This study was carried out in Bule Hora Town, which is located West Guji Zone of the Oromia region in the lowlands of southern Ethiopia. According to the West Guji Zone health and demographic surveillance department dataset, this is derived from the 2007 census by 2.4% inflation. ¹⁴ The town, which is located 465 kilometers from the capital city of Addis Ababa along the main connecting road, is under zonal administration and has a total population of 157,148. The town is characterized by an urban setting, a mostly agrarian economy, and a population that is mostly from the Oromo ethnic group, namely the Guji agro-pastoral community. A facility-based unmatched case-control study design was employed. To address the determinants of early antenatal care initiation, the study followed specific procedures conducted between June and July 2025.

The study population consisted of pregnant women who attended ANC at Bule Hora University Teaching Hospital during the study period and were randomly selected. Cases were defined as women who initiated ANC before 12 weeks of pregnancy (early initiators), while controls were those who started ANC at or after 12 weeks (late initiators). ¹⁵ Eligible participants were women aged 18 and older, as well as mature minors under 18, who had resided in the study area for more than six months. Women with unknown gestational ages at their first visit, hearing or speech impairments, serious pregnancy complications, or mental illness were excluded from the study. The study participants were approached by the principal investigator and interviewers with extensive experience in quantitative research.

Dependent/Outcome variable: Early ANC initiation, defined as initiation before 12 weeks of pregnancy (early initiators/cases = 1), versus initiation at or after 12 weeks (late initiators/controls = 0).

Independent/Exposure variables: Sociodemographic factors (age, residence, marital status, personal and husband’s education and occupation, income, and distance to a health facility); knowledge-related factors (awareness of ANC importance, timing, number of visits, services, danger signs, and information sources); individual and obstetric factors (parity, gravidity, previous and current ANC timing, delivery location, history of abortion/stillbirth, past ANC use, family planning, pregnancy planning, complications, and transportation access); and finally, health facility and cultural factors (facility proximity, service costs, drug availability, waiting times, provider attitude, husband accompaniment, community support and norms, and cultural barriers).

The sample size was calculated using a double population proportion formula in Epi-Info version 7.2.2.6, based on a key determinant from prior literature which is educational level (Exposure proportion: 35.7% among controls (late ANC initiators after 12 weeks). ¹⁶ The first calculation produced 416 participants with a power of 90%, a 95% confidence level, a case-to-control ratio of 1:3, an exposure proportion of 35.7% among controls, and an adjusted odds ratio of 2.15. The final sample size was 456 (114 cases and 342 controls) after accounting for a 10% non-response contingency. Simple random sampling was used to choose study participants from the hospital’s antenatal care register using computer-generated random numbers generator to select eligible pregnant women.

Questionnaires were refined before deployment on KoboToolbox. After deployment on digital KoboToolbox. The questionnaire was pretested among 23 (5%) participants at a neighboring health facility before the actual data collection to ensure clarity, consistency, and applicability of the tool to ensure tool use, data accuracy, and ethical procedures for two days. Internal consistency of the knowledge assessment items was evaluated using Cronbach’s alpha coefficient, and the reliability value was found to be acceptable (Cronbach’s alpha = 0.78). Built-in validation checks (such as skip logic, range limits, and required fields) were applied to minimize entry errors. Supervisors reviewed submitted forms daily for completeness and consistency, whereas the principal investigator performs random cross-checks and downloaded the data periodically to identify anomalies or missing values. After collection, all data was cleaned, verified for completeness, and exported to statistical software for final consistency checks before analysis.

The data was collected using face-to-face interviews from pregnant women who can provide informed consent using Kobocollect app. The questionnaire was adapted from Ethiopian Demographic and Health Survey and different related literature. ^{17 , 18} Informed consent was obtained before conducting the interviews. The data collectors are professionally Nurses or health officer, or midwives have experience of working in the ANC unit from other facilities were participated and one supervisor was recruited from zonal health office.

All complete questionnaires were checked for accuracy, consistency, and completeness before being entered into KoboToolbox and exported to SPSS version 27 for analysis. To find factors linked to Early ANC initiation, descriptive analysis like percentage, frequency and mean was computed to show the data characteristics.

Binary and multivariable logistic regression analysis was used to identify associations between dependent and independent variables. Variables associated with early ANC initiation at p < 0.25 in bivariable analysis, together with variables supported by previous literature and biological plausibility, were considered for multivariable logistic regression analysis. To minimize model overfitting, the number of predictors included in the final model was evaluated considering the events-per-variable recommendation for logistic regression. Finaly age, Knowledge, waiting time, Marital structure, Community support, Residence, Education was considered for multivariable regression. The variables that show p value <0.05 were declared as statistically significant variables in multivariable logistic regression analysis.

