Women’s determinant factors for preferred place of delivery in Dodoma region Tanzania: a cross sectional study

The study was conducted from November 2013 to June 2014 in Dodoma Region, Tanzania.Administratively; Dodoma Region is divided into seven districts namely Kondoa, Bahi, Mpwapwa, Kongwa, Chamwino, Chemba and Dodoma Urban. The study was conducted in Dodoma Urban, Kondoa, Kongwa and Chamwino districts.

According to the 2012 Nation Census [11], the population of Dodoma Region was 2,083,588. The Average Annual Intercensual Growth Rate is 2.1% and population density is 50 people per sq.km. The percentage of women of reproductive age is 44.1% of the total female population in Dodoma Region. The Region has a total of seven hospitals (5 public; 2 private), 32Health Centres (26 public; 6 private) and 269 dispensaries.

The study employed a cross sectional survey design, targeting women of reproductive age (15–49 years old) who had at least one child in the past five years. Women who were sick or had mental health problems and those who were not willing to participate were excluded from the study.

The sample of 800 women was determined using aformula (N = z² p (1-p)/e²) where N = Sample size; Z = Z score for 95% confidence level, which is 1.96; p = Prevalence in previous studies (percentage of women delivered at the health facility in Dodoma was 45.1%); e = tolerable error (set at 5%). The estimated sample size was N = 380, 10% was added due to non-response fear and multiplied by a design effect of 2 due to the multistage nature of the sampling method. Therefore, the minimum sample size based on the formula was 798, but a total of 800 participants were studied.

A multistage sampling technique was used to select sample units from the targeted population. In the initial stage, the list of all districts in Dodoma Region was obtained. Dodoma Municipality was selected purposively because it is an urban district with people of different backgrounds. Three other districts were selected randomly namely, Kondoa, Chamwino and Kongwa. In the second stage, two sub districts called (wards) were randomly selected from each of the four selected districts using a table of random numbers and a total of eight wards were obtained. In the third stage, two villages/streets were selected randomly from each of the eight wards using a table of random numbers, making a total of 16 villages/streets. From each of the 16 selected village/streets, all households with women who had children below five years of age were eligible for the study. Therefore, from each of 16 selected villages/streets, 50 participants were selected from the randomly selected households. Within the households, only one respondent was selected. In the households where there were more than one woman qualifying to participate in the study, a woman who had the youngest child was selected assuming she could remember most of the events that occurred during child delivery.

Structured questionnaire with open- and closed-ended questions was prepared in English and translated to Kiswahili. This was administered by two research assistants who collected data via face to face interview techniques. Both research assistants were graduates working as primary school teachers, and underwent training for two days before data collection. Data collection tools were pre-tested to ensure consistency of the variables included in the study. Data recorded was cleaned, edited and coded before data analysis. However, data exploration was done to make sure of its suitability for analysis.

A total of 800 women with children less than five years old were recruited and participated in the study from the four selected districts. The majority of respondents 499 (62.4%) were Gogo by tribe, 193 (24.1%) were Rangi and 108(13.5%) were other tribes. Age-wise, 531 (67.3%) respondents were between 20 and 34 years old; 209 (26.5%) were between 35 and 49 years; 49 (6.1%) were Teenagers between 15 and 19 years.

In terms of marital status, majority of respondents 438 (54.9%) were married, 243 (30.5%) cohabited, 69 (8.6%) were single and 48 (6.0%) were separated/widowed/divorced.Regarding number of children, 171 (21.4%) respondents had 1–2 children, 439 (54.8%) had 3–4 and 190 (23.8%) had five or more children. Results show that 148 (18.5%) women in Dodoma Region had no formal education. A high proportion of women, 562 (70.3%) had primary education while a smaller proportion, about 89(11.1%) had secondary and tertiary education. About 633 (79. 1%) respondents were peasants and 167(20.9%) had formal employment/self employment (Table 1 ).

