Determinants of delayed antenatal visit attendance in rural Burkina Faso: a cross-sectional study

21

The World Health Organization recommends a minimum of eight antenatal care (ANC) 22 contacts, with the first visit occurring before the 12th week of gestation, as a strategy to enhance the 23 preparedness of women for institutional delivery and improve perinatal outcomes. The present study 24 aims to assess the prevalence of delayed ANC attendance among pregnant women in rural Burkina 25 Faso and identify associated risk factors. 26

This is a secondary analysis of clinical data collected from a randomised-controlled trial 27 (clinicaltrials.gov ref: NCT03199547); conducted between 2018 and 2021 in rural Burkina Faso. We 28 estimated gestational age (GA) at the first ANC visit based on recall information on the last menstrual 29 period provided by study participants or, when such information was unavailable, symphysis-fundal 30 height measurements taken by ANC nurses. We used descriptive methods followed by unadjusted and 31 adjusted logistic regression, informed by an original conceptual framework, to determine the 32 prevalence and risk factors associated with delayed first ANC visit, defined as occurring after the 12th 33 week of gestation. A significance threshold was set at 0.05. 34

Out of the 5250 women enrolled in the study, 2480 (47.2%) had data available from their first 35 ANC visit, and 90.6% ( 2248/2480) of th ose women had gestational age estimates. Most women 36 (n=2037/2248, 9 0.6%) attended their first ANC after the 1 4th week of gestation. The main factors 37 associated with this delay were multiparity ≥ 4 pregnancies (OR=2.26, 95%CI [1.48 – 3.4], p < 0.001) 38 and first ANC visit attended during the dry season (OR=1.79, 95%CI [1.34 – 2.39], p < 0.001). 39

Our study highlights that most pregnant women in rural Burkina Faso attended their first 40 ANC visit later than the WHO recommended timeline , increasing their risk of poor delivery outcome . 41 Although we identified some factors that increased this risk of late ANC attendance, awareness raising 42 interventions are required for the whole population as starting late seems to be the norm. 43 . CC-BY 4.0 International licenseIt is made available under a perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint The copyright holder for thisthis version posted October 24, 2025. ; https://doi.org/10.1101/2025.10.22.25338535doi: medRxiv preprint NOTE: This preprint reports new research that has not been certified by peer review and should not be used to guide clinical practice. PregnAnZI-2_ANC_preprint 2

44 The antenatal and peripartum periods carry significant risks of morbidity and mortality for both 45 mothers and their newborns. In 2020, the global maternal mortality rate (MMR) was 216 deaths 46 per 100,000 live births [1]. In sub-Saharan Africa (SSA), the MMR was substantially higher, at 47 536 maternal deaths per 100,000 live births [2]. Furthermore, the neonatal mortality rate (NMR) 48 in SSA was estimated at 27 deaths per 1,000 live births in 2021 and this accounted for 45% of 49 global neonatal deaths [3]. 50 Professional antenatal care (ANC) is crucial for improving pregnancy outcomes. The World 51 Health Organization (WHO) recommends a minimum of eight ANC contacts, with the first visit 52 occurring before 12 weeks of gestation or during the first trimester of pregnancy [4]. Early 53 attendance and regular ANC follow -ups play a crucial role in preventing and managing 54 maternal health problems and preparing women for institutional delivery [5–7]. Indeed, primary 55 risk factors for perinatal mortality can be identified during pregnancy, and the utilisation of at 56 least one early ANC visit by a skilled provider reduces the neonatal mortality rate by 39% in 57 SSA [7, 8]. A systematic review of 74 studies published between 2008 and 2018 across 23 58 countries in SSA, including West African nations (Burkina Faso, Nigeria, Benin, Ghana, and 59 Niger), revealed that more than two-thirds of pregnant women attended ANC late, and up to 1 60 in 14 had their first visit during the third trimester of pregnancy [9]. 