Correlates of late initiation and underutilisation of the recommended eight or more antenatal care visits among women of reproductive age: insights from the 2019 Ghana Malaria Indicator Survey.

OBJECTIVE: This study assessed the correlates of late initiation and underutilisation of the WHO's recommended eight or more antenatal care visits among women in Ghana.
DESIGN: We analysed secondary data from 2163 women in the 2019 Ghana Malaria Indicator Survey, which collected data on malaria and antenatal care indicators among women of reproductive age across the previous 10 regions of Ghana. SETTING AND PARTICIPANTS: Women of reproductive age across the 10 regions of Ghana. MAIN OUTCOME MEASURES: Late initiation and underutilisation of the recommended eight or more antenatal care visits among women of reproductive age.
RESULTS: About half (49%) of the participants were between the ages of 25 and 34 years; mean (±SD)=30 (±7.10). The majority (57%) of the participants obtained less than eight antenatal care visits, while 32% initiated antenatal care visits after the first trimester. The significant factors associated with the late initiation of antenatal care visits were age, region and parity (p<0.05). Factors associated with underutilisation of the recommended eight or more antenatal care visits were marital status, wealth index, parity, region and place of residence (p<0.05).
CONCLUSION: A majority of the women underused antenatal care services. A significant minority of the women started antenatal care visits late. Socio-demographic factors, parity and socioeconomic factors were identified as the significant factors associated with the late initiation and underutilisation of antenatal care services. Maternal health interventions should prioritise young, multiparous and poor women. © Author(s) (or their employer(s)) 2022. Re-use permitted under CC BY-NC. No commercial re-use. See rights and permissions. Published by BMJ.

This study used nationally representative data, hence making the findings generalisable to the population of women of reproductive age in Ghana.

Standard data collection procedures were used to ensure reliability; hence the survey estimates accurately represent the household situations.

The cross-sectional nature of the study makes it impractical to draw causal relationships.

We analysed secondary data which did not include some potential predictors of antenatal care (ANC) utilisation, including distance to health facilities, community factors and health-seeking behaviours.

ANC visits and time of initiation were self-reported, hence there were possibilities of social desirability and recall biases.

Introduction

Globally, the maternal mortality rate is still unacceptably high. In 2020, the global maternal mortality ratio was 152 deaths per 100 000 live births.1 Even though most maternal deaths are preventable, a large percentage (94%) occur in low-resource settings.2 Sub-Saharan Africa (SSA) alone accounts for two-thirds of the global maternal deaths.2 West Africa has the highest number of maternal deaths, which is about 20% of all global maternal deaths.3 Despite global efforts to improve access to maternal and child healthcare services, women in low-income and middle-income countries still experience a high risk of pregnancy-related deaths.4 5 Utilisation of antenatal care (ANC) services is a major indicator for tracking progress towards achieving the Sustainable Development Goals (SDGs).6 Target 3.1 of the SDGs aims to decrease maternal deaths to less than 70 per 100 000 live births by 2030.7

Early initiation and adequate utilisation of ANC services are crucial for reducing negative pregnancy outcomes, including maternal death and stillbirth.8 9 Previously, the WHO recommended a minimum of four or more ANC visits, and the first visit should occur before the 16th week of gestation.10 However, WHO in 2016 recommended eight or more ANC visits and the first visit should occur in the first trimester of pregnancy.11 This new recommendation seeks to provide an opportunity for adequate maternal and fetal assessment which can help health professionals to detect complications early and provide timely interventions to reduce maternal and perinatal deaths.12 Inadequate ANC utilisation breaks the critical link in the continuum of care and affects the health of the mother and baby.13 Despite the ramifications of late initiation and underutilisation of ANC services, many women in SSA start ANC visits in the second or third trimester of pregnancy.12

Ghana adopted the previous WHO-focused antenatal care (FANC) model14 which recommends four ANC visits. Yet, some women in Ghana are unable to attain this recommendation due to barriers such as financial constraints and poor road infrastructure.15–17 Despite the institutionalisation of the FANC model in Ghana for over a decade, the maternal mortality ratio remains high (343 per 100 000 live births).18 This raises concerns about the implementation of the latest recommendation of eight or more ANC visits during pregnancy.

