• Open Access

Pregnancy intention in an urban Australian antenatal population

Authors


Abstract

Objective: To determine the prevalence of unintended pregnancy in women presenting for antenatal care to a large metropolitan hospital in Sydney, Australia, and to investigate health behaviours and demographic factors associated with unintended pregnancy.

Methods: From October 2010 to April 2011, a self-administered questionnaire covering pregnancy intention, contraceptive use and demographic information was given to 1,554 women. A total of 1,218 women (78.4%) completed all questions in the validated pregnancy intention instrument.

Results: Two-thirds of pregnancies (67.6%) were clearly intended, 30.0% were ambivalent and more than 2% experienced an unplanned pregnancy. Those more likely to experience an unintended pregnancy were women under 25 years old (OR 1.86, 95% CI 1.10–3.14), unmarried women (OR 6.08, 95% CI 3.40–10.87) and women of Asian background (OR: 2.45, 95% CI 1.76–3.42). More than one-third of women (34.6%) did not take any health actions such as stopping smoking before pregnancy.

Conclusions: Unintended pregnancies in this population were associated with young age, being unmarried and being of Asian background. This study confirms the idea that many women do not take health actions before pregnancy.

Implications: Experts believe that an effective strategy to address unintended pregnancy is to improve access to long-acting reversible contraceptives, which do not require daily compliance.

Of all pregnancies destined to end in birth, about one-third are unintended.1,2 Unintended pregnancies are a significant public health issue because of the associated financial and social costs, including costs to the health and welfare systems. An unintended pregnancy is one that is either unwanted (not wanted at all) or mistimed (not wanted at this time) and has been used in this study to encompass both ambivalent and unplanned pregnancies according to the pregnancy intention scale developed and reported previously in the UK.1 It should be emphasised that an unintended pregnancy does not necessarily equate to an unwanted child. An intended pregnancy is both desired and occurring at the correct time.

According to a US study, taxpayers spend US$9.6–12.6 billion a year on unintended pregnancy (including abortion, unintended fetal losses and antenatal care, births and infant medical care).3 This is a conservative estimate, as it does not account for non-medical benefits received by parents, the numerous benefits received beyond infancy by children of unintended pregnancies, or the increased likelihood of poverty and welfare requirements of these children.3

Apart from societal costs, unintended pregnancy often affects the lives of individuals with implications for educational attainment and workforce participation.4 This is particularly so for teenage mothers who are less likely to graduate from high school and attend post-secondary schooling, compared to women having their first child at 30 or older.5 Unintended pregnancies are also potentially associated with maternal health behaviours that can increase risks for the developing fetus, such as poor glycaemic control in women with diabetes and excessive alcohol intake.6 While some women have a healthy lifestyle prior to conception, with a diet high in folate and minimal alcohol consumption that does not require change, for some women there is a public health benefit to folic acid supplementation that has been repeatedly demonstrated in population studies.7

There have been a number of attempts to measure pregnancy intention. A six-item pregnancy intention instrument was devised and validated in the UK (Table 1)8 and has been used to determine the prevalence of unintended pregnancy in women attending antenatal care and abortion clinics in Scotland.1,9

Table 1. Measurement of pregnancy intention.1,8,9
QuestionAnswerScore
  1. Notes:

  2. a

    Health actions include folic acid supplementation, smoking cessation or reduction, alcohol cessation or reduction and seeking medical advice

At the time of conceptionAlways use contraception
Inconsistently use contraception

Not use contraception
0
1
2
In terms of becoming a motherWrong time
An OK time, but not quite right

Right time
0
1
2
Just before falling pregnantNot intend to become pregnant
Did not mind either way

Intend to get pregnant
0
1
2
Just before falling pregnantNot want a baby
Have mixed feelings about having a baby

Want a baby
0
1
2
Before falling pregnant, had you and your partnerNever discussed children
Discussed children but no firm agreement

Agreed to the pregnancy
0
1
2
Health actions before falling pregnantaNo actions
1 action

2 or more actions
0
1
2

We aimed to determine the prevalence of intended, ambivalent and unplanned pregnancies in an urban Australian population of women presenting to a public hospital antenatal service within the Sydney Local Health District (SLHD) and to investigate the social and demographic factors associated with unintended pregnancy. The health behaviours of women who do and do not intend their pregnancies were also investigated. We hypothesised that unintended pregnancies would comprise one-third of the total pregnancies in this population, a figure consistent with previous international studies of this kind. Further, it was expected that demographic factors would have a significant effect on pregnancy intention – most prominently, younger, unmarried women and those of a low socioeconomic status would experience a comparatively greater rate of unintended pregnancy.

