Background
An understanding of the individual determinants (patient-related factors) of antenatal care (ANC) utilisation may assist the pursuit of adequate levels of care recommended for every pregnancy. ANC is important because it enables early and continuing risk assessment, health promotion and medical and psychosocial follow-up [
1]. Despite its value, some women do not make proper use of ANC [
2].
According to Andersen and Newman ‘s health behavioural model, individual determinants of health care utilisation can be divided into predisposing, enabling [
3] and need components [
4]. With respect to ANC, predisposing determinants refer to individual characteristics which exist prior to the pregnancy and affect the propensity to use care. Previous studies have concluded that low maternal age [
4‐
7], being single [
7], low educational level [
6‐
9], lack of a paid job [
9], foreign ethnic background [
6,
9] or origin [
2,
5,
8], poor language proficiency [
1,
7], (little) support from a social network [
1] and lack of knowledge of the health care system [
1] are associated with inadequate ANC utilisation. Enabling determinants refer to conditions which make ANC available to pregnant women. Absence of health insurance [
6,
7], planned pattern of ANC [
6], hospital type at booking [
6], personalized communication and knowledge of cultural practices of the care provider [
1] have been found to be associated with inadequate ANC. The pregnancy-need component of the determinants include pregnancy related elements explaining the degree of care needed/used. Inadequate use of ANC seems to be related to high parity [
5‐
7], unplanned pregnancy [
7], no previous premature birth [
6], discontinuity of care [
8], late recognition of pregnancy [
6] and behavioral factors such as smoking during pregnancy [
6,
9].
The measurement of ANC utilisation varies across studies, therefore results must be interpreted cautiously. The initiation of care [
1,
5‐
7,
9], the number of antenatal visits [
6,
7] and several indices based on the timing of initiation of ANC, the total number of antenatal visits and the gestational age at birth [
2,
6‐
8] have been used previously to define ANC use. Since there is no consensus about the number of antenatal visits [
10], it is preferable to take into account elements of the content and timing of care during the pregnancy. One recent study measured ANC more comprehensively using the content and timing of care during pregnancy (CTP) tool [
8].
Previously defined determinants of ANC use should be interpreted in relation to the context of these studies. In addition to individual determinants, health care utilisation depends on resources (e.g. number of care providers available) and the organisation of the national health care system, such as the nature of referrals between health care providers [
3]. Feijen-de Jong et al. identified the need for comparative research in several countries with varying antenatal health care arrangements as these might explain differences in the effects of individual determinants on ANC use [
6]. Therefore in this study, we compared ANC between two groups of ANC attendees in two different countries (Belgium and the Netherlands) with a different health care system. In the Netherlands, most women with uncomplicated pregnancies receive ANC from primary care midwives who act as gatekeepers to secondary obstetric care [
11]. They receive fixed remunerations for follow-up during the full length or part of the pregnancy [
12]. In Belgium, most women access an obstetrician directly for ANC as they do not need preauthorisation to gain access to specialist care [
5]. The majority of general practitioners, specialists and independent midwives in Belgium are paid on a fee-for-service basis [
13].
This study aims to 1) compare ANC utilisation in Belgium and the Netherlands as measured by the CTP tool and 2) to identify its predisposing, enabling and pregnancy-related determinants.
Results
Characteristics of the women
The final data set consisted of 642 women. Chi-squared tests indicated significant differences between the two subsamples for marital status, occupational status, nationality, educational level of the partner, nationality of the partner, equivalent income, health insurance or additional health insurance cover, parity, desire for pregnancy and attendance of antenatal information classes (
p < 0.05) (Table
1).
