckling well, high temperature, infectious cord, unclear eyes, and yellow skin colour [26].

Analysis of the level of paternal involvement in maternal and newborn care focused on 6 key variables as recommended [27], selected as practices that men could engage in to show participation in maternal and newborn health, based on the woman’s statement that the man: 1) did some household chores, 2) provided money for health care needs, 3) paid a house help because the woman could not engage in heavy household work, 4) reminded and encouraged the woman to attend her hospital visits, 5) accompanied her for at least one visit to the hospital and 6) had an HIV test together.

For the 3 outcome variables, a chi-square test to examine the association between household income source and the key variables was carried out.

The contribution of the MoH in Uganda to the reduction of maternal and newborn morbidity and mortality is guided by the set targets in the Uganda maternal newborn and child health sharpened plan [3]. At the end of every 5 years of health care implementation, the Ugandan government commissions a nationwide health and demographic status survey [2] to ascertain the level of the services uptake. A comparison of the study results with the national targets set by the sharpened plan and the current national status provided by the Uganda Demographic and Health Survey (UDHS) was also done.

Of the children included in the sample, 50.2% were male and 49.8% were female (see Table 2). The mean age of the children under 5 years included in the study was 28.92 months (SD = 15.60).

Regarding the income for the man the households depend on, 35.7% of the household’s source of income was categorized as other, which included depending on seasonal farming and donations from relatives and friends. Even though 29.3% of households depended on a monthly salary as income, only 5.1% were permanent jobs, with the rest being temporary, with uncertainty of continuity. More than a quarter (35%) of the households depended on small businesses.

In total, 92.8% of women had received a Child Health card from the HF, and 71.5% of women presented the card to the study team. Even though 84.8% of the women had an ANC card at the time of study, 5.2% of the women did not have both the ANC card and the Child Health card.

Results from Table 3 show that only 24.3% of the women had attended their first ANC in the first trimester, implying that 75.7% of the women in the Hoima District missed out on early uptake of the ANC services offered. Women who had four or more ANC visits during the period of pregnancy were 55.1%. Even though the rest had one, two or three visits, 11.1% of the women had never attended any ANC visit throughout their most recent pregnancy.

Only 0.4% of women presented evidence of having received all five goal-oriented ANC services. Almost all women (91%) confirmed that they had received at least one to two of the services, with HIV counselling and testing being the most frequently provided service (91%), IPT being the second service (78.9%), and iron-folic acid supplementation contributing only 28.8%.

Goal-oriented ANC scored low because only 13.3% of the women reported having secured all six components of the birth plan. More than half (62.1%) reported having secured at least four to five of the six components mentioned.

Only 3% was able to describe all seven danger signs during pregnancy. It is important to note that 45.3% of women could only mention three to four danger signs, and 23.3% of the women could point out five to six danger signs.

Results in Table 4 show that only 0.5% of the women benefited from all 7 ENC practices in their most recent childbirth. A total of 20.1% of the women never practised any of the 7 ENC practices. The majority of the women (79.9%) reported having practised at least one or two ENC, and these were mostly childbirth at the HF and skilled birth attendance.

The study also examined how each of the seven actions was practised by the women. However, 20% of the women interviewed had not given birth at the HF, and the same percentage was not assisted by a skilled birth attendant. For those who delivered at the HF, 95.2% practised skin to skin contact, 89% used a clean birthing kit, but only 59.8% of the women reported having initiated breastfeeding within the first hour of life. For length of stay at the HF after delivery, 24.9% of the women were discharged from the HF within 6 h after birth, and another 34.9% of the women were discharged within 24 h. Less than a quarter (20.1%) of women who delivered at the HF were discharged beyond 24 h after delivery.

Only 2.4% of women understood and recalled all the six danger signs of the newborn. Individual examination of the danger signs was not very successful, with identification of an infectious cord scoring 14.8%, followed by baby not suckling well with 13.4%, and identification of yellow skin scoring the lowest with 7.7%.

Results from the study as shown in Table 5 indicate that only 11.1% of the women declared that the man supported them with all 6 aspects. At least 46.7% reported that the man was involved in one to three aspects during pregnancy and the immediate newborn period. It is important to note that 21.1% of the women reported that their male partners did not provide any support at all. Of the 6 components, allocation of money to the woman for healthcare needs stood out as the most widely practised behaviour (63.7%). The study further revealed that 52.5% of women had their male partners take time off from their schedules to assist with household work and child care as the pregnant woman rested, and 49.2% of the woman confirmed that they had a household helper during the newborn period, paid for by the male counterpart.

Of the 6 components, 3 focused on practices that male counterparts should do to contribute to goal-oriented ANC care. Women reported that only 39.9% of their male counterparts encouraged them to actually seek care focusing on the ANC visits. This was further confirmed by the low levels of male counterparts (36.9%) accompanying their pregnant women to ANC. Even though the MoH guidelines encourage couple counselling and testing for HIV, to accelerate the reduction of woman to child transition of HIV, only 26.2% of male counterparts took an HIV test with their pregnant women during ANC.

Table 6 shows the percentages of participants that meet all indicators for the three key variables, according to household income source. The chi-square tests show that household income source significantly predicts male involvement, with males being more often involved in the pregnancy and newborn care if the income comes from a small business or temporary job, compared to households where the income comes from a permanent job or other sources.

As shown in Fig. 1, in the Hoima District, earlier first attendance at ANC is 26% below the national target, but only 5% below the national achievement (UDHS), meaning that most woman attend their first ANC visit after 12 weeks of gestation. The Hoima District has a low 4th ANC attendance percentage, which is 5% below the national average and 25% below the Ugandan national target. The Hoima District PNC practices within the first 48 h after birth are 50% below the national target and 34% below the national practice. The Hoima District early initiation of breastfeeding is 20% below the national target but only 6% below the national average.