More than three quarters of the respondents 629 (78.6%) reported to have delivered at the health facilities (dispensaries, health centres and hospitals), while 171 (21.4%) delivered at home/on the way to health facility. Reasons for delivering at home include: abrupt occurrence of labour pain 89(52%); long distance to the health facilities 43 (25.1%); transport unavailable 23 (13.5%); lack of money to pay for transport 9 (5.3%) and unfriendly experience with the health care providers 7(4.1%). Results also showed that majority of women 617(77.1%) were assisted by skilled health care providers during delivery, 135(16.9) by traditional birth attendants, 26(3.2) by relatives and 22(2.8) had no assistance. Some women 12(1.5%) delivered at the health facilities but ended-up with a TBA’s assistance. Majority of respondents 502 (62.8%) reported that both husband and wife decided on the place of delivery, 195(24.4%) reported making the decision on their own, 65(8.1%) reported husbands making the decision and 38(4.8%) reported the relatives made the decision. Regarding estimated distance to the nearby health facilities, 696 (87%) live within five kilometres while 104 (13%) live more than 5 km (Table 2).

As stated in the methodology section, univariate analysis based on simple logistic regression model was first used to identify the predictors of the preferred place of delivery. The results of the model presented in Table 3 indicated that number of children (p < 0.000; mother’s education level (p = 0.0126), estimated distance to the nearby health facility (p < 0.0001), occupation (p = 0.0022) and cost of transport (p < 0.0001), were strong predictors of the preferred place to delivery. The results have shown that the preference to deliver at the health facility is inversely related with the number of children a woman has. For women with 1–2 children, the odds of choosing health facility was almost 3 times that of women with at least 5 children (OR = 3.043, p < 0.0001). Compared to women with secondary education or higher, those with primary were reported to be significantly less likely to prefer health facility as place to deliver (OR = 0.359, p = 0.0051). However, the odds of the preference to deliver at the health facility among women with 3–4 children was about 2 times significantly greater than that of women with 5 or more children (OR = 1.908, p = 0.0010). With respect to distance to the nearby health facility, the results revealed that women stayed with 5 km (OR = 2.977, p < 0.0001) were significantly more prevalent to prefer to deliver at the health facility in comparison to women who stayed at distance of above 5 km. Occupation of the mother was another predictor of the preferred place of delivery, with employed women havinga significantly higher chance of preferring the health facility as the placed to deliver than peasant women (OR = 2.159, 0.0022).Regarding the cost of transport to the health facility, the results have shown that the preference to deliver at the health facility was significantly lower among women who paid for transport compared to those who did not pay for transport (OR = 0.239, <0.0001).

The results of the multivariable logistic regression model displayed in Table 4 showed that, accounting for possible confounder the effect of mother education (p = 0.8813), estimated distance to the nearby health facility(p = 0.4666), and occupation (p = 0.2921) on the preferred place of delivery were no longer significant. On the other hand, the number of children that a woman has (p = 0.0078) and transport cost (p < 0.0001) were the significantly potential predictors of the preferred place of delivery. The results of the multivariable logistic regression model indicated that women with 1–2 (aOR = 2.437, p = 0.0025) and 3–4 (aOR = 1.554, p = 0.0350) children were significantly more likely to prefer health facility as a place to deliver than women with at least 5 children. With regard to cost of transport to the health facility, the results confirmed that the preference for delivering at the health facility was significantly lower among women who paid for transport compared to those who did not pay for transport (OR = 0.270, <0.0001).

The findings from this study indicate that more than three quarters, (78.6%) of respondents had delivered at the health facilities in their last pregnancy. This result is consistent with the findings of studies conducted elsewhere. For instance, a study conducted across the three ecological zones of Ghana showed that75.8% of women from Dodowa, 61.6% from Kintampo, and 89.3% from Navrongo delivered at the health facilities [12]. Another study conducted in the Democratic Republic of Congo showed that high proportion of women (93.8%) delivered at the health facilities [13].

The Kenya Demographic and Health Survey report [14] has also shown that the proportion of women who delivered at the health facilities increased from 43% in 2008–2009 to 61% in 2014. Again, a similar finding was observed from a community-based cross-sectional study conducted in Nepal which showed the changing trends of place of delivery. It was found that within five years’ time interval, the percentage of women who delivered at the health facilities increased from 45.13% to 63.88% [15]. The reasons reported for the increase of institutional delivery were the implementation of various programs along with safe motherhood, free services for institutional delivery, Safe Delivery Incentive Program (SDIP) and establishment of birthing centers in rural areas [12‐15].

The 2015–16 TDHS-MIS [9] result showed that the proportion of respondents who delivered at the health facility in the study area is 69%, an increase from 45.1% in TDHS 2010 [16]. The noted increase in health facility delivery in Dodoma Region might be due to the implementation of various reproductive and child health interventions in the community such as improved primary health facilities with maternity health care services and fee exemption for maternal health services [17]. Other contributing factors include on-going health promotion campaigns through media, health care providers and community leaders.