61 Late ANC attendance is associated with perinatal complications and adverse birth outcomes 62 such as increased risk of stillbirth, low birth weight, neonatal mortality and preterm delivery [5, 63 10–12]. These serious outcomes are important public health challenges in Burkina Faso, with 64 31 preterm births per 1000 live births, and 22 stillbirths for 1000 total births in 2020 [13]. In 65 Burkina Faso, data from the Ministry of Health showed that the proportion of pregnant women 66 who attended ANC late varied across geographic areas, ranging from 52% to 74% in 2018 [14] 67 and approximately 60% in 2020 [13]. Despite the implementation of free health care policy in 68 Burkina Faso in 2016, which aimed to improve access to health care services coverage and 69 timing of ANC visits [15], there is no significant improvement of the situation [13]. To reduce 70 the perinatal mortality rate even with the free policy, it is necessary to assess the determinants 71 of late ANC attendance in order to be able to implement strategies to improve earlier ANC 72 attendance in line with the WHO recommendations that the first prenatal consultation should 73 be carried out during the first trimester (before 12 weeks of gestation) of pregnancy in an effort 74 to reduce perinatal mortality . Hence, it is important to determine the prevalence of delayed 75 . CC-BY 4.0 International licenseIt is made available under a perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint The copyright holder for thisthis version posted October 24, 2025. ; https://doi.org/10.1101/2025.10.22.25338535doi: medRxiv preprint PregnAnZI-2_ANC_preprint 3 ANC attendance and assess its main determinants in rural Burkina Faso, in the context of a free 76 healthcare policy for pregnant women and lactating mothers. 77

78 Study design 79 This is a secondary data analysis of a descriptive cohort from the PregnAnZI -2 project 80 (clinicaltrials.gov ref: NCT03199547); the latter is a phase III, double-blind, placebo-controlled 81 randomised clinical trial conducted in The Gambia and Burkina Faso from 2017 to 2021. The 82 trial procedures were reported elsewhere [16]. Briefly, women were approached for consent 83 during ANC visits but were enrolled into the trial only during labour when oral intrapartum 84 azithromycin or placebo was administered, hence the allocation did not influence the attendance 85 or timing of ANC and both trial arms were included in this study. However, we have included 86 here only the Burkina Faso cohort as it is more representative of a rural West African population 87 compared to peri-urban Gambian participants. 88 Study site 89 Burkina Faso is a West African country with approximately 20 million inhabitants, where the 90 main livelihood is agriculture. There are two seasons: a dry season (November to May) and a 91 rainy season (June to October). Women of childbearing age represent 24.1% of the population, 92 with a fertility rate in 2020 of 5 children per woman [17, 18]. At the national level, in 2020 93 nearly three-quarters of pregnant women attended one ANC visit , and among them, one-quarter 94 attended during the first trimester of pregnancy [13]. Regarding adverse perinatal outcomes in 95 Burkina Faso, the stillbirth rate was 22 per 1000 live births, NMR of 64 per 1000 live births, 96 and MMR of 147 per 100,000 live births in 2020 [13]. Since 2016, all services offered at health 97 facilities are free of charge for pregnant women, lactating mothers, and children under 5 years. 98 The present study was conducted in the Nanoro Health District catchment area located at 99 approximately 85 km from Ouagadougou, the capital of Burkina Faso. This is a rural area with 100 ANC services provided by midwives at government primary health centres. 101 Participants 102 This study included pregnant women aged 16 and above recruited in the PregnANZI-2 trial and 103 who delivered at the study health facility with ANC services available. Women were excluded 104 if: they had a known acute or chronic condition (e.g. HIV infection, diabetes mellitus); had a 105 planned caesarean section or anticipated referral to a tertiary referral facility; known severe 106 . CC-BY 4.0 International licenseIt is made available under a perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint The copyright holder for thisthis version posted October 24, 2025. ; https://doi.org/10.1101/2025.10.22.25338535doi: medRxiv preprint PregnAnZI-2_ANC_preprint 4 foetal congenital malformation; macrolide allergy or use of medication known to increase the 107 QT interval during two weeks preceding delivery [16]. 108 Study procedures 109 Written informed consent for participation in the PregnAnZI -2 trial was taken during ANC 110 visits and verbally confirmed during labour. Socio-demographic, epidemiological, and clinical 111 information were collected by trained research nurses during labour and the immediate post -112 partum period (typically 6 -24h after delivery) using the handheld Maternal and Child Health 113 book and maternal direct questioning, as per PregnAnZI-2 trial procedures [16]. 