Table 1
Study sample characteristics, comparison between both groups of antenatal care attendees (n = 642)
Predisposing characteristics | | | | |
Age (years) | | | | 1.000 |
< =20 | 8 (1.2) | 4 (1.2) | 4 (1.2) | |
21-35 | 528 (82.2) | 264 (82.2) | 264 (82.2) | |
> 35 | 106 (16.5) | 53 (16.5) | 53 (16.5) | |
Marital status | | | |
.000
|
Co-habiting or married | 604 (94.1) | 291 (90.7) | 313 (97.5) | |
Single | 38 (5.9) | 30 (9.3) | 8 (2.5) | |
Educational level | | | | 1.000 |
Up to secondary | 376 (58.6) | 188 (58.6) | 188 (58.6) | |
Tertiary | 266 (41.4) | 133 (41.4) | 133 (41.4) | |
Occupational status | | | |
.000
|
Employed | 419 (65.3) | 149 (46.4) | 270 (84.1) | |
Unemployed | 223 (34.7) | 172 (53.6) | 51 (15.9) | |
Nationality | | | |
.000
|
Belgian/Dutch | 475 (74.0) | 184 (57.3) | 291 (90.7) | |
All other nationalities | 167 (26.0) | 137 (42.7) | 30 (9.3) | |
Educational level partner | | | | .001 |
No partner | 38 (5.9) | 30 (9.3) | 8 (2.5) | |
Up to secondary | 334 (52.0) | 163 (50.8) | 171 (53.3) | |
Tertiary | 270 (42.1) | 128 (39.9) | 142 (44.2) | |
Nationality of the partner | | | |
.000
|
No partner | 38 (5.9) | 30 (9.3) | 8 (2.5) | |
Belgian/Dutch | 441 (68.7) | 170 (53.0) | 271 (84.4) | |
All other nationalities | 163 (25.4) | 121 (37.7) | 42 (13.1) | |
Enabling characteristics | | | | |
Equivalent incomea
| | | |
.000
|
Low | 112 (17.4) | 92 (28.7) | 20 (6.2) | |
Moderate | 451 (70.2) | 151 (47.0) | 300 (93.5) | |
High | 79 (12.3) | 78 (24.3) | 1 (0.3) | |
Health insurance coverage | | | |
.000
|
Yes | 623 (97.0) | 302 (94.1) | 321 (100.0) | |
No | 19 (3.0) | 19 (5.9) | 0 (0.0) | |
Additional health insurance coverage | | | |
.000
|
Yes | 431 (67.1) | 151 (47.0) | 280 (87.2) | |
No | 211 (32.9) | 170 (53.0) | 41 (12.8) | |
Pregnancy-related characteristics | | | | |
Parity | | | |
.001
|
Primiparae | 284 (44.2) | 121 (37.7) | 163 (50.8) | |
Multiparae | 358 (55.8) | 200 (62.3) | 158 (49.2) | |
Wish for pregnancyb
| | | |
.002
|
Wanted pregnancy | 628 (98.0) | 308 (96.3) | 320 (99.7) | |
Unwanted pregnancy | 13 (2.0) | 12 (3.8) | 1 (0.3) | |
Planned pregnancy | | | | .239 |
Yes | 512 (79.8) | 250 (77.9) | 262 (81.6) | |
No | 130 (20.2) | 71 (22.1) | 59 (18.4) | |
COCc
| | | | .253 |
< 50 % | 463 (72.1) | 238 (74.1) | 225 (70.1) | |
> =50 % | 179 (27.9) | 83 (25.9) | 96 (29.9) | |
Attending antenatal information courses | | |
.000
|
Yes | 238 (37.1) | 71 (22.1) | 167 (52.0) | |
No | 404 (62.9) | 250 (77.9) | 154 (48.0) | |
The majority of the women in the final data set werre aged between 21 and 35 years (82.2 %), werre co-habiting or married (94.1 %), employed (65.3 %), did not have tertiary education (58.6 %). 42.1 % did not have a foreign nationality (Table
2). Of the women, 42.1 % had a partner with tertiary education and 25.5 % had a partner with a foreign nationality.
Table 2
Study sample characteristics, chi-squared test reporting significance level for association with antenatal care utilisation, ordinal regression analysis reporting adjusted OR for being assigned into a higher CTP category
Predisposing characteristics | | | | | | | |
Age (years) | | | | | | 0.32(a) | (b) |
≤20 | 8 (1.2) | 0 | 0 | 6 (75.0) | 2 (25.0) | | |
21–35 | 528 (82.2) | 41 (7.8) | 40 (7.6) | 172 (32.6) | 275 (52.1) | | |
>35 | 106 (16.5) | 8 (7.5) | 6 (5.7) | 36 (34.0) | 56 (52.8) | | |
Marital status | | | | | | 0.14(a) | (b) |
Co-habiting or married | 604 (94.1) | 44 (7.3) | 45 (7.5) | 197 (32.6) | 318 (52.6) | | |