Methods

From October 2010 to April 2011, reception staff invited all women presenting to the antenatal clinic, ultrasound service and the birth centre at a large metropolitan hospital in the SLHD to participate in this study. Consent was obtained in the form of a patient information sheet. The questionnaires were printed in English and were also translated into Chinese, Korean, Bengali, Turkish and Thai languages, the most commonly spoken languages in patients attending this hospital. Women who were unable to satisfactorily read, understand and complete questionnaires in these languages or women who appeared distressed were excluded from the study. Women who indicated they had completed the questionnaire at another location were not given the questionnaire a second time.

The self-administered questionnaire comprised 15 questions, six of which assessed pregnancy intention, according to the validated pregnancy intention instrument (Table 1). The questionnaire was designed such that each response to the six questions was scored out of two, and summed to give a final pregnancy intendedness score between zero and 12. The authors of the instrument suggest that intention scores can be divided into three groups: zero-three (unplanned), four-nine (ambivalent) and ten-twelve (planned).8

The remaining questions covered the pregnancy service where the questionnaire was completed and the type of contraception used at the time of conception by those not intending to fall pregnant. Demographic information was collected by asking participants to select an age group, marital status, employment status and education level. Other questions asked about cultural background, religion and postcode in a free-response format. Age was grouped into those under 25 years of age, and 25 and over, as those under 25 are widely recognised to be the age group most at risk of sexual and reproductive ill health.10,11 Information on parity was not collected. Respondents were attributed a socioeconomic status (SES) according to their postcode, using the deciles in the Socio-economic Indexes for Areas (SEIFA), Index of Relative Socio-economic Advantage and Disadvantage, 2006.12 Deciles one to three were considered low SES, four to seven middle SES and eight to ten high SES.

Results of the questionnaires were entered into a purpose-designed database and no identifying data was attached to the questionnaire.

The Sydney South West Area Health Services Ethics Review Committee and The University of Sydney Human Research Ethics Committee approved the questionnaire and study design.

Statistical analysis

The data were analysed using SPSS version 19.0.13 The prevalence of unintended pregnancy was calculated using data from women who completed all six questions in the pregnancy intention instrument. In this study, women with pregnancy intendedness scores less than 10 (including both ambivalent and unplanned pregnancies) were considered unintended. Univariate analyses were run between demographic factors or uptake of health behaviours and pregnancy intention. Degree of association was assessed using chi-squared tests. The demographic characteristics with a p value <0.25 were then examined using logistic regression analyses. Odds ratios and 95% confidence intervals were calculated to determine the degree of association between demographic characteristics and pregnancy planning (alpha-level=0.05).

Results

Of the 1,554 questionnaires collected, 1,309 (84.2%) women chose to participate in the study. The remainder indicated their non-participation by returning a blank questionnaire. A total of 1,218 (78.4%) women completed the six questions in the pregnancy intention instrument and the following analysis of data is based on those 1,218 completed questionnaires.

Women were sampled from one of three antenatal areas; the antenatal clinic (701; 57.5%), the ultrasound service (353; 29.0%) or the birth centre (164; 13.5%).