114 Outcome measures and variables of interest 115 The outcome of interest is a binary variable; delayed ANC attendance (yes or no) defined as 116 first visit after 1 4th week of gestation. This variable was generated using the gestational age 117 (GA) recorded during the first ANC visit to classify weeks of pregnancy. We estimated GA at 118 the first ANC visit based on recall information of the last menstrual period (LMP) provided by 119 study women or when such information was not recalled by the women, we used symphysis -120 fundal height measurements assessed by the government facility nurses. The correlation 121 between fundal height and the GA in weeks was previously reported to have a prediction error 122 ranging between 13.9 and 14.9 days [19]. Gold standard dating methods using first trimester 123 ultrasound scanning were not available either as part of routine antenatal care or in association 124 with PregnAnZI-2 trial. 125 To explore factors associated with delayed ANC attendance, we developed an original 126 conceptual framework based on the literature review and plausible associations (Figure 1). 127 Although existing frameworks exist for measuring ANC quality of care [20] and for ANC in 128 well-resourced settings [21], we did not identify any conceptual frameworks relevant to our 129 population of West African women living in a rural low-resource setting. Hence, we developed 130 an original one. The conceptual framework guided our choice of variables for the univariate, 131 unadjusted analyses and informed the development of adjusted analy tical models . The 132 categories for these variables were selected based on a preliminary analysis of the data to 133 consider small cell sizes, and similar outcomes in adjacent ordered or related categories. 134 135 . CC-BY 4.0 International licenseIt is made available under a perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint The copyright holder for thisthis version posted October 24, 2025. ; https://doi.org/10.1101/2025.10.22.25338535doi: medRxiv preprint PregnAnZI-2_ANC_preprint 5 Data management and statistical analyses 136 All data were electronically captured using the REDCap software, version 8.9.2, with internal 137 validation and consistency checks. Descriptive statistics (percent age, mean , and standard 138 deviations) were performed to determine the prevalence of all independent variables. Logistic 139 regression was used to assess the association between each covariate and to generate crude odds 140 ratio and 95% confidence intervals (CI) . Variables with p -value < 0.20 from the unadjusted 141 analyses that were part of the conceptual framework were included in the adjusted models. A 142 p-value < 0.05 was considered as statistically significant. All analyses were conducted using 143 Stata 15. 144

145 Characteristics of the study population 146 Overall, 42.8% (2248/5250) of enrolled women had available information regarding the timing 147 of their first ANC visit and were included in the analysis. Out of them, 1% recalled their LMP 148 and in the remaining 99% fundal heigh t measurements were taken by a study nurse. 149 Demographic and epidemiological characteristics of women included, and those who were not 150 included in the study were not statistically significant ly different, except that the included 151 women were slightly younger (Supplementary Table 1). For included women, mean age was 152 26.5 years, and more than half were multiparous . Around 58% of pregnant women attended 153 their first ANC during the dry season (November-May), 9% and 3% had history of miscarriage 154 and stillbirth respectively (Table 1). Overall, 2037/2248 (90.6%) of women attended their first 155 ANC visit after 1 4 weeks of gestation . The majority of these pregnant women (75.7%, 156 1701/2248) had their first ANC visit during the second trimester, and 14.9% (336/2248) during 157 the third trimester of their pregnancy (Figure 2). 158 Determinants of delayed ANC attendance 159 Parity and season of the first ANC visit were associated with delayed ANC attendance 160 according to the adjusted regression model. Multiparous women pregnant for fourth or more 161 time had an increased odds of starting ANC visits at > 14 weeks of gestation compared to 162 women in their second pregnancy (OR=2. 26 95%CI [1.48 – 3.35], p < 0.001). Women who 163 attended their first ANC visit during the dry season were more likely to attend after 14 weeks 164 of gestation compared to those attending for the first time during the rainy season (OR=1.7 9, 165 95%CI [1.3 4 – 2.39], p < 0.001) (Table 1). Even though , we found statistically significant 166 differences for the prevalence of delayed ANC attendance between parity groups, less than 15% 167 . CC-BY 4.0 International licenseIt is made available under a perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint The copyright holder for thisthis version posted October 24, 2025. ; https://doi.org/10.1101/2025.10.22.25338535doi: medRxiv preprint PregnAnZI-2_ANC_preprint 6 of women within each parity group had their first ANC visit on time; and the same applies to 168 both seasons. 169