Several studies16 17 19 20 have assessed factors associated with the previous recommendation of four or more ANC visits among women in Ghana. Also, a recent study in Ghana assessed the prevalence and socioeconomic inequalities in the latest recommendation of eight or more ANC visits.21 Prior studies elsewhere have shown that several factors are associated with late initiation and underutilisation of the recommended eight or more ANC visits. These factors include marital status, maternal age, parity, maternal education, husband’s education, employment status, health insurance coverage and socioeconomic status.5 12 22 Moreover, barriers such as long distances to health facilities, lack of transportation, financial constraints, lack of knowledge and misconceptions about ANC affect ANC utilisations.5 12 22

To the best of our knowledge, no study in Ghana has assessed factors associated with late initiation and underutilisation of the recommended eight or more ANC visits, using nationally representative data. To help improve maternal and reproductive health outcomes in Ghana and contribute to meeting the SDGs, there is a need for more contextual and recent evidence to inform maternal health policies and programmes. Therefore, this study aimed to assess factors associated with late initiation and underutilisation of the recommended eight or more ANC visits, using secondary data from the 2019 Ghana Malaria Indicator Survey.

Methods

Data source and study design

Ghana is one of the countries in West Africa and shares boundaries with Togo to the east, Cote d’Ivoire to the west, Burkina Faso to the north and the Gulf of Guinea to the south. Ghana has a total population of about 30.8 million with the majority being women (50.7%).23 In Ghana, malaria is endemic and perennial. Ghana is one of the countries with the highest malaria burden. Malaria in pregnancy is a public health concern and accounts for 9% of all maternal deaths in Ghana.24 We analysed data from the 2019 Ghana Malaria Indicator Survey (GMIS). The 2019 GMIS was the second of its kind, with the first survey conducted in 2016. GMIS is a national survey designed to collect information on malaria indicators (ie, ownership and use of mosquito bed nets, intermittent preventive treatment for pregnant women and malaria knowledge) to inform strategic planning and evaluation of malaria interventions in Ghana. Moreover, GMIS provides information on malaria prevention, treatment and prevalence in the country.

The 2019 GMIS was a nationwide survey that collected data across the previous 10 regions of Ghana, including rural and urban areas. Data collection was done in two phases, including the selection of 200 enumeration areas in the first phase and interviews in the second phase. During, the first phase, the households were listed, and information was collected on Global Positioning System coordinates and names of household heads. During the second phase, questionnaires were administered to eligible women aged 15–49 years.

Data collection started on 24 September 2019 and ended on 25 November 2019, by trained fieldworkers using questionnaires programmed into a Computer-Assisted Personal Interviewing application. The 2019 GMIS was conducted by the Ghana Statistical Service and Ministry of Health/Ghana Health Service with technical support from the Inner-City Fund through the Demographic and Health Survey programme.

Population and sampling

The target population for the GMIS was women between the ages of 15 and 49 years residing in Ghana at the time of the survey. The 2010 population and housing census sampling list was used as the sampling frame for the selection of households. The former 10 administrative regions of Ghana were divided into 20 strata, including rural and urban areas. The first stage comprised the selection of 200 enumeration areas (97 in urban areas and 103 in rural areas) using the proportional to size approach with independent selection in each stratum. The stratum was divided into enumeration areas or clusters.