Demographic characteristics

Table 2 shows the frequency distribution of demographic characteristics of the women participating in this study. Women aged under 25 years comprised a small portion of the sample (92; 7.6%) compared to those over 25 years (1,121; 92.0%). Similarly, single and separated or divorced women comprised a small portion of the sample (75; 6.2%) compared to de facto and married women (1,139; 93.5%). Employed women (883; 72.5%) constituted a larger portion of the sample than unemployed women (328; 26.9%). In terms of education, the largest proportion had a post-secondary education or diploma (930; 76.4%), fewer had attained school only (221; 18.1%) and even fewer had a foreign qualification (44; 3.6%). Women with a religion (681; 55.9%) involved a greater proportion of the sample than women with no religion (426; 35.0%). Cultural backgrounds were grouped into Australian (494; 40.6%), Asian (283; 23.2%), European and American (234; 19.2%), other Oceania (53; 4.4%) and African and Middle Eastern (48; 3.9%). Finally, almost all women were considered to be of high SES (1,161; 95.3%), very few were middle SES (26; 2.1%) and even fewer were low SES (8; 0.7%).

Table 2. Frequency distribution of demographic characteristics.
NFrequency
Age, n (%)
 16–17 years old
 18–24 years old
 25–34 years old
 35–39 years old
 40–45 years old
 >45 years old
 Data missing

2 (0.2)
90 (7.4)
726 (59.6)
338 (27.8)
56 (4.6)
1 (0.1)
5 (0.4)
Marital status, n (%)
 Single
 De facto
 Married
 Separated/divorced
 Data missing

53 (4.4)
285 (23.4)
854 (70.1)
22 (1.8)
4 (0.3)
Employment status, n (%)
 Working full-time
 Working part-time
 Studying full-time
 Studying part-time
 Unemployed
 Pension
 Data missing

576 (47.3)
307 (25.2)
45 (3.7)
25 (2.1)
257 (21.1)
1 (0.1)
7 (0.6)
Religion, n (%)
 No religion
 Christian
 Buddhist
 Hindu
 Muslim
 Minority religion
 Data missing

426 (35.0)
538 (44.2)
46 (3.8)
52 (4.3)
41 (3.4)
4 (0.3)
111 (9.1)
Cultural group, n (%)
 European & American
 Asian
 African & Middle Eastern
 Other Oceania
 Australian
 Data missing

234 (19.2)
283 (23.2)
48 (3.9)
53 (4.4)
494 (40.6)
106 (8.7)
Socioeconomic status, n (%)
 Low
 Medium
 High
 Data missing

8 (0.7)
26 (2.1)
1161 (95.3)
23 (1.9)
Education attained, n (%)
 School only
 Any foreign qualification
 Post-secondary qualification
 Data missing

221 (18.1)
44 (3.6)
930 (76.4)
23 (1.9)

Pregnancy intention

Overall, 394 (32.4 %) pregnancies in the antenatal population were unintended, of which 365 (30.0 %) were ambivalent and 29 (2.4%) were unplanned. The remaining 824 (67.6 %) pregnancies were considered intended.

Health behaviours

Approximately one-third of women (34.6%; 422) did not take on any positive health actions such as stopping smoking, taking folic acid, reducing alcohol intake or seeking medical advice prior to pregnancy. The uptake of positive health behaviours was associated with pregnancy intention status, with 23.1% of women (n=91) with unintended pregnancies taking up any health actions compared with 85.6% of women (n=705) with intended pregnancies (p<0.001).

Contraceptive use

Of the 394 women experiencing an unintended pregnancy, 203 (51.5%) were not using contraception at the time of conception. Of the 191 (48.5%) women using contraception at the time of conception, approximately half (96; 50.3%) were using contraception consistently and the other half were using it intermittently (95; 49.7%). The contraceptive methods used by these 191 women were condoms (86; 45.0%), the oral contraceptive pill (47; 24.6%), withdrawal (12; 6.3%), natural methods (12; 6.3%), implants (2; 1.0%), diaphragm (2; 1.0%) and intrauterine devices (IUDs; 2; 1.0%). Data were missing for 28 women (14.7%).

Univariate analysis

Table 3 shows the univariate association between unintended pregnancy and demographic characteristics. The demographic factors associated with unintended pregnancy on univariate analysis were age, marital status, cultural background, employment status, education level and religion.