Single | 38 (5.9) | 5 (13.2) | 1 (2.6) | 17 (44.7) | 15 (39.5) | | |
Occupational status | | | | | |
<0.001
| |
Employed | 419 (65.3) | 20 (4.8) | 26 (6.2) | 120 (28.6) | 253 (60.4) | | |
Unemployed | 223 (34.7) | 29 (13.0) | 20 (9.0) | 94 (42.2) | 80 (35.9) | |
0.49 (0.34-0.70)
|
Educational level | | | | | |
<0.001
| |
Up to secondary | 376 (58.6) | 35 (9.3) | 33 (8.8) | 139 (37.0) | 169 (44.9) | |
0.60 (0.43-0.82)
|
Tertiary | 266 (41.4) | 14 (5.3) | 13 (4.9) | 75 (28.2) | 164 (61.7) | | |
Nationality | | | | | |
0.009
| (b) |
Belgian/Dutch | 475 (74.0) | 29 (6.1) | 36 (7.6) | 149 (31.4) | 261 (54.9) | | |
All other nationalities | 167 (26.0) | 20 (12.0) | 10 (6.0) | 65 (38.9) | 72 (43.1) | | |
Educational level partner | | | | | |
<0.001
| (b) |
No partner | 38 (5.9) | 5 (13.2) | 1 (2.6) | 17 (44.7) | 15 (39.5) | | |
Up to secondary | 334 (52.0) | 33 (9.9) | 30 (9.0) | 120 (35.9) | 151 (45.2) | | |
Tertiary | 270 (42.1) | 11 (4.1) | 15 (5.6) | 77 (28.5) | 167 (61.9) | | |
Nationality of the partner | | | | | |
0.003
| (b) |
No partner | 38 (5.9) | 5 (13.2) | 1 (2.6) | 17 (44.7) | 15 (39.5) | | |
Belgian/Dutch | 441 (68.7) | 27 (6.1) | 29 (6.6) | 133 (30.2) | 252 (57.1) | | |
All other nationalities | 163 (25.4) | 17 (10.4) | 16 (9.8) | 64 (39.3) | 66 (40.5) | | |
Region | | | | | |
0.009
| |
Brussels Metropolitan | 321 (50.0) | 31 (9.7) | 26 (8.1) | 118 (36.8) | 146 (45.5) | | 0.90 (0.64-1.26) |
Urban Dutch regions | 321 (50.0) | 18 (5.6) | 20 (6.2) | 96 (29.9) | 187 (58.3) | | |
Enabling characteristics | | | | | | | |
Equivalent income | | | | | |
<0.001
| (b) |
Low | 112 (17.4) | 17 (15.2) | 9 (8.0) | 51 (45.5) | 35 (31.3) | | |
Moderate | 451 (70.2) | 29 (6.4) | 33 (7.3) | 141 (31.3) | 248 (55.0) | | |
High | 79 (12.3) | 3 (3.8) | 4 (5.1) | 22 (27.8) | 50 (63.3) | | |
Health insurance cover | | | | | |
0.008(a) | (b) |
Yes | 623 (97.0) | 46 (7.4) | 46 (7.4) | 202 (32.4) | 329 (52.8) | | |
No | 19 (3.0) | 3 (15.8) | 0 (0.0) | 12 (63.2) | 4 (21.1) | | |
Additional health insurance | | | | | |
<0.001
| (b) |
Yes | 431 (67.1) | 24 (5.6) | 29 (6.7) | 130 (30.2) | 248 (57.5) | | |
No | 211 (32.9) | 25 (11.8) | 17 (8.1) | 84 (39.8) | 85 (40.3) | | |
Pregnancy-related characteristics | | | | | | |
Parity | | | | | |
0.042
| (b) |
Primiparae | 284 (44.2) | 19 (6.7) | 16 (5.6) | 84 (29.6) | 165 (58.1) | | |
Multiparae | 358 (55.8) | 30 (8.4) | 30 (8.4) | 130 (36.3) | 168 (46.9) | | |
Wish for pregnancyd
| | | | | | 0.51(a) | (b) |
Wanted pregnancy | 628 (98.0) | 49 (7.8) | 44 (7.0) | 210 (33.4) | 325 (51.8) | | |
Unwanted pregnancy | 13 (2.0) | 0 (0.0) | 2 (15.4) | 4 (30.8) | 7 (53.8) | | |
Planned pregnancy | | | | | |
0.013
| (b) |
Yes | 512 (79.8) | 35 (6.8) | 34 (6.6) | 161 (31.4) | 282 (55.1) | | |
No | 130 (20.2) | 14 (10.8) | 12 (9.2) | 53 (40.8) | 51 (39.2) | | |
COCe
| | | | | |
0.041
| |
<50 % | 463 (72.1) | 42 (9.1) | 39 (7.8) | 158 (34.1) | 227 (49.0) | |
0.60 (0.42-0.84)
|
≥50 % | 179 (27.9) | 7 (3.9) | 10 (5.6) | 56 (31.3) | 106 (59.2) | | |
Attending antenatal information classes | | | | |
<0.001
| |
Yes | 238 (37.1) | 11 (4.6) | 7 (2.9) | 72 (30.3) | 148 (62.2) | | |
No | 404 (62.9) | 38 (9.4) | 39 (9.7) | 142 (35.1) | 185 (45.8) | |
0.67 (0.47-0.94)
|
With regard to the enabling characteristics, 70.2 % of the women had a moderate equivalent income, 97.0 % had health insurance cover and 32.9 % had no additional health insurance cover.