The degree of association between independent and dependent variables was assessed using crude OR and adjusted OR (AOR) with 95% CI. The Hosmer Lemeshow test was used to check the model fitness, the Hosmer–Lemeshow goodness-of-fit test indicated adequate model fit (p = 0.80). Multicollinearity was checked (variance inflation factor < 10), indicating that there was no multicollinearity among the variables in this study.

Antenatal care: ANC is defined as the care provided by skilled health care professionals to pregnant women and adolescent girls to ensure the best health conditions for both mother and baby during pregnancy. ¹⁵

Early ANC booking was defined as those pregnant women initiation ANC before or at 12 weeks of pregnancy, whereas, late ANC initiation was defined as those pregnant women starting ANC after 12 weeks of their pregnancy. This is taken from the last menstrual period of the woman. ¹⁵

Cases: Cases were pregnant women visited the ANC unit before 12 weeks of gestational age (ie, early initiators of ANC) while, controls were pregnant women who initiate ANC after 12 weeks.

Knowledge of ANC: Was assessed using structured questions covering early booking, danger signs, pregnancy complications, timing and importance of ANC, ANC services, and gestational age. Participants scoring ≥50% of the total knowledge questions were categorized as having good knowledge, while those scoring <50% were categorized as having poor knowledge. ¹⁹ The knowledge classification threshold was adapted from previous Ethiopian maternal health studies. ¹⁹

Long Waiting Time: Waiting time was defined as the cumulative time a pregnant woman spent waiting at registration, triage, laboratory/diagnostic units, and before consultation with the healthcare provider during the ANC visit. Based on previous Ethiopian outpatient service studies, waiting time was categorized as short (<120 minutes) and long (≥120 minutes). ²⁰

Ethical clearance was obtained from Bule Hora University following the Declaration of Helsinki and was approved by Institutional Research Ethics Review Committee (BHU/IRERC) with protocol number of BHU/IRERC/030/17. A support letter was obtained from the Bule Hora Town health department to the hospital. All participants were informed about the purpose and the procedures of the study. All responses were kept confidential and anonymous. The study protocol was approved by the Bule Hora University Institutional Review Board (IRB). Prior to enrollment, informed consent was obtained from all participants: written consent was obtained from those with formal literacy, while witnessed verbal consent was obtained from participants unable to read or write.

Four hundred forty-eight pregnant women participated in the study (112 cases and 336 controls) with a 98.4% response rate. The mean age of the respondents was 27.42 years (±5.72 SD). Most participants were urban residents, 394 (87.9%). The majority were married, 441 (98.4%), with slightly fewer in controls, 108 (96.4%) vs. 333 (99.1%) in cases. Secondary education (9–12 grades) was the most common level, 123 (27.5%), and housewife was the dominant occupation, particularly among controls, 208 (61.9%) vs. 37 (33.0%) in cases. Husbands of 142 (31.7%) participants had higher education, slightly more among controls, 41 (36.6%). Less than 10 km distance to nearest health facility was reported by 321 (71.7%) participants, more in cases, 108 (96.4%) vs 213 (63.4%) in controls ( Table 1).

Most participants knew the importance of early ANC, 436 (97.3%), and its benefits, 433 (96.7%). Knowledge of Early ANC initiation was higher among 106 cases (94.6%), than controls, 154 (45.8%). Very small number of controls, 67 (19.9%), knew the correct time to start ANC (<12 weeks) or first trimester, compared to 71 (63.4%) cases. Regarding the benefits of early ANC initiation, monitoring maternal and fetal health was reported by 106 (94.6%) cases, and by 154(45.5%) controls. Regarding the benefits of early ultrasound imaging before 24 weeks was reported by 111 (99.1%), while 233 (73.3%) significantly reported high by cases. Regarding the recommended number of ANC visits, most participants believed that less than four visits were sufficient, reported by 82 (73.2%) cases and 194 (57.7%) controls. In contrast, awareness of the currently recommended 8 or more ANC visits was very low, reported by only 2 (1.8%) cases and 3 (0.9%) controls. Knowledge of pregnancy danger signs was 189 (42.2%), with vaginal bleeding being the most recognized sign, 125 (66.1%). The main sources of ANC information were community health workers, 170 (37.9%), and health workers, 116 (25.9%) ( Table 2).