Table 3. Univariate associations between unintended pregnancy and potential covariates.
NUnintended pregnancyp-value*
 Yes

(n= 394)a
No

(n= 824)a
 
  1. Notes:

  2. * p-value for categorical variables, calculated by χ2 test

  3. a

    N varies between factors due to missing data

Age, n (%)
 <25 years old
 >25 years old

47 (51.1)
345 (30.8)

45 (48.9)
776 (69.2)
<0.001
Marital status, n (%)
 Single or separated or divorced
 De facto or married

51 (68.0)
341 (29.9)

24 (32.0)
798 (70.1)
<0.001
Employment status, n (%)
 Unemployed
 Employed

127 (38.7)
264 (29.9)

201 (61.3)
619 (70.1)
0.004
Religion, n (%)
 Religious
 Not religious

233 (34.2)
122 (28.6)

448 (65.8)
304 (71.4)
0.053
Cultural group, n (%)
 European & American
 Asian
 African & Middle Eastern
 Other Oceania
 Australian

69 (29.5)
122 (43.1)
13 (27.1)
19 (35.8)
133 (26.9)

165 (70.5)
161 (56.9)
35 (72.9)
34 (64.2)
361 (73.1)
<0.001
Socioeconomic status, n (%)
 Low
 Medium
 High

2 (25.0)
10 (38.5)
374 (32.2)

6 (75.0)
16 (61.5)
787 (67.8)
0.722
Education attained, n (%)
 School only
 Any foreign qualification
 Post-secondary qualification

96 (43.4)
16 (36.4)
274 (29.5)

125 (56.6)
28 (63.6)
656 (70.5)
<0.001

Multivariate analysis

Table 4 shows the multivariate associations of predictor variables of unintended pregnancy. Unintended pregnancy was significantly associated with age, marital status and cultural background. There was no association between unintended pregnancy and employment status, religious affiliation or education level.

Table 4. Multivariate associations of predictor variables of unintended pregnancy.
 OR (95% CI) 
  1. Notes:

  2. a

    Compared to post-secondary education or diploma

  3. b

    Compared to Australians

  4. *p<0.05; ** p<0.01

Age <251.86 (1.10–3.14)*
Marital status: Single/ Separated/ Divorced6.08 (3.40–10.87)**
Education: Foreign qualificationa0.89 (0.42–1.89) 
Education: School onlya1.35 (0.92–1.65) 
Employment: Unemployed1.02 (0.74–1.42) 
Religion: Religious1.26 (0.94–1.68) 
Cultural background: Other Oceaniab1.42 (0.72–2.81) 
Cultural background: European and Americanb1.28 (0.89–1.85) 
Cultural background: African and Middle Easternb1.16 (0.58–2.33) 
Cultural background: Asianb2.45 (1.76–3.42)**

Discussion

Of women attending antenatal services at this large metropolitan hospital in the SLHD during the study period, only two-thirds of pregnancies were clearly intended (both desired and occurring at the correct time). The remaining third of pregnancies was unintended (mistimed or unwanted). The pregnancy intention was ambivalent in just less than one- third of all pregnancies, and one in fifty pregnancies were unplanned. This is consistent with previous studies conducted in the UK and the US that found one-third of pregnancies destined to end in childbirth are unintended.1,14,15

In our study, more than half of all pregnancies in women under 25 years old were unintended. This finding is unsurprising and has been documented in a number of previous studies.1,15 Younger women have higher fertility, higher frequency of sexual intercourse and higher rates of contraceptive failure.16 In contrast, older women are more likely to be married and starting families and may be more able to accommodate a child.15 Accordingly, we found that women who were single, separated or divorced were more than six times as likely to experience an unintended pregnancy. Married women are better able to control their fertility, as greater relationship intimacy is positively associated with contraception use at last sexual encounter and use of more reliable methods of contraception.17

In our study, Asian women experienced significantly more unintended pregnancies than other ethnicities. This may reflect differences in contraceptive practices, as previous studies have shown Asian women have lower rates of contraceptive use than non-Asians.18 Migrant Asian women are also more likely to use unreliable methods of contraception such as condoms,19 and less likely to use the most effective methods like the IUD.18

Surprisingly, on univariate analysis, SES was not shown to have an effect on prevalence of unintended pregnancy. On multivariate analysis, education level and employment status were not shown to have a significant effect on the prevalence of unintended pregnancy. The women in this study belonged to a very narrow portion of the socioeconomic spectrum, which reflects the SLHD as a whole. The eight Local Government Areas that comprise the SLHD are considered to be high SES according to the SEIFA Index of Relative Socio-economic Advantage and Disadvantage, 2006.12 Therefore, this sample is not suitable for studying the relationship between unintended pregnancy and SES and its contributing factors, education level and employment status.