The second stage of the recruitment included the selection of a fixed number of 30 households from each enumeration area. Replacement of non-responding households was not allowed. In all, 6000 households were selected, out of which 5833 were occupied. Of the total occupied households, 5246 women were eligible, out of which 5181 were interviewed, equivalent to a 99% response rate. The focus of this study was on women aged 15–49 who had a live birth in the last 5 years preceding the survey. Therefore, women who did not meet this criterion were dropped from the data set before the analysis. Also, non-response and ‘do not know’ responses were dropped from the data set. A total of 2939 participants out of the 5181 were dropped from the data set, the remaining 2242 participants (unweighted). After adjusting for the sample weight, clustering and stratification using the ‘svy’ Stata command, a weighted sample size of 2163 women was included in the analysis. Details about the 2019 GMIS can be found elsewhere.25

Measurement

The outcome variables were late initiation of ANC visits and underutilisation of the recommended eight or more ANC visits. All the outcome variables were originally coded as continuous variables. We recoded ANC initiation (How many months pregnant were you when you first received antenatal care for this pregnancy?) into a categorical variable: ‘0–3 months’=0: ‘early initiation’; and ‘4–9 months’=1: ‘late initiation’. We also recoded the number of ANC visits (How many times did you receive antenatal care during this pregnancy?) into a categorical variable: ‘1–7 ANC visits’=1: ‘underutilisation’ and ‘8–16 ANC visits’=0: ‘optimum utilisation’. The independent variables included: age, educational status, religion, ethnicity, wealth index, region, type of place of residence, health insurance coverage and parity. The independent variables were selected from the literature. Also, we focused on independent variables that were available in the data set.

Statistical analysis

Data were analysed in three levels, including univariate, bivariate and multivariable. The sample weight was adjusted to offset challenges associated with oversampling, undersampling and survey weight. Descriptive statistics, including frequencies, percentages, mean and SD were computed and presented in a table. Chi-square (Χ2) analysis was computed to check for associations between the independent and dependent variables. At the multivariable level, binomial logistic regression analysis was computed to identify the predictors of the outcome variables. Two regression models were computed. In model 1, the outcome was initiation of ANC visits. We simultaneously adjusted for the independent variables, including age, educational status, religion, ethnicity, wealth index, region, type of place of residence, health insurance coverage and parity. In model 2, the outcome was ANC utilisation and we simultaneously adjusted for the independent variables (same as model 1). The Hosmer-Lemeshow test showed that both models fitted well (p>0.05). All statistical significance and ORs were reported at the 95% CI and the 0.05 significance level. All statistical analyses were done with the aid of Stata/SE, V.16 (StataCorp, College Station, Texas, USA).

Patient and public involvement

The women or the public were not involved in the designing, conducting and reporting of this study.

Results

Descriptive statistics for participant characteristics

About half (49%) of the participants were aged 25–34 years; mean (±SD)=30 (±7.10). More than half (53%) of the participants had attained secondary education, 76% of the participants were Christians and 56% resided in rural areas. A higher proportion of the participants were from the Ashanti region (17%), belonged to the Akan ethnic group (41%) and were in the poorest wealth quintile (21%). Most of the participants (59%) were covered by health insurance. Close to a quarter (24%) of the participants had given birth to one child, while 12% had given birth to six or more children. Regarding utilisation of ANC services, the majority (56%) of the women obtained less than eight ANC visits. Concerning initiation of ANC visits, almost one-third (32%) of the women started ANC visits after the first trimester of pregnancy (table 1).

Table 1

Descriptive statistics of participants’ characteristics

Response	Frequency	Percentage
Region
Western Central Greater Accra Volta Eastern Ashanti Brong-Ahafo Northern Upper East Upper West	2261593222192343751842799768	1071510111791353
Educational level completed
No education Primary Secondary Higher	4474481137131	2021536
Wealth index categories
Poorest Poorer Middle Richer Richest	446447478415377	2121221917
Religion
Christian Islam Others (ie, traditional)	163545177	76204
Type of place of residence
Urban Rural	9581205	4456
Ethnicity
Akan Ewe Mole-Dagbani Others (ie, Grusi)	892318477477	41152222
Age groups (years)
15–24 25–34 35–49	4941055615	234928
Health insurance coverage
No Yes	8121151	4159
Number of children ever born
1 2 3 4 5 6 or more	511518361303209261	232417141012
ANC utilisation
<8 ANC visits (underutilisation) ≥8 ANC visits (optimum utilisation)	1222942	5644
ANC initiation
<4 months (early initiation) ≥4 months (late initiation)	1472691	6832

ANC, antenatal care.