The pregnancy-related characteristics revealed that 55.8 % of the women were multiparae. Pregnancy was wanted for 98.0 % of the women but unplanned for 20.2 %. A lower continuity of care provider, represented by a COC index < 50 %, was observed for 72.1 % of the women, while 62.9 % did not attend antenatal information classes.
Comparison of ANC utilisation between both regions
ANC utilisation differs significantly between regions (
p = 0.009) (Tables
2 and
3). According to the classification by the CTP tool, 9.7 % of the women from the Belgian subsample had an inadequate care trajectory compared with 5.6 % in the Dutch subsample. Furthermore, only 45.5 % of the women in Belgium, compared to 58.3 % of Dutch women, were assigned to the appropriate ANC group (Table
3).
Table 3
Comparison of antenatal care utilization between regions (N = 642)
Content and Timing of Pregnancy care | | | |
Inadequate | 49 (7.6) | 31 (9.7) | 18 (5.6) |
0.009
|
Intermediate | 46 (7.2) | 26 (8.1) | 20 (6.2) | |
Sufficient | 214 (33.3) | 118 (36.8) | 96 (29.9) | |
Appropriate | 333 (51.9) | 146 (45.5) | 187 (58.3) | |
Individual determinants of ANC utilisation
The predisposing characteristics of occupational status (
p < 0.001), educational level and nationality of the women (
p < 0.001;
p = 0.009 respectively) and their partners (
p < 0.001;
p = 0.003 respectively) were found to be significantly associated with ANC utilisation (Table
2). Appropriate ANC use was higher among women with tertiary education (61.7 %), who were employed (60.4 %) and who were native (54.9 %) compared with women with secondary level education (44.9 %), who were unemployed (35.9 %) and had a foreign nationality (43.1 %) respectively.
Concerning the enabling characteristics, results showed that the higher the equivalent income, the higher the proportion of women with appropriate ANC utilisation (p < 0.001). More than half of the women with moderate (55.0 %) or high equivalent income (63.3 %) received appropriate ANC. This proportion was 31.3 % among women with low equivalent income. Women with health insurance and additional health insurance cover received appropriate content and timing of pregnancy care more often than women without this coverage (52.8 % versus 21.1 % and 57.5 % versus 40.3 % respectively) (p = 0.008 and p < 0.001 respectively).
With respect to pregnancy-related characteristics, appropriate care use was higher among primiparae (58.1 %), women with a planned pregnancy (55.1 %), women who had a COC index ≥ 50 % (59.2 %) and women who attended antenatal information classes (62.2 %) compared with multiparae (46.9 %), women with an unplanned pregnancy (39.2 %), women who had a COC index < 50 % and women who did not attend antenatal information classes (45.8 %) respectively (p < 0.05).
In the final model of the multivariate analysis, after adjustment for confounding variables (Table
2), the overall regional variable – the Belgian versus the Dutch subsamples – did not remain significantly associated with ANC use. However, four variables were significantly associated with ANC utilisation when controlling for the other variables. Women with no more than a secondary education (OR: 0.60; 95 % CI 0.43–0.82) and unemployed women (OR: 0.49; 95 % CI 0.34–0.70) had lower odds of being assigned to a higher CTP category compared with women with tertiary education and employment respectively.
In the final model no enabling characteristics remained significantly associated with the content and timing of ANC.
Women with a COC index < 50 % (OR: 0.60; 95 % CI 0.42–0.84) and women who did not attend antenatal information classes (OR: 0.67; 95 % CI 0.47–0.94) had lower odds of obtaining a higher CTP classification compared with women with a COC index ≥ 50 % and those attending antenatal information classes respectively.
Discussion
This study compares ANC utilisation as classified by the CTP tool between two groups of ANC attendees in two different countries and identified predisposing, enabling and pregnancy-related determinants based on a pooled data set. To our knowledge this is the first international comparative study that has considered these three factors related to the content and timing of ANC. Unadjusted analysis reveal that women in urban Dutch regions receive more appropriate ANC than women in the Brussels Metropolitan Region. However, multivariate analysis do not indicate that the region in itself is a determinant of ANC utilisation when controlling for all individual characteristics. This finding makes the study unique. Irrespective of the region, adequate content and timing of ANC is associated with higher educational level, employed status, higher continuity of care and attendance of antenatal information classes.