170 This large cohort study highlights that most pregnant women in rural Burkina Faso do not attend 171 their first ANC visit before the 14th week of gestation, even after the implementation of a free 172 national healthcare policy. This delay occurs across all age groups, ethnic groups, parities and 173 seasons. 174 In our study, at least 9 0% of relatively healthy pregnant women living in rural Burkina Faso 175 attended their ANC visits after the 1 4th week of gestation, later than the WHO recommended 176 timeline. This figure exceeds the delayed ANC attendance previously reported from Burkina 177 Faso at national level (62.2% to 64.3%) [22]. Similar rates of late ANC attendance were 178 reported at another health district level (62.9%) located in the Cent re-North of Burkina Faso 179 [23]. Prevalence of delayed ANC attendance also varies between countries across West Africa, 180 ranging from 85.3% in Nigeria [24] to 57.7% in Ghana [25]. Elsewhere in Africa, delayed 181 ANC attendance rates are also a concern as reported in rural Ethiopia, Eastern Africa (60.5%) 182 [23, 26, 27], albeit not as high as observed in our study. 183 Reasons for the high prevalence of delayed ANC attendance in our rural West African study 184 population are likely multifactorial. Potential factors include local beliefs that revealing 185 pregnancy too early would result in miscarriage, financial constraints for transport to health 186 facilities, and limitation in the availability and quality of the ANC services. These issues are 187 not unique to Burkina Faso as they have been observed also in South Africa, Tanzania and other 188 African countries [28–30], as depicted in the framework. This framework could facilitate the 189 identification of factors for delay at first ANC visit in similar contexts for other studies. The 190 widespread delay in ANC attendance poses significant risks as it exposes pregnant women to 191 adverse pregnancy outcomes [31] by missing early preventive care and the timely identification 192 of maternal, foetal and obstetric complications. To address this, emphasising the importance of 193 early ANC visits both at health facility level and in the communities through primary health 194 workers could enhance awareness and promote earlier engagement with ANC services , as 195 demonstrated in Ethiopia [32] and Uganda [33], where pregnant women with knowledge of 196 appropriate ANC timings were more likely to initiate care timely. In Burkina Faso, despite free 197 ANC services, economic barriers such as the cost of specialised tests, transportation and the 198 indirect economic impact of time spent away from formal or domestic work still pose a 199 significant challenge. Innovative evidence -based solutions are needed to overcome these 200 . CC-BY 4.0 International licenseIt is made available under a perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint The copyright holder for thisthis version posted October 24, 2025. ; https://doi.org/10.1101/2025.10.22.25338535doi: medRxiv preprint PregnAnZI-2_ANC_preprint 7 barriers and ensure more women can access timely to ANC. Also, introducing a more flexible 201 ANC schedule system could further improve the early attendance. 202 Although most women were late to their first ANC visit, we identified two important 203 determinants: multiparity and season of first ANC visit. Consistent with previous studies from 204 Ethiopia [32] and Nigeria [34], multiparous women in our study were more than twice as likely 205 to have delayed ANC attendance compared to women during their second pregnancy. A 206 possible explanation, supported by a qualitative study in Cameroon, is that women with 207 extensive previous experience of pregnancy gain confidence in their own ability to manage their 208 pregnancies and perceive early healthcare as less critical [35]. An alternative explanation could 209 be that attitudes were shaped by the women’s previous negative experiences in the health 210 facility, such as long waiting times and poor quality of care, as highlighted in a study from 211 Papua New Guinea [36]. 