Association between participant characteristics and adherence to the recommended ANC visits

The results showed that underutilisation of the recommended ANC visits was significantly associated with region, educational status, wealth index, religion, ethnicity, place of residence, age and parity (p value≤0.05). For instance, 76% (highest) of women in the Upper West region and 33% (lowest) of women in the Upper East region did not attain the recommended ANC visits. About 7 in 10 women with no education (66%) and those in the poorest wealth index (67%) did not attain the recommended ANC visits. In addition, initiation of ANC visits was significantly associated with educational status, wealth index and age (p value≤0.05). For example, 34% of women with no education and 18% of women with higher education started ANC visits after the first trimester of pregnancy. Also, 36% of women in the poorer wealth quintile started ANC visits after the first trimester of pregnancy. Similarly, 43% of women aged 15–24 years started ANC visits after the first trimester (table 2).

Table 2

Factors associated with underutilisation and late initiation of antenatal care (ANC) visits

Participants’ characteristics	ANC utilisation			ANC initiation
	Optimum utilisationn (%)	Under utilisationn (%)	P value	Early initiationn (%)	Late initiationn (%)	P value
Region
Western Central Greater Accra Volta Eastern Ashanti Brong-Ahafo Northern Upper East Upper West	141 (63)51 (32)190 (59)58 (27)95 (41)175 (47)79 (43)71 (25)65 (67)16 (24)	85 (37)109 (68)132 (41)161 (73)139 (59)200 (53)105 (57)208 (75)32 (33)51 (76)	<0.01	153 (68)113 (71)214 (66)153 (70)166 (71)251 (67)113 (61)184 (66)76 (78)49 (73)	73 (32)46 (29)108 (34)66 (30)68 (29)124 (33)71 (39)95 (34)21 (22)18 (27)	0.27
Educational level completed
No education Primary Secondary Higher	150 (34)177 (39)530 (47)85 (65)	297 (66)271 (61)607 (53)48 (35)	<0.01	296 (66)306 (68)762 (67)108 (82)	151 (34)142 (32)375 (33)24 (18)	0.05
Wealth index categories
Poorest Poorer Middle Richer Richest	147 (33)149 (33)203 (42)207 (50)236 (63)	299 (67)298 (67)275 (58)208 (50)141 (37)	<0.01	302 (68)286 (64)322 (67)274 (66)287 (76)	144 (32)161 (36)156 (33)141 (34)89 (24)	0.02
Religion
Christian Islam Others (ie, traditional)	754 (46)164 (36)24 (31)	881 (54)287 (64)54 (69)	<0.01	1121 (69)299 (66)52 (67)	513 (31)153 (34)25 (33)	0.71
Type of place of residence
Urban Rural	457 (48)484 (40)	501 (52)720 (60)	0.03	653 (68)820 (68)	306 (32)385 (32)	0.98
Ethnicity
Akan Ewe Mole-Dagbani Others (ie, Grusi)	439 (49)131 (41)177 (37)195 (41)	453 (51)187 (59)300 (63)283 (59)	0.02	619 (69)226 (71)320 (67)308 (64)	272 (31)92 (29)157 (33)170 (36)	0.42
Age groups (years)
15–24 25–34 35–49	168 (34)480 (46)293 (48)	326 (66)574 (54)322 (52)	<0.01	284 (57)732 (69)457 (74)	211 (43)323 (31)158 (26)	<0.01
Health insurance coverage
No Yes	358 (44)519 (45)	454 (56)633 (55)	0.80	546 (67)799 (69)	266 (33)352 (31)	0.44
Number of children ever born
1 2 3 4 5 6 or more	226 (44)249 (48)169 (47)115 (38)85 (40)98 (38)	284 (56)269 (52)193 (53)188 (62)125 (60)162 (62)	0.05	339 (66)359 (69)254 (70)200 (66)147 (70)173 (66)	172 (34)159 (31)108 (30)104 (34)62 (30)88 (34)	0.48

