Reasons for Obstetric Scan Among Pregnant Women in Keffi, Nigeria: A Cross-Sectional Study of a Low-Cost Ultrasound Centre

Ultrasound is a safe diagnostic tool used in pregnancy to detect significant causes of foeto-maternal mortality. We explored the reasons for an obstetric scan in 400 randomly selected pregnant women attending a low-cost ultrasound centre in Nigeria. We used a cross-sectional analysis using a structured interviewer-administered questionnaire. The results showed that the participants were aged 27.37±5.96 years, 76% literate, 68.5% Muslims, 50.5% small-scale self-employees, and 31.45% grand-multiparous. The reasons for an obstetric scan were foetal well-being (44.2%), referral by a health worker (34%), gender determination (9.7%), vaginal bleeding (4.8%), confirmation of pregnancy (2.6%), due-date determination (2.5%), foetal weight estimation (2.1%), and foetal number (0.1%). Spouse education with foetal wellbeing (p = .001), gender determination (p = .002), and dating of gestation (p = .002); participants’ occupation with referral by a health worker (p = .026) and confirmation of pregnancy (p = .044); spouse occupation with foetal wellbeing (p = .004) and vaginal bleeding (p < .001); daily household income with referral by a health worker (p < .001) and pregnancy dating (p = .005); and religion with gender determination (p = .007). Health education and improved access to ultrasound services will go a long way in improving maternal/child health.


Introduction
Nigeria is the second country with the worst maternal and child mortalities having daily maternal and under-five deaths of over 145 women and 2,300 children respectively (Odesanya et al., 2015). Globally, Nigeria ranks second with the worst maternal mortality rate of 800 deaths per 100000 live births (Bakari et al., 2015). In Nigeria, emergency obstetric care is available in only 20 per cent of health facilities while skilled birth attendants cover only 35% of deliveries (Akinyemi et al., 2015).
The medical indications for an ultrasound scan in pregnancy depend on the gestational age. In early pregnancy, many studies asserted that confirmation of pregnancy location, viability, duration, and complications were the commonest reasons for an obstetric scan (Kagan et al., 2015). In mid and late gestation, the leading indications are dating, foetal position, wellbeing, growth and malformation, placental condition, and liquor volume assessment (Bakalis et al., 2015;Reddy et al., 2015;Royer, 2015). Postpartum ultrasound can detect retained placenta, uterine rupture, sepsis, and involution, the position of intrauterine contraceptive devices, and assessment and follow-up of congenital malformations (Chapman et al., 2015). Ultrasound is used to inform sophisticated neonatal diagnostic and therapeutic procedures and follow-up (Sananes et al., 2016;Worley et al., 2008).
A growing global non-medical indication for an obstetric ultrasound, especially in developing countries like Nigeria, is determination of foetal sex. This phenomenon could have negative public health implications such as sex selection leading to increased induced or criminal abortions and potential maternal complications including maternal mortality. Inyang-Etoh and Ekanem (2016) demonstrated that about 65% of expectant mothers preferred to have a male child in their index pregnancy for reasons like inheritance, satisfaction of spouse or spouse family members, and consolidation of marriage. Okeke et al. (2015) asserted that 90% of primigravidas wanted an ultrasound scan for sex determination and 62% preferred a male child. Similarly, Larsson et al. (2017) found that 95.8% of 2289 Swedish women had discussed sex determination with the partner before the ultrasound scan, and 57% of them and their partners wanted an ultrasound for foetal sex determination.
However, there is a dearth in the literature regarding the precise motivations behind acquiring an ultrasound. In this study we will further investigate if the intention of an acquiring an ultrasound scan is foetal well-being or sex determination, which could lead to sex selection.

Materials and Methods
The study utilized primary data collected in 2016 using an interviewer-administered questionnaire. The analysis involved descriptive statistics and tests of associations using appropriate statistical tests. A statistical power of 80%, Confidence Interval (CI) of 95%, Level of Significance of 0.05, and Standard Deviation (SD) of .5 were used for sample size selection to adequately control type I and type II errors (Suresh and Chandrashekara, 2012). Participants less than 18 years of age were excluded on ethical and legal grounds. Random sampling was employed to specifically address selection bias that threatens validity and generalization of findings in cross-sectional studies.