212 We observed an association between season of first ANC visit and timely attendance, with 213 increased risk of delay during the dry season. This finding may be linked to domestic work 214 demands for pregnant women, as reported in an Ethiopian study where the time constraints 215 linked to household activities represented more than 24% of reasons for delayed ANC 216 attendance [27]. During the dry season in rural Burkina Faso there are various cultural activities 217 which coincide with the harvest campaign, including mask dances and funeral ceremonies [37–218 39], hence women are busy with domestic duties and this may impact on the attendance of their 219 first ANC visit. Even though we identified two determinants of delayed first ANC visit, most 220 women were at risk of being late and therefore any intervention to improve early ANC 221 attendance should ideally target all of them. 222 This study has some limitations that need to be considered when interpreting the results. Firstly, 223 we used data from a selected population participating in a randomised controlled trial, which 224 could have introduced potential bias by excluding women with serious health problems or those 225 not willing to participate. Secondly, gestation al age was estimated using inaccurate measures 226 (last menstrual period and symphysis -fundal height), limiting the accuracy of the estimations 227 and potentially misclassifying the primary outcome. However, we mitigated this risk by using 228 an established method with precision of 2 weeks to minimise the misclassification errors. A 229 third limitation is the selection bias towards pregnant women with available data to estimate 230 the gestational age. However, comparison of baseline characteristics for included and excluded 231 pregnant women did not identify any major differences in their characteristics , hence bias is 232 assumed to be minimal . Lastly, as this analysis used an existing dataset, we were unable to 233 . CC-BY 4.0 International licenseIt is made available under a perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint The copyright holder for thisthis version posted October 24, 2025. ; https://doi.org/10.1101/2025.10.22.25338535doi: medRxiv preprint PregnAnZI-2_ANC_preprint 8 assess all potential variables associated with delayed ANC attendance . Hence, there is a need 234 for further qualitative and quantitative research to understand determinants of timely attendance 235 of ANC for rural West African populations. 236

237 Our study identified a very high proportion of pregnant women in rural Burkina Faso attending 238 their ANC visits later than the WHO recommended timeframe with multiparity and dry season 239 the main determinants of delayed ANC initiation . This delay poses significant public health 240 concerns by limiting the opportunities for maternal and antepartum preventive services. Merely 241 providing free access to ANC or implementing national policies has proven to be insufficient 242 in ensuring women starting their ANC on time. There is a pressing need for further research to 243 understand the context-specific reasons behind these delays to optimise maternal and perinatal 244 health outcomes effectively. 245 DECLARATIONS 246 Ethics approval and consent to participate 247 The trial was approved by the ‘’Comité d’Ethique pour la Recherche en Santé (CERS)’’ and the Ministry 248 of Health of Burkina Faso, The Gambia Government/MRCG (Medical Research Council Unit The 249 Gambia) Joint Ethics Committee and the LSHTM Ethics Committee. All women provided written 250 informed consent during antenatal care visits and were free to withdraw at any time. 251 Consent to publish 252 Not applicable. 253 Availability of data and materials 254 Data may be obtained from a third party and are not publicly available. Qualified researchers may 255 request access with the Gambia Government/MRC Joint Ethics Committee. The review process and 256 release of data will be facilitated by MRC Unit The Gambia (http://www.mrc.gm/) through the Head of 257 Governance at MRCG. Access will not be unduly restricted. 258 Competing interests 259 The authors declare that they have no competing interests. 260 Funding 261 The PregnAnZI-2 trial was funded by a grant from the UKRI under the Joint Global Health Trial Scheme 262 (JGHT) (ref: MC_EX_MR/P006949/1) and the Gates Foundation (Ref: OPP1196513). The funders and 263 study sponsor (MRCG) had no role in the study design, collection, analysis, or interpretation of data, 264 writing of the article nor the decision to submit for publication. 265 . CC-BY 4.0 International licenseIt is made available under a perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint The copyright holder for thisthis version posted October 24, 2025. ; https://doi.org/10.1101/2025.10.22.25338535doi: medRxiv preprint PregnAnZI-2_ANC_preprint 9 Authors’ contributions 266 J.D.B. conceptualised this study with input from U .dA., H .B., H .T., A .R. Data collection was 267 coordinated by J .D.B., and PregnAnZI-2 field teams in Burkina Faso. J .D.B. performed the analysis 268 with full access to the data and input from T .R. J.D.B. drafted the initial manuscript with input from 269 H.B., A.R., H.T., T.R., A.M.S., G.J.W.N., E.Y.S, D.L. All authors contributed to the final version. H.T. 270 and A.R. gave oversight to the work as guarantors and accepted full responsibility for the finished work 271 and controlled the decision to publish. 272