There was also no significant effect of religion on the prevalence of unintended pregnancy. There is conflicting information regarding religion and its relationship to pregnancy intention. Some studies have shown no effect of religion on unintended pregnancy rates,20 while others show that women who are not religious are more likely to experience an unintended pregnancy, and more likely to end it in termination.21

Contraceptives were used by just less than half of women experiencing an unintended pregnancy. The most common contraceptives to have failed were condoms and the oral contraceptive pill. This is a product of the prevalence of their use and their ease of misuse. These data support findings from the US and Europe where the ‘failure’ of methods requiring daily decisions still contributes to half to two-thirds of unintended pregnancies.22,23 International experts believe that an effective strategy to address this issue is to improve access to long-acting reversible contraception including implants, injectables and IUDs, which do not require daily compliance.24 These methods are highly reliable, reversible and are also cost-effective, principally because they reduce the risk of unintended conceptions.

As part of the pregnancy intention scale, women were asked whether they had taken on any positive health actions, and the women who had done so were significantly more likely to have planned their pregnancy. However, a significant proportion of those with a planned pregnancy were not taking on any positive health actions. This is consistent with other studies finding that although awareness of folic acid and its relation to neural tube defects may be high, this is not matched by peri-conceptual uptake. Pregnancy planning is one of the main barriers. Public health experts in Australia and elsewhere have recognised this gap and fortification of the flour in breads has occurred since September 2009 in Australia.25 Where folic acid fortification of grains occurs, it may be sufficient to prevent neural tube defects, with a recent US study indicating that folic acid supplementation may not offer further benefit for reducing risk of spina bifida.26 It is therefore imperative that policy makers are made aware of the high prevalence of unintended pregnancy and the requirement for strategies that do not involve active behavioural changes on the behalf of the mother, in order to prevent adverse pregnancy outcomes.

A limitation of this study is the difficulty in applying this data from one metropolitan hospital to the general Australian population. For a more accurate estimate on the total rate of unintended pregnancy in Australia, women who seek abortions would need to be included. Furthermore, the women in this study were largely of high SES. Repetition of this study in other centres including women from areas of relative socioeconomic disadvantage, and from varying ethnic groups, would again allow a more accurate estimate of the extent of unintended pregnancy in Australia. Such studies have occurred in the UK and US, where the data inform and drive policy to reduce unintended pregnancies, particularly in teenagers. Another issue with this study is response bias; participants may not want to indicate their pregnancy was unintended, as they may equate this with an unwanted pregnancy. We did not collect information on parity, which may have affected pregnancy intention, as higher parity has been associated with increased rates of unintended pregnancy.27 Finally, the pregnancy intention instrument used in this study has not been validated within the Australian population and the demographic characteristics on which the instrument was validated vary from those of the population of Australian women in this study, which may have an impact on interpretation of the questions and, subsequently, women's answers to these questions.

Conclusions

This study was an initial investigation into the state of pregnancy intention in an Australian antenatal population using a previously validated instrument. One-third of pregnancies destined to end in childbirth were unintended. More than one-third of women did not take any health actions prior to pregnancy. This has implications for the implementation of public health strategies directed at improving pre-conceptual care. While further efforts are needed to inform the population about the benefit of lifestyle modifications and folate supplementation prior to a planned pregnancy, policy makers should also be made aware of the prevalence of unintended pregnancy and the requirement of strategies that do not rely upon conscious behaviour change – such as folate fortification of bread flours – in order to protect against adverse pregnancy outcomes. The findings also have implications for the provision of contraceptive services, in that greater emphasis should be placed on encouraging use of more reliable methods of contraception such as long-acting reversible contraceptives, particularly for young women who are most at risk of unintended pregnancy.

Ancillary