The study used primary data only. These are data in their natural states from original sources that have not been collected before. The survey instrument was an interviewer-administered questionnaire. The questionnaire was suitably designed to produce accurate answers. The research questions satisfied the SMART(er) and FINER criteria (Aslam and Emmanuel, 2010;Riva et al., 2012). They were specific, measurable, achievable, relevant, time-bound, feasible, interesting, novel, ethical, and could be evaluated and revised. They were straightforward and easy to translate into Hausa local language and can be administered by research assistants (Jacobsen, 2014). Additionally, the questions were based on evidence from previous studies and contributions from classmates and the instructor (Farrugia et al., 2010). Finally, the questions could fill in the identified gaps in the body knowledge on the topic (Robinson et al., 2011). The questionnaire had sections on socio-demographics and the reasons for an obstetric ultrasound.
The study was conducted in Hamamah Diagnostic Services Centre (HDSC), Keffi, Nigeria. Keffi is a town 50km from Abuja, the capital city of Nigeria. It is the headquarters of Keffi LGA of Nasarawa State. It is a cosmopolitan settlement because of its proximity to Abuja. HDSC was selected because of its high patronage by pregnant women, the study population of interest that forms a sample frame for recruiting participants. It is the cheapest ultrasound centre in the locality; it charges less than $2 per ultrasound scan compared to around $10 in other centres in the locality.
The minimum sample size of 384 was selected using a CI of 95%, Level of Significance of 0.05, and assumed SD of 0.5. Other methods were not used since the population of targeted pregnant women in rural areas was not known.
Primary data was collected using a structured interviewer-administered questionnaire. The interviewer-administration was chosen to maximize participation. The questionnaire was written in the English language only but translated verbally into Hausa local language where necessary. The approach allowed the participants to brainstorm the questions and establish a rapport (Thompson et al., 2015). It also enabled a quick collection of a large amount of data on several parameters at a low cost. A feasibility study of available medical records at HDSC indicated that 400 samples were achievable. Two research assistants were used for data collection only. The data was collected over six weeks during the study period.
The questionnaire had quantitative closed-ended queries that aided a systematic analysis. Firstly, the answers were coded numerically to permit codebook entries. Secondly, the data was directly entered into SPSS. Thirdly, Double-entry was done to ensure data accuracy. Fourthly, the data was cleaned to correct for typographical and other errors. Finally, necessary data re-coding was done during analysis.
The dataset had independent and dependent variables that were nominal, ordinal or continuous. The independent variables were the socio-demographic profile of the participants, while the dependent variables were the reasons for an ultrasound scan. The socio-demographic factors included age, gravidity, education, spouse education, occupation, spouse occupation, daily household income level, and religion.
The analysis initially involved a univariate descriptive analysis using frequencies and percentages to show the distribution of responses with appropriate tables, charts, or graphs. Additionally, measures of central tendency and dispersion were used to analyse the quantitative variables. The socio-demographic factors were tested for possible relationships using bivariate correlation. X 2 -Test was used for categorical variables and logistic regression to control for confounders.
The reasons for an obstetric scan (Table 2) were foetal well-being (44.2%), referral by a health worker (34%), gender determination (9.7%), and vaginal bleeding (4.8%). Other reasons included confirmation of pregnancy (2.6%), due date determination (2.5%), foetal weight estimation (2.1%), and lastly foetal number (0.1%). X 2 -Test revealed statistically significant associations between the socio-demographic variables and reasons for a scan. Gravidity was a predictor of vaginal bleeding (p = .041) and confirmation of pregnancy (p = .040) while the educational level of respondents predicted foetal wellbeing (p = .040) and referral by a health worker (p = .041). Similarly, there were associations between spouse education and foetal wellbeing (p = .001), gender determination (p = .002), and dating of gestation (p = .002). Occupation was related to referral by a health worker (p = .026) and confirmation of pregnancy (p = .044). Spouse occupation was a determinant of foetal wellbeing (p = .004) and vaginal bleeding (p < .001). Daily household income was associated with a referral by a health worker (p < .001) and pregnancy dating (p = .005). Finally, religion was a predictor of foetal gender determination (Phi=.136, Odds Ratio [OR] = 2.051, p = .007).