Reasons mentioned by those women who delivered at home such as abrupt labor pain, long distance to the health facilities, unavailability of transport, lack of money to pay and unfriendly experience with health care providers are similar to those given by women in the existing studies [18‐20]. The fact that those women who delivered at home are not attended by skilled health professionals may contribute to high maternal morbidity and mortality. However, this study also showed that some women delivered at the health facilities but ended-up with a TBA’s assistance. This may be due to shortage of health care providers at the health facilities especially in the rural areas. All women need access to skilled care during childbirth as timely management and treatment can make the difference between life and death [21]. The government through the Ministry of Health, Community Development, Gender,the Elderly and Children (MoHCDGEC) need to increase the number of well-trained health workers in both rural and urban areas.

The results of the simple logistic regression model indicate that mothers’ education level, number of children, cost of transport, the estimated distance to the nearby health facility and occupation, were strong predictors of the preferred place of delivery. However, after controlling the potential confounder the multivariable logistic regression model demonstrated a significant association between a delivery at the health facility and the number of children and transport cost.

With regard to the education level of a woman, it showed that the likelihood of delivering at the health facility increases with an increase in mother’s education level. Women with secondary education and above were more likely to deliver at health facility compared to those with primary education. This finding is consistent with results of previous studies conducted in other countries [22‐25]. The increased likelihood of delivering at the health facility with increase in mother’s education level could be related to increased access to right information concerning maternal health as well as increased women’s autonomy in decision making on the preferred place of delivery. Efforts to improve women literacy should be intensified.

With regard to the number of children, the study results have shown that the likelihood of a woman to deliver at the health facility decreases with an increase in the number of children. This is consistent with the findings of other studies which show that the likelihood of women delivering at the health facility decreased consistently as the number of children ever born increased [26‐30]. Perhaps, the likelihood of women with high parity to deliver at the health facility decreased due to increased family responsibility because of large family size which can limit both time and money to seek hospital delivery and other reproductive health care services. Another possible explanation could be due to previous experiences in maternity health services. The attitude of health care providers towards women is a major influence on women’s decision-making about whether or not to use a particular type of maternal health care service. Health care providers need to understand that the manner in which they interact with women, particularly those with high parity, impacts maternal health care utilization, depending on whether a woman perceives her experience as positive or negative. As such, health care providers working in RCH clinics and labour wards need to urgently address care issues in order to improve services provided. Respectful care and education needs to be customized for multiparous women. This care must be free from misconceptions and negative attitudes regarding multi-parity.

The cost of transport and the estimated distance to nearby health facility was inversely related to women’s preferred place of delivery. The finding of this study showed that women who paid transport cost to the health facility and those who stayed long distance to nearby health facilities were less likely to deliver at the health facilities compared to those who did not pay and stayed nearby health facilities. This finding is similar to those from other studies which showed that transport cost and long distances from the household to the health facilities limit women from delivering at the health facilities [31‐33]. A study conducted in Kongwa District showed that majority of women either walked or used carts or donkey as a means of transport to the District Hospital due to high transport cost [30]. These challenging conditions reduce the ability of poor people to access maternal health services, which may be contributing to high maternal and child mortality experienced in Tanzania and other developing countries. Though availability of emergency transport and immediate referral services could increase the access to and use of delivery care services by women, often than not there are cost implication as families are compelled to contribute for fuel. The permanent solution is for the government through the MoHCDGEC to increase the number of health facilities including maternity waiting homes in both rural and urban areas.

The results of this study showed that the likelihood of women to deliver at the health facility was higher among women who were employed compared to peasants. This finding is consistent with results of previous study which showed that employed women were more likely to have greater knowledge about pregnancy and childbirth due to freedom of movement and interaction with other people outside the homestead [34, 35]. Employment status can increase women’s economic autonomy and decision making power on preferred place of delivery.

The study employed cross sectional survey design. The cross-sectional nature of the data does not allow making causal inferences about the relationship between delivery place and the determinant factorsfor preferred place of delivery. However, the study was also conducted in four district of Dodoma Region and findings might not reflect the situation of the rest of the country due to the contextual variations in availability and accessibility of maternal health care services.