Knowledge of ANC was assessed using 12 structured questions covering timing of ANC initiation, recommended number of visits, ANC services, and pregnancy danger signs. Each correct response was given a score of 1 and incorrect responses 0. The total score was converted into a percentage, and respondents scoring ≥50% were classified as having good knowledge, while those scoring <50% were classified as having poor knowledge.

As shown in the figure below, overall knowledge of ANC was higher among early ANC utilizers (cases) with 92 (82.1%) demonstrating good knowledge, compared to 164 (48.8%) of late utilizers (controls). Overall, 256 (57.1%) of participants had good knowledge, while 192 (42.9%) had poor knowledge of ANC ( Figure 1).

Out of total participants, 42 (37.5%) mothers with cases and 97 (28.9%) mothers with controls were primigravida and 309 (69.0%) multigravidas. One third of case 31(27.87) and 104 (31.0%), respectively were given birth. In their last pregnancy, only 49 (17.2%) initiated ANC in the first trimester, while 142 (49.8%) started in the second trimester, and 29 (10.2%) in the third. Place of previous delivery showed that 12 (10.0%) of the mother with cases and 57 (17.0%) of mothers with controls had a history of birth at home for the preceding birth, whereas 245 (54.7%) gave birth at a health institution and 9 (2.0%) on the way. A total of 70 (15.6%) had a history of abortion and 33 (7.4%) experienced stillbirths. Family planning use before pregnancy was reported by 250(55.8%) and ANC attendance in a previous pregnancy by cases and 201 (63.8%) controls. Pregnancy was planned by 331 (73.9%).

Among the participants, 309 (69.0%) reported proximity to health facility, with a higher proportion among cases 101 (90.2%) compared to controls 208 (61.9%). Waiting times (>120 minutes) were experienced predominantly among controls 169 (50.3%) versus only 20 (17.9%) cases. Most participants rated provider attitudes as good or very good: 421 (94.0%). In terms of socio-cultural and support factors, husband accompaniment to ANC was reported by 221 (52.1%), and 321 (71.7%) were exposed to maternal health messages through media. Perceived community support for ANC was reported by 106 (94.6%) cases and 253 (75.3%) controls, and 397 (88.6%) believed ANC attendance was a community norm. Decision-making regarding ANC was self- or jointly made by 403 (90.0%) participants.

Perceived community support for ANC was measured using a Likert-scale questionnaire, and total scores were summed. Participants were classified as having “good” or “poor” support based on whether their score was above or below the mean.

In the bivariate analysis, variables associated with early ANC and (p < 0.25) were considered for the multivariable logistic regression. After adjusting for potential confounders, the multivariate logistic regression analysis showed that maternal age, knowledge of ANC, waiting time before consultation, marital structure, and community support were significantly associated with Early ANC initiation. Our study showed that women aged 15–24 years were nearly three times more likely to initiate ANC early compared with women aged 35–44 years (AOR = 2.79; 95% CI: 1.12–6.96). The odds of early initiation of ANC visit among pregnant women with good knowledge of ANC (AOR = 4.19; 95% CI: 2.24–7.82) were four times more likely to initiate early ANC (<12 weeks) compared with women who had poor knowledge. Women who waited less than 120 minutes before provider consultation were more likely to initiate ANC early compared with women who experienced waiting times of 120 minutes or more (AOR = 4.02; 95% CI: 2.23–7.24). Women in monogamous marriages had nearly 4 times higher odds of initiating ANC early compared with women in polygamous marriages (AOR = 3.74; 95% CI: 1.53–9.16). Similarly, women who reported good community support for ANC were more likely to initiate ANC early compared with women who reported poor community support (AOR = 3.63; 95% CI: 1.44–9.17) ( Table 3).

A major strength of this study was its comprehensive assessment of factors associated with early antenatal care (ANC) initiation, incorporating socio-demographic, obstetric, and healthcare-related variables to provide a broader understanding of Early ANC initiation. However, institutional-based case-control study has several limitations. Firstly, although case-control studies are useful for identifying associations, temporal relationships between some exposures and early ANC initiation cannot be definitively established. Secondly, since it was conducted in a referral hospital, the findings may not be generalizable to all women in the community. In addition, recall and social desirability biases may affect the accuracy of self-reported data, particularly regarding missed doses and socioeconomic status while incomplete ANC registration records affected the accuracy of gestational age data.