The occupation of the participants was a significant predictor of referral and confirmation of pregnancy as reasons for an ultrasound (p < .05). The effect size was small in both cases (Cramer's V [CV] =.154, p = .023 versus CV=.164, p = .013 respectively). Housewives and other occupations were twice more likely to do a scan because of a referral than the unemployed and employed respondents (Table 3). Housewives were twice more likely to scan for confirmation of pregnancy than women in small-scale self-employed jobs while the employed and unemployed were twice more likely than housewives. The effect was highly concentrated at the employed class (Standardized Residuals [SR] = 2.4). Both foetal well-being and vaginal bleeding were significantly predicted by spouse occupation (X 2 (2, N=400) = 11.221, p = .004; and X 2 (2, N=400) = 17.299, p < .001 respectively). The impact was more on vaginal bleeding (CV = 208, p < .001 against CV=.167, p = .004). The significance (SR = 2.2) lies in the wives of unemployed husbands who were less likely to request a scan for foetal wellbeing than those of the employed ( Table 3).
The unemployed spouse category had the highest percentage of 37.5% and other occupation the least. Participants of unemployed spouses were thrice more likely to do a scan for vaginal bleeding than those in the other two categories. The difference was in the unemployed category (SR = 3.9).
Daily household income as a measure of poverty was significantly related to referral by a health worker and dating of pregnancy. The p values were very low in both cases (p = .001-.005) but the effect size small p = .001). The percentages of referrals increased down the ladder of income level (Table 3). Both effects were concentrated around those households living above poverty line (SR = 2.8 versus 3.1 respectively). The likelihood of scanning for dating pregnancy increased with a rising income level.
There was a significant relationship between religion and doing a scan for gender determination (X 2 (1, N=400) = 7.367, SR = 2, OR = 2, p = .007). The effect size was small (Phi=.136, p = .007). The percentage of Christians interested in foetal gender determination was 25% as against 14% for Muslims (Table 3). The practical implication was that for every probability of Muslims requesting for foetal gender; there was a twice likelihood for Christians. This trend might not be unconnected with male-dominant inheritance practices among Nigerian Christians in a patriarchal society.

Discussion
In this study, the main reason for doing an obstetric scan was foetal well-being (44.2%), followed by referral by a healthcare worker (34.0%), gender determination (9.7%), vaginal bleeding, (4.8%), confirmation of pregnancy (2.6%), dating pregnancy (2.5%), and foetal weight estimation (2.1%). This is an important finding because the safety of the foetus is a positive effect of ultrasound screening while sex determination can potentially lead to more harmful practices like sex selection.
These findings are consistent in many respects with previous studies. Enakpene et al. (2009) found foetal viability and foetal gender determination as the key reasons for doing ultrasound in pregnancy. On the other hand, in Ikeako et al. (2014) study, foetal gender determination (17.8%) and foetal position (15.4%) were the leading reasons for an obstetric scan. In contrast, Oboro et al. (2015) reported pregnancy confirmation (87.5%), foetal presentations (70.3%), location of pregnancy (66.8%), confirmation of miscarriage (66.6%), foetal number (65.3%), and foetal growth restriction (60.0%) as the most common reasons for obstetric scans as described by Nigerian health workers. These differences may be connected with the participants and the study settings. Oboro et al. (2015) targeted health workers, while the other two studies focused on pregnant women in urban areas. However, this study concentrated on pregnant women residing in rural areas assessing low-cost obstetric ultrasound services.