Predictors of underutilisation of ANC services

Women living in the other regions had higher odds of underusing the recommended ANC visits compared with those in the Upper East region. For example, women in Greater Accra (adjusted OR (AOR)=2.61, 95% CI: 1.32 to 5.18) and the Volta region (AOR=8.58 95% CI: 4.03 to 18.24) were over two and eight times more likely to underuse ANC services respectively compared with those in the Upper East region. In addition, compared with women aged 35–49 years, those aged 15–24 (AOR=2.67, 95% CI: 1.68 to 4.23) and 25–34 years (AOR=1.48, 95% CI: 1.09 to 2.01) had higher odds of underusing ANC services. Women who were in the poorest wealth quintile (AOR=2.22, 95% CI: 1.22 to 4.02) had higher odds of underusing ANC services compared with those in the richest wealth quintile. Women living in rural areas (AOR=1.44, 95% CI: 1.01 to 2.06) had higher odds of underusing ANC services compared with those in urban areas. We also found that women who had given birth to six or more children (AOR=2.74, 95% CI: 1.61 to 4.66) had higher odds of underusing ANC services compared with those who had given birth to one child (table 2).

Predictors of late initiation of ANC visits

The results demonstrated that women in the other regions had higher odds of starting ANC visits late compared with those in the Upper East region. For example, women in Greater Accra (AOR=3.08, 95% CI: 1.54 to 6.17) were three times more likely to start ANC visits late than those in the Upper East region. Also, women in the Western region (AOR=2.27, 95% CI: 1.15 to 4.48) were two times more likely to start ANC visits late compared with those in the Upper East region. In addition, women aged 15–24 years (AOR=3.51, 95% CI: 2.12 to 5.82) and those aged 25–34 years (AOR=1.78, 95% CI: 1.25 to 2.53) had higher odds of starting ANC visits late compared with women aged 35–49 years. It was also revealed that women who had given birth to six or more children had higher odds (AOR=2.74, 95% CI: 1.61 to 4.66) of starting ANC visits late compared with women who had given birth to one child (table 3).

Table 3

Predictors of underutilisation and late initiation of antenatal care (ANC) visits