273 The authors wish to thank Medical Research Council Unit, The Gambia at LSHTM for allowing 274 access to the data, the research staff, and all the participants of Nanoro health district for accepting to 275 take part in the study.276 . CC-BY 4.0 International licenseIt is made available under a perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint The copyright holder for thisthis version posted October 24, 2025. ; https://doi.org/10.1101/2025.10.22.25338535doi: medRxiv preprint PregnAnZI-2_ANC_preprint 10

277 1. Alkema L, Chou D, Hogan D, Zhang S, Moller A-B, Gemmill A, et al. Global, regional, 278 and national levels and trends in maternal mortality between 1990 and 2015, with scenario-279 based projections to 2030: a systematic analysis by the UN Maternal Mortality Estimation 280 Inter-Agency Group. Lancet Lond Engl. 2016;387:462–74. https://doi.org/10.1016/S0140-281 6736(15)00838-7. 282 2. Maternal mortality rates and statistics. UNICEF DATA. 2021. 283 https://data.unicef.org/topic/maternal-health/maternal-mortality/. Accessed 10 Mar 2023. 284 3. UNICEF. Levels and trends in child mortality: United Nations Inter-Agency Group for 285 Child Mortality Estimation (UN IGME), Report 2023. UNICEF DATA. 2023. 286 https://data.unicef.org/resources/levels-and-trends-in-child-mortality/. Accessed 9 Mar 2024. 287 4. OMS. Recommandations de l’OMS concernant les soins prénatals pour que la grossesse 288 soit une expérience positive. 2016. 289 5. Millogo T, Ouédraogo GH, Baguiya A, Meda IB, Kouanda S, Sondo B. Factors associated 290 with fresh stillbirths: A hospital-based, matched, case-control study in Burkina Faso. Int J 291 Gynaecol Obstet Off Organ Int Fed Gynaecol Obstet. 2016;135 Suppl 1:S98–102. 292 https://doi.org/10.1016/j.ijgo.2016.08.012. 293 6. Bountogo M, Sié A, Zakané A, Compaoré G, Ouédraogo T, Lebas E, et al. Antenatal care 294 attendance and risk of low birthweight in Burkina Faso: a cross-sectional study. BMC 295 Pregnancy Childbirth. 2021;21:825. https://doi.org/10.1186/s12884-021-04310-6. 296 7. Tekelab T, Chojenta C, Smith R, Loxton D. The impact of antenatal care on neonatal 297 mortality in sub-Saharan Africa: A systematic review and meta-analysis. PloS One. 298 2019;14:e0222566. https://doi.org/10.1371/journal.pone.0222566. 299 8. Chalumeau M. Can clinical risk factors for perinatal mortality in West Africa be detected 300 during antenatal care or during labor? J Gynecol Obstet Biol Reprod (Paris). 2002;31:63–9. 301 9. Okedo-Alex IN, Akamike IC, Ezeanosike OB, Uneke CJ. Determinants of antenatal care 302 utilisation in sub-Saharan Africa: a systematic review. BMJ Open. 2019;9:e031890. 303 https://doi.org/10.1136/bmjopen-2019-031890. 304 10. Lingani M, Zango SH, Valéa I, Somé G, Sanou M, Samadoulougou SO, et al. Low birth 305 weight and its associated risk factors in a rural health district of Burkina Faso: a cross 306 sectional study. BMC Pregnancy Childbirth. 2022;22:228. https://doi.org/10.1186/s12884-307 022-04554-w. 308 11. Meda N, Traore GS, Meda HA, Curtis V, Cousens SN, Mertens TE. Perinatal mortality in 309 Burkina Faso: risk factors in an urban environment of Bobo-Dioulasso. Ann Soc Belg Med 310 Trop. 1991;71:307–16. 311 12. A O, Cm O, A O, J L. [Factors associated with preterm birth in an urban African 312 environment: A case-control study at the University Teaching Hospital of Ouagadougou and 313 Saint Camille Medical Center]. Med Sante Trop. 2015;25. 314 https://doi.org/10.1684/mst.2015.0465. 315 13. Ministère de la santé. Annuaire statistique 2020. Périodique. Burkina Faso: Direction 316 générale des études et des statistiques sectorielles; 2021. 317 14. Ministère de la santé. Annuaire statistique 2018. Direction générale des études et des 318 statistiques sectorielles. Burkina Faso; 2019. 319 . CC-BY 4.0 International licenseIt is made available under a perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint The copyright holder for thisthis version posted October 24, 2025. ; https://doi.org/10.1101/2025.10.22.25338535doi: medRxiv preprint PregnAnZI-2_ANC_preprint 11 15. De Allegri M, Ridde V, Louis VR, Sarker M, Tiendrebéogo J, Yé M, et al. Determinants 320 of utilisation of maternal care services after the reduction of user fees: a case study from rural 321 Burkina Faso. Health Policy Amst Neth. 2011;99:210–8. 