Only 0.1% of respondents did a scan for foetal number contrary to the findings of Oboro et al. (2015) where 65.3% of respondents gave foetal number determination as a reason of doing an obstetric ultrasound. This difference might be due to the high incidence of twin pregnancy among the Yoruba ethnicity of the southwest Nigeria (Aduloju et al. 2015), the setting for the Oboro et al. (2015) study.
Socio-demographic characteristics have been known to influence knowledge, attitude, and practice (Rollins et al., 2016). This study found significant associations between socio-demographic variables and reasons for a scan. Participants' occupation and daily household income, religion and spouse occupation were determined to be significant predictors of referral by a health worker, gender determination, and vaginal bleeding as reasons for a scan respectively. Participants' occupation was a predictor of choosing HDSC because of low cost and quality of service while spouse occupation was a predictor of recommendation by a friend or relative as reasons for choosing HDSC. Age, gravidity, spouse education, religion, participants' and spouse occupation were significantly associated with distance travelled while daily household income level and religion were significant predictors of satisfaction level.
Other studies had established some of these relationships. Enakpene et al. (2009) found that age, occupation, level of education were significant predictors of foetal viability on X 2 -Test but not on logistic regression. However, in this study educational background of participants and their spouse, and spouse occupation were significant predictors on bivariate analysis with spouse education not retained by logistic regression but retained by hierarchical multiple regression (R 2 change=30%, β=.205, CI = .016-.059. F(1,396) = 12.179, p = .001).
Foetal gender determination was predicted by spouse education and religion but only religion retained on both logistic regression (Muslims versus Christians, β = -.706, OR = .494. CI = 247-.987, p = .046) and hierarchical model (R 2 Change = 20%, β=.141, CI = .035-.196, F(1,397) = 7.966, p = .005). In contrast, Enakpene et al. (2009) found age, religion, occupation and gravidity as predictors of gender determination on X 2 -Test while only gravidity and occupation remained statistically significant in the multivariate model. In both studies, Christians were more likely to request a scan for gender determination than Muslims, though none of these studies explored the emerging issue of sex selection abortive practices. The study by Enakpene et al. (2009) did not go further to analyse the predictors of the other reasons, but this study analysed the predictors of all the outcome variables.
The financial burden of foetal gender determination is mostly covered by out-of-pocket payment and not funded by the government or any health insurance schemes. The ultrasound request for foetal gender is prevalent in Nigeria at the expense of serious pregnancy-related risk assessment for due dates, foetal well-being, and antepartum bleeding. This trend prolongs the waiting time for obstetric ultrasound services with reports of massive haemorrhage and even deliveries happening during waiting time (Bindman et al., 1991;Umar et al., 2011). There is a need to address this situation in the light of ethics, morality, cultural norms, and social justice (Ohagwu et al., 2014).

Conclusion
The study had achieved its objectives of discovering the reasons of seeking ultrasound services by pregnant women attending HDSC Keffi and testing associations between the socio-demographic characteristics of the participants against the reasons for a scan.
The primary reasons for a scan were to assess foetal well-being, foetal gender, and because of vaginal bleeding. Christians were more likely to check foetal gender than Muslims. The association between gravidity and vaginal bleeding calls for the need for governments and non-governmental organizations to provide more support for family planning services to prevent fatal antepartum and postpartum haemorrhage associated with grand multiparity.
The study clearly demonstrated the role of educational level of both pregnant women and their spouses in predicting foetal wellbeing. Improved maternal and spouse education will improve the wellbeing of not only foetuses, but also the general wellbeing of all children, mothers, and the community at large. Hence the need for more campaigns and opportunities for education such as scholarships and grants to disadvantaged populations in fighting foetal and maternal morbidity and mortality.
The association between occupation and referral by a health worker suggests the role of occupational groups in enhancing access to healthcare services. Therefore, public health campaign organizations should involve occupational groups in disseminating health information. Market women in Africa can be used as public health partners in disseminating health information on maternal and child health. Knowledge is empowering.