Characteristic	Underutilisation		Late initiation
	AOR (95% CI)	P value	AOR (95% CI)	P value
Region
Upper East Western Central Greater Accra Volta Eastern Ashanti Brong-Ahafo Northern Upper West	1 (ref)1.55 (0.73 to 3.27)6.03 (2.88 to 12.59)2.61 (1.32 to 5.18)8.58 (4.03 to 18.24)4.79 (2.36 to 9.71)3.35 (1.61 to 6.98)2.98 (1.33 to 6.69)6.03 (3.30 to 11.01)6.54 (3.44 to 12.45)	0.246<0.0010.006<0.001<0.0010.0010.008<0.001<0.001	1 (ref)2.27 (1.15 to 4.48)1.82 (0.90 to 3.67)3.08 (1.54 to 6.17)2.19 (1.12 to –4.28)1.94 (1.00 to 3.75)2.20 (1.09 to 4.42)2.81 (1.42 to 5.54)1.82 (0.99 to 3.32)1.34 (0.75 to –2.39)	0.0180.0940.0020.0220.0480.0260.0030.0500.307
Educational level completed
No education Primary Secondary Higher	1.50 (0.84 to 2.67)1.25 (0.70 to 2.22)1.35 (0.82 to 2.23)1 (ref)	0.1670.4330.234	1.35 (0.70 to 2.61)1.04 (0.56 to 1.95)1.50 (0.75 to 3.00)1 (ref)	0.3610.8870.249
Wealth index categories
Poorest Poorer Middle Richer Richest	2.22 (1.22 to 4.02)2.46 (1.47 to 4.11)1.88 (1.25 to 2.83)1.35 (0.86 to 2.14)1 (ref)	0.0090.0010.0030.185	1.51 (0.88 to 2.59)1.57 (1.01 to 2.43)1.13 (0.83 to 2.09)1.38 (0.95 to 2.01)1 (ref)	0.1300.0410.2240.087
Religion
Christian Islam Others (ie, traditional)	1.24 (0.84 to 1.84)1 (ref)1.34 (0.66 to 2.72)	0.2680.404	1.11 (0.50 to 2.45)1.12 (0.49 to 2.51)1 (ref)	0.7840.782
Type of place of residence
Urban Rural	1 (ref)1.44 (1.01 to 2.06)	0.044	1.12 (0.86 to 1.46)1 (ref)	0.388
Ethnicity
Akan Ewe Mole-Dagbani Others (ie, Grusi)	1.13 (0.72 to 1.77)1 (ref)1.14 (0.89 to 2.36)1.21 (0.77 to 1.91)	0.5880.1290.393	1.11 (0.69 to 1.80)1 (ref)1.44 (0.71 to 2.90)1.28 (0.62 to 2.63)	0.6450.3020.496
Age groups (years)
15–24 25–34 35–49	2.67 (1.68 to 4.23)1.48 (1.09 to 2.01)1 (ref)	<0.0010.011	3.51 (2.12 to 5.82)1.78 (1.25 to 2.53)1 (ref)	<0.0010.001
Health insurance coverage
No Yes	1 (ref)1.00 (0.75 to 1.33)	0.99	1.07 (0.80 to 1.42)1 (ref)	0.638
Parity
1 2 3 4 5 6 or more	1 (ref)1.01 (0.71 to 1.41)1.23 (0.86 to 1.77)1.96 (1.18 to 3.24)1.60 (0.92 to 2.76)1.80 (1.13 to 2.87)	0.9580.2440.0090.0890.013	1 (ref)1.09 (0.77 to 1.54)1.27 (0.78 to 2.07)2.15 (1.18 to 3.90)1.76 (0.98 to 3.16)2.74 (1.61 to 4.66)	0.6160.3180.0120.057<0.001

Discussion

This study assessed factors associated with late initiation and underutilisation of the recommended eight or more ANC visits among women in Ghana. This study found that 32% of the women started ANC visits late. This prevalence (32%) is lower than findings in the rural part of Ghana (43%)19 and other SSA countries, including Ethiopia (71.2%),12 (66.3%)26 and (52.5%);27 Zambia (86.6%);28 and Tanzania (70.4%).29 These divergent findings could be attributed to the socio-economic, cultural and timing differences among the study populations.19 30 31 Late initiation of ANC visits could deny the mother of accessing the full benefits of preventive measures and early detection of disease (HIV/AIDS and other sexually transmitted infections). Moreover, mothers may miss essential services such as folate and iron supplements, which can help prevent neural tube defects.32

This study revealed that more than half (56%) of the women underused ANC services. Our finding contradicts prior studies in Africa. For instance, a recent study in Nigeria revealed that 83% of pregnant women were unable to attain the eight or more ANC visits22 while in Benin 92% of women underused ANC services.33 Also, studies outside the African continent found a high prevalence of underutilisation. For example, a study conducted in Bangladesh revealed that 94% of pregnant women did not attain the eight or more ANC visits.34 Another study in Myanmar revealed that 82% of pregnant women underused ANC services.5 Though the rate of underutilisation in this study was lower than what prior studies found, there is still the need to intensify maternal health interventions that would help increase ANC coverage.

The study also revealed geographical variations in ANC utilisation. Our findings showed that women in the Upper East region reported the highest ANC visits. Evidence shows that the Upper East region has the highest ANC and health insurance coverage in Ghana.25 35 The differences in the findings may be attributed to the high health insurance coverage in the region, which can help remove financial barriers to accessing ANC services. We also found that the Upper East region had the highest rate of early ANC initiation compared with the other regions. This might explain the low underutilisation of ANC services in the Upper East region.