322 https://doi.org/10.1016/j.healthpol.2010.10.010. 323 16. Camara B, Bognini J, Nakakana U, Some A, Jagne I, Tahita M, et al. Pre-delivery 324 administration of azithromycin to prevent neonatal sepsis and death: a phase iii double-blind 325 randomized clinical trial (PregnAnZI-2 trial). Int J Clin Trials. 2022;9:34. 326 https://doi.org/10.18203/2349-3259.ijct20220110. 327 17. Comité National du Recensement IN de la S et de la D. Cinquième Recensement Général 328 de la Population et de l’Habitation du Burkina Faso : Synthèse des résultats définitifs. 329 Ouagadougou: Institut National de la Statistique et de la Démographie; 2022. 330 18. Fertility rate, total (births per woman) - Burkina Faso | Data. 331 https://data.worldbank.org/indicator/SP.DYN.TFRT.IN?locations=BF. Accessed 7 Dec 2022. 332 19. Pugh SJ, Ortega-Villa AM, Grobman W, Newman RB, Owen J, Wing DA, et al. 333 Estimating gestational age at birth from fundal height and additional anthropometrics: a 334 prospective cohort study. BJOG Int J Obstet Gynaecol. 2018;125:1397–404. 335 https://doi.org/10.1111/1471-0528.15179. 336 20. Lattof SR, Moran AC, Kidula N, Moller A-B, Jayathilaka CA, Diaz T, et al. 337 Implementation of the new WHO antenatal care model for a positive pregnancy experience: a 338 monitoring framework. BMJ Glob Health. 2020;5:e002605. https://doi.org/10.1136/bmjgh-339 2020-002605. 340 21. Betrán AP, Bergel E, Griffin S, Melo A, Nguyen MH, Carbonell A, et al. Provision of 341 medical supply kits to improve quality of antenatal care in Mozambique: a stepped-wedge 342 cluster randomised trial. Lancet Glob Health. 2017;6:e57–65. https://doi.org/10.1016/S2214-343 109X(17)30421-7. 344 22. Ministère de la santé et de l’hygiène publique. Annuaire statistique 2021. Périodique. 345 Ouagadougou; 2022. 346 23. Somé A, Baguiya A, Coulibaly A, Bagnoa V, Kouanda S. Prevalence and Factors 347 Associated with Late First Antenatal Care Visit in Kaya Health District, Burkina Faso. Afr J 348 Reprod Health. 2020;24:19–26. https://doi.org/10.29063/ajrh2020/v24i2.2. 349 24. Olufemi Olayinka T, Sebutu Bello I, Oluwafemi Olajubu T, Oloyede Oyegbade O, 350 Omobolanle Olajubu A, Tamunotonye Ezeoma I. Factors Influencing the Booking Gestational 351 Age Among Antenatal Clinic Attendees at Primary Health Centers in South West, Nigeria: A 352 Cross-Sectional Study. SAGE Open Nurs. 2022;8:23779608221139078. 353 https://doi.org/10.1177/23779608221139078. 354 25. Kotoh AM, Boah M. “No visible signs of pregnancy, no sickness, no antenatal care”: 355 Initiation of antenatal care in a rural district in Northern Ghana. BMC Public Health. 356 2019;19:1094. https://doi.org/10.1186/s12889-019-7400-2. 357 26. Paudel YR, Jha T, Mehata S. Timing of First Antenatal Care (ANC) and Inequalities in 358 Early Initiation of ANC in Nepal. Front Public Health. 2017;5:242. 359 https://doi.org/10.3389/fpubh.2017.00242. 360 27. Weldemariam S, Damte A, Endris K, Palcon MC, Tesfay K, Berhe A, et al. Late antenatal 361 care initiation: the case of public health centers in Ethiopia. BMC Res Notes. 2018;11:562. 362 https://doi.org/10.1186/s13104-018-3653-6. 363 . CC-BY 4.0 International licenseIt is made available under a perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint The copyright holder for thisthis version posted October 24, 2025. ; https://doi.org/10.1101/2025.10.22.25338535doi: medRxiv preprint PregnAnZI-2_ANC_preprint 12 28. Zon H, Pavlova M, Groot W. Factors associated with access to healthcare in Burkina 364 Faso: evidence from a national household survey. BMC Health Serv Res. 2021;21:148. 365 https://doi.org/10.1186/s12913-021-06145-5. 366 29. Mgata S, Maluka SO. Factors for late initiation of antenatal care in Dar es Salaam, 367 Tanzania: A qualitative study. BMC Pregnancy Childbirth. 2019;19:415. 368 https://doi.org/10.1186/s12884-019-2576-0. 369 30. Jinga N, Mongwenyana C, Moolla A, Malete G, Onoya D. Reasons for late presentation 370 for antenatal care, healthcare providers’ perspective. BMC Health Serv Res. 2019;19:1016. 371 https://doi.org/10.1186/s12913-019-4855-x. 372 31. Alemayehu BA, Fenta SL, Gessesse SS, Gezahegn TW, Kassahun EA, Balcha WF. 373 Proportion and factors associated with fetal adverse birth outcome among mothers who gave 374 birth at Felege Hiwot Comprehensive Specialized Hospital, Bahir Dar city, Northwest, 375 Ethiopia 2019. Pan Afr Med J. 2022;42:76. https://doi.org/10.11604/pamj.2022.42.76.34686. 376 32. Tesfaye M, Dessie Y, Demena M, Yosef T. Late antenatal care initiation and its 377 contributors among pregnant women at selected public health institutions in Southwest 378 Ethiopia. Pan Afr Med J. 2021;39:264. https://doi.org/10.11604/pamj.2021.39.264.22909. 