Previous studies have shown that a low educational level is associated with late initiation of ANC [
7,
9], a low number of antenatal visits [
6,
28], receiving no care at all [
6] and a lower probability of being in a higher CTP category [
8]. Lack of a paid job [
9] and type of occupation [
29] have also been related to inadequate ANC use. Choté et al. suggested that education may influence ANC use due to the level of general health knowledge and health literacy [
9]. The knowledge and skills acquired through education may create better access to information, stimulate receptiveness to health education messages and thus enable to access and communicate with health care providers [
30].
The social network, which may be less extended in unemployed women might be a mechanism explaining the association of employment with ANC use. Information and encouragement received through a social network may stimulate women to use care [
31,
32].
No enabling characteristics, such as income, was retained in our final model. The compulsory universal cover offered by health insurers, which includes basic ANC in both Belgium [
13] and the Netherlands [
33] may play a part. However, the provision of universal cover seems to be insufficient to offset disparities in ANC utilisation [
29]. The use of health care services can be measured in terms of realised access to these services [
4]. Inequitable access occurs when important structural aspects of society determine who receives appropriate ANC. However, a sole focus on measures designed to alter these aspects – such as educational level and employment status – for the sake of promoting equitable access, is hard due to their low mutability [
4]. Other measures, such as the promotion of health literacy and knowledge from an early age through the education system or the training of health professionals in communication skills to adapt to the health literacy level of the care seeker, may encourage better utilisation of care [
34].
With regard to pregnancy-related determinants, this study demonstrates that a lower continuity of ANC provider is associated with a lower CTP category. This index is calculated without differentiating between the type of primary caregiver – in Belgium most often an obstetrician and in the Netherlands a midwife. These results indicate that the continuity of care provider is important for the appropriateness of care irrespective of the type of provider. Attending antenatal classes is related to receiving more appropriate ANC, although the number and content of these classes were not considered. While non-attenders are not convinced that antenatal classes might benefit them, attenders consider them to be valuable [
35]. Similarly, non-attenders may be less convinced of the importance of and need for ANC, which may hinder appropriate ANC use. Non-attenders of antenatal education classes are found to come from more vulnerable groups, with a low level of education or being unemployed [
36]. Enhancing the awareness of the importance of appropriate follow-up and the advantages of antenatal classes may stimulate care use.
Cross-border data-sharing enabled the study of ANC utilisation in two countries. However there are some limitations to the study. The number of variables used in this study was restricted by the variables equally examined and operationalised in the original studies [
8,
14]. For example, origin or ethnicity could not be examined in this study due to different operationalization of the variables in both datasets, although previous studies have identified these variables as important determinants of ANC use [
2,
5,
6,
8,
9]. These differences in the data sets could lead to possible bias of the results. Furthermore, it would be valuable to extend the set of determinants with more elements of the health care system (eg main care provider, reimbursement system) to unravel their role in relation to antenatal care utilisation. In both studies only women that seek care were included. Therefore we are unable to draw conclusions in this specific group of women.
Conclusions
While it could be expected that the country women live in, with a specific health care system, would have an impact on the appropriateness of antenatal care use, personal characteristics seemed to have a larger impact. The results of our study demonstrate that educational level and employment status are important factors in obtaining appropriate content and timing of ANC in both regions. One way to promote appropriate ANC and influence practice would be to introduce measures encouraging women to attend antenatal classes, for example by providing classes free of charge to socially vulnerable women. The organisation of public education about the (importance of) antenatal care is another recommendation for practice. Furthermore, it is important to systematically create maternal health care models in which the continuity of care provider is ensured. All are modifiable factors that will contribute to more appropriate care use and can be considered by perinatal health care practitioners.
This is the first study measuring received content and timing of care in pregnancy (CTP) across countries. Despite the value of this study, more cross-border studies are required including other/more countries with varying health care systems. A pan-European approach would be appropriate in order to perform collaborative research aiming at increasing the uptake of antenatal care. Further other individual determinants, such as origin, social network and health beliefs with regard to pregnancy and care could be examined. These future studies should also use a larger sample including women residing in both urban and non-urban regions. To achieve this, systematic and routine data collection that provides information on elements of the CTP tool and the individual characteristics of pregnant women will be required.
Acknowledgements
We want to thank all women that agreed to participate in the initial studies in order to make this comparison possible. Furthermore we want to thank the Vrije Universiteit Brussel and the VU University Medical Center, Amsterdam to make this international comparison possible.