Further, women of low socioeconomic status had higher odds of underutilisation compared with women of high socioeconomic status. This finding is consistent with a recent study in Nigeria36 where women of high socioeconomic status were more likely to attain the recommended ANC visits. Another study in Myanmar revealed that pregnant women in the highest wealth quintile were three times more likely to attain the recommended eight or more ANC visits compared with women in the lowest wealth quintile.5 Our finding, however, contradicts a study in Ethiopia, where it was revealed that socioeconomic status did not influence the utilisation of ANC services.37 A plausible explanation is that women of high socioeconomic status can afford maternal health services. Despite, the Free Maternal Health Policy in Ghana, pregnant women still make unapproved payments during ANC visits.38 In addition, women of high socioeconomic status can afford transportation costs and other charges which may pose barriers to accessing ANC services.

Women in rural areas were more likely to underuse ANC services compared with those in urban areas. This finding is consistent with studies conducted in low-income and middle-income countries.22 34 Evidence shows that there is a significant difference in rural–urban utilisation of maternal healthcare services including ANC, which may be due to inequities in the distribution of healthcare resources between rural and urban areas.39–42 In addition, women in rural areas may face challenges such as long distances to health facilities and poor road infrastructure.39 In a quest to increase the utilisation of ANC services in rural areas, stakeholders, including the Ministry of Health, should focus on strengthening the Community‐based Health Planning and Services (CHPS) programme. The CHPS programme aims to increase access to healthcare services, especially in rural areas.43 44

We also found that young women (15–24 years) were more likely to underuse ANC services as well as start ANC visits late. Similarly, studies in Myanmar5 and Southwest Ethiopia12 revealed that young women especially adolescents had a higher risk of late initiation of ANC visits. The finding of this study could be explained by the fact that younger women may not recognise pregnancy signs and symptoms early. They may also try to conceal pregnancy leading to late initiation of ANC visits.5 45–47

Women who had given birth to six or more children were more likely to underuse ANC services as well as start ANC visits late. Mothers who have had multiple births might have gathered some experiences, hence they may delay starting ANC visits or underuse maternal health services.48 49 The low utilisation of ANC services among multiparous women may be due to negative experiences with previous ANC visits, taking care of children and constraints on family resources.49 50 Moreover, based on experiences from previous pregnancies, multiparous women may think that they know much about the signs of pregnancy-related complications, hence they may delay in accessing ANC services. It is, therefore, crucial to prioritise multiparous women in programmes that seek to increase ANC utilisation.49 The essence of early initiation of ANC visits should be re-echoed as well as the consequences of late initiation and underutilisation of ANC services.

Recommendations and implications for public health and policy

This study provides empirical evidence on the correlates of late initiation and underutilisation of the WHO’s recommended eight or more ANC visits among Ghanaian women. We analysed nationally representative data; hence the findings of this study can be generalised to the population as well as inform national maternal health policies and programmes. This study revealed the gaps in ANC coverage in Ghana using WHO’s recommendation as a yardstick. Going forward, it is crucial to strengthen existing maternal health interventions, including the Free Maternal Health Policy and CHPS initiative. Investing more resources in the CHPS initiative can help bridge the urban–rural inequalities in the utilisation of maternal health services. It is also recommended that interventions aimed at increasing ANC utilisation should prioritise women of low socioeconomic backgrounds.

Conclusion

This study has demonstrated that the utilisation of ANC services in Ghana is below WHO’s standard. The majority of women in Ghana attain less than eight ANC visits during pregnancy and one-third make their first ANC visit after the first trimester of pregnancy. Underutilisation of ANC services was influenced by parity, maternal age, socioeconomic status, place of residence and geographical region. Late initiation of ANC visits was influenced by parity, socioeconomic status, maternal age and geographical region. Going forward, maternal health interventions should prioritise young, multiparous and women of poor socioeconomic backgrounds to help increase ANC coverage in Ghana. Future research should focus on household and health system factors that influence the utilisation of ANC among women of reproductive age. Moreover, future studies may explore health professionals’ and mothers’ perspectives on barriers to early initiation and optimum utilisation of ANC services.