379 33. Kisuule I, Kaye DK, Najjuka F, Ssematimba SK, Arinda A, Nakitende G, et al. Timing 380 and reasons for coming late for the first antenatal care visit by pregnant women at Mulago 381 hospital, Kampala Uganda. BMC Pregnancy Childbirth. 2013;13:121. 382 https://doi.org/10.1186/1471-2393-13-121. 383 34. Oladokun A, Oladokun RE, Morhason-Bello I, Bello AF, Adedokun B. Proximate 384 predictors of early antenatal registration among Nigerian pregnant women. Ann Afr Med. 385 2010;9:222. https://doi.org/10.4103/1596-3519.70959. 386 35. Warri D, George A. Perceptions of pregnant women of reasons for late initiation of 387 antenatal care: a qualitative interview study. BMC Pregnancy Childbirth. 2020;20:70. 388 https://doi.org/10.1186/s12884-020-2746-0. 389 36. Andrew EVW, Pell C, Angwin A, Auwun A, Daniels J, Mueller I, et al. Factors affecting 390 attendance at and timing of formal antenatal care: results from a qualitative study in Madang, 391 Papua New Guinea. PloS One. 2014;9:e93025. https://doi.org/10.1371/journal.pone.0093025. 392 37. Leguy C. Chants funéraires des Mossi (Burkina Faso). L’Homme Rev Fr 393 D’anthropologie. 2016. 394 38. Dacher M. Le deuil du père en pays goin (Burkina Faso). Systèmes Pensée En Afr Noire. 395 1989;:75-104. https://doi.org/10.4000/span.1131. 396 39. Sissao AJ. Masques, alliances et parentés à plaisanterie au Burkina Faso : le jeu verbal et 397 non verbal. 2019. 398 399 400 401 402 403 404 . CC-BY 4.0 International licenseIt is made available under a perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint The copyright holder for thisthis version posted October 24, 2025. ; https://doi.org/10.1101/2025.10.22.25338535doi: medRxiv preprint PregnAnZI-2_ANC_preprint 13 Figures 405 406 407 Figure 1. Conceptual framework to understand determinants of pregnant women’s timely use 408 of antenatal care services 409 410 411 Figure 2: Flowchart showing the inclusion of relatively healthy pregnant women in the study.412 . CC-BY 4.0 International licenseIt is made available under a perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint The copyright holder for thisthis version posted October 24, 2025. ; https://doi.org/10.1101/2025.10.22.25338535doi: medRxiv preprint PregnAnZI-2_ANC_preprint 14 Tables 413 Table 1: Characteristics associated with delayed ANC attendance in rural Burkina Faso. 414 Total Delayed ANC visit attendance Unadjusted analysis Adjusted analysis Characteristics Total 2248 (100%) No 211 (9.4%) Yes 2037 (90.6%) OR CI 95% p-value OR CI 95% p-value Age of the pregnant woman (Years) 0.01* 0.28 Age (mean, SD) 26.51 (6.33) 24.49 (5.69) 26.72 (6.35) 20 – 35 1,680 (74.73) 159 (9.46) 1521 (90.54) 1 1 15 – 19 362 (16.11) 43 (11.88) 319 (88.12) 0.77 [0.54 -1.10] 1.33 [0.81 – 2.20] > 35 206 (9.16) 09 (4.37) 197 (95.63) 2.28 [1.15 – 4.55] 1.49 [0.73 – 3.05] Mothers’ ethnicity 0.37 Mossi 2,077 (92.39) 197 (9.48) 1,880 (90.52) 1 Gurunsi 144 (6.41) 10 (6.94) 134 (93.06) 1.40 [0.72 – 2.71] Others 27 (1.20) 4 (14.81) 23 (85.19) 0.60 [0.20 – 1.75] Husbands’ ethnicity 0.55 Mossi 2,070 (92.08) 197 (09.52) 1,897 (90.48) 1 Gurunsi 143 (6.36) 10 (06.99) 133 (93.01) 1.39 [0.72 – 2.70] Others 35 (1.56) 4 (11.43) 31 (88.57) 0.81 [0.28 – 2.33] Period of the ANC visit commencement < 0.001* < 0.001 Rainy (June-October) 947 (42.13) 116 (12.25) 831 (87.75) 1 1 . CC-BY 4.0 International licenseIt is made available under a perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint The copyright holder for thisthis version posted October 24, 2025. ; https://doi.org/10.1101/2025.10.22.25338535doi: medRxiv preprint PregnAnZI-2_ANC_preprint 15 Dry (November-May) 1,301 (57.87) 95 (7.30) 1,206 (92.70) 1.77 [1.33 – 2.35] 1.79 [1.34 – 2.39] Parity < 0.001* < 0.001 Para 2 347 (15.44) 43 (12.39) 204 (87.61) 1 1 Primiparous 438 (19.48) 59 (13.47) 379 (86.53) 0.90 [0.59 – 1.38] 0.77 [0.46 – 1.27] Para 3 335 (14.90) 44 (13.13) 291 (86.87) 0.93 [0.59 – 1.46] 0.93 [0.59 – 1.47] Para ≥ 4 1,128 (50.18) 65 (5.76) 1,063 (94.24) 2.31 [1.54 – 3.47] 2.26 [1.48 – 3.45] History of miscarriage 0.23 No 2,043 (90.88) 187 (9.15) 1,856 (90.85) 1 Yes 205 (9.12) 24 (11.71) 181 (88.)29 0.65 [0.59 – 4.58] History of stillbirth 0.33 No 2,181 (97.02) 207 (9.49) 1,974 (90.51) 1 Yes 67 (2.98) 4 (5.97) 64 (94.03) 1.96 [0.61 – 6.31] * Variables included in the adjusted model. 415 Abbreviations: ANC = Antenatal care; CI = Confidence interval; OR = Odds ratio; SD = Standard deviation 416 . CC-BY 4.0 International licenseIt is made available under a perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint The copyright holder for thisthis version posted October 24, 2025. ; https://doi.org/10.1101/2025.10.22.25338535doi: medRxiv preprint