Moreau 2016
Moreau 2016
Moreau 2016
⁎
risk perceptions☆ Caroline Moreaua,b,c, , Aline Bohetb aDepartment of Population, Family and
Reproductive Health, Johns Hopkins Bloomberg School of Public Health, 615 North Wolfe Street, Baltimore, MD, 21205, USA b Gender, Sexual
and Reproductive Health, CESP Centre for Research in Epidemiology and Population Health, U1018, INSERM, F-94807, Kremlin, Bicêtre,
France c Institut National d'Etudes Démographiques, F-75020, Paris, France Received 27 October 2015; revised 26 February 2016; accepted 28
February 2016
Abstract
Objectives: This study is to assess frequency and correlates of women's reports of unintended pregnancy risk in the general
population in France. Study design: Data are drawn from the FECOND survey, a national probability survey on sexual and
reproductive health conducted in France in 2010. We identified 2969 women ages 15–49 years who had heterosexual intercourse
in the 4 weeks prior to the survey and who were at potential risk of unintended pregnancy. We evaluate women's reports of
unintended pregnancy risk in the last 4 weeks and identify correlates of such reports using logistic regression modeling. Results:
Fifteen percent of women thought that they could have become pregnant in the last 4 weeks without wanting to do so. Reports of
unintended pregnancy risk were higher among women in very difficult financial situations (OR=1.87 [1.32–2.65]) and
foreign-born women (OR=1.53 [1.03–2.29]). Exposure in the form of contraceptive practices and errors in use of contraception
were the strongest correlates of women's reports of unintended pregnancy risk, yet among the 9.8% of women who reported
inconsistent use of contraception or unprotected intercourse in the last 4 weeks, 63% did not think they could have become
pregnant unintentionally. Conclusions: Significant discrepancies between pregnancy exposure and women's report of unintended
pregnancy risk call for better SRH educational programs to improve pregnancy awareness in the general population. On the other
hand, targeted interventions toward women who report being at risk of unintended pregnancy may contribute toward reducing
unintended pregnancies given the frequency of such events. © 2016 Elsevier Inc. All rights reserved.
Keywords: Pregnancy intention; Risk perception; Family planning; Europe
1. Introduction
While contraceptive use is widespread in France, with only 3% of women with an unmet need for contraception and
high levels of modern method usage [1], unintended pregnancies remain frequent with more than one in three
pregnancies reported as such in 2010 [2]. Most of these unintended conceptions are due to lack of use, error of use or
inconsistent use of contraception [3,4]. Interventions to reduce the risk of pregnancy after unprotected intercourse or
errors of use of contraception (for instance by providing advance supplies of emergency contraception) have had
little impact on unintended pregnancy rates at the population level [5,6]. The primary reason is that women do not
perceive
themselves as being at risk of pregnancy and therefore do not use backup contraception to reduce their risk [7,8].
Likewise, studies exploring reasons for unmet need for contraception suggest that many women believe they have
very limited risks of becoming pregnant despite having unprotected intercourse [5,7]. While some women are at low
risk because they have infrequent sexual intercourse [9], others mistakenly think they are at low risk because they
have not become pregnant from previous unprotected intercourse [7]. Misper- ceptions of pregnancy risk may be a
critical target for family planning interventions, yet there is limited knowledge about the frequency and correlates of
women's perceptions of being at risk of an unintended pregnancy at the population level. Thus, the aim of this study
is to identify women who report they could have become pregnant over the last 4 weeks without wanting to do so
and identify the correlates of such
☆
None of the authors have competing interests. ⁎ Corresponding author. Tel.: +1-4105028951.
perceptions. We consider proximate factors related to pregnancy exposure in the form of frequency of intercourse,
E-mail address: cmoreau2@jhu.edu (C. Moreau).
contraceptive use and errors or inconsistent use, as well as
http://dx.doi.org/10.1016/j.contraception.2016.02.029 0010-7824/© 2016 Elsevier Inc. All rights reserved.
Contraception 94 (2016) 152–159
sociodemographic and past or current sexual and reproductiveisk.
characteristics that inform both exposure and perceptions of
professional and financial situation and health insurance), as
well as their past and current sexual and reproductive
2. Materials and methods haracteristics (parity, history of unintended pregnancy,
pregnancy intentions). We also assessed women's reports of
Data are drawn from the FECOND survey, a national being at risk of an unintended pregnancy as a function of
survey on sexual and reproductive health carried out in France exposure, including frequency of sexual intercourse, contra-
in 2010. A sample of 5272 women ages 15–49 years residing in ceptive usage and errors of use/nonprotected intercourse in the
France were included in the study using random digit dialing. last 4 weeks. We also considered information on date of last
The refusal rate was estimated at 20% [10]. The survey was menstrual cycle to explore how reports of recent unintended
approved by the Commission Nationale de l'Informatique et des pregnancy risk could vary according to week of menstrual
Libertés and secondary analysis was approved by the cycle.
Bloomberg School of Public Health Institutional Review Board We conducted a multivariate logistic regression to assess
at Johns Hopkins University. the independent effects of the sociodemographic and sexual and
We restricted our sample to the 3934 women who had reproductive factors on women's reports of unintended
heterosexual intercourse in the last 4 weeks, in order to study pregnancy risk in the absence or in the presence of factors
recent perceptions of pregnancy risk. Women who were assessing pregnancy exposure (frequency of intercourse,
pregnant or trying to conceive (n= 478) and women who were contraceptive usage and errors of use). We also present results
sterile, had a tubal ligation or whose partner's were sterile or had from the most parsimonious model (using Akaike information
a vasectomy (n=403) were not asked about their risk of criterion) to evaluate the robustness of our results.
becoming pregnant unintentionally in the last 4 weeks. We In a final analysis, we examined reports of unintended
further excluded women who dropped out of the survey before pregnancy risk among the 277 women who indicated they were
responding to the question on pregnancy risk (n= 79) or did not not using contraception or an error/inconsistent use of
provide an answer to the question (n= 5). Thus, our final study contraception in the last 4 weeks. We only present results from
population comprised 2969 women who had had heterosexual bivariate analysis because we found that none of the factors
intercourse in the 4 weeks prior to the survey and who were at were related to unintended pregnancy risk in this subgroup of
potential risk of an unintended pregnancy. the population.
Women responded by phone to a multithematic ques- tionnaire,
which lasted on average 41 min. They provided information on
their sociodemographic background and were questioned on a 3. Results
number of sexual and reproductive health topics including their
pregnancy history, recent sexual activity and current The characteristics of women who were sexually active
contraceptive usage. Women described their current method of in the last 4 weeks and potentially at risk of an untended
contraception, how satisfied they were with their method and if pregnancy are presented in Table 1. The mean age of these
they had a problem using their method in the last 4 weeks women was 33.7 years [33.3–34.1]. Three quarters of the
(missed pills, condom slippage or breakage). Additionally, women were living with a partner at the time of the survey and
women indicated if they had had unprotected intercourse in the two thirds had at least one child (Table 2).
last 4 weeks and, if so, if they had used subsequent protection A third of women reported a history of unintended
including emergency contraception. In another section of the pregnancy and 5% indicated they had had a sexually transmitted
questionnaire (the sexual health module), women responded to infection (STI) (excluding vaginal yeast infec- tion) in the last 5
the question “do you think you could have become pregnant in years (Table 2). Half of women were on the pill at the time of
the last 4 weeks without wanting to do so?”. This question was the survey, a quarter was using long-acting methods (23.6%
used to define our outcome measure of “reports of unintended using IUDs and 2.9% using the implant), 12.4% used condoms
pregnancy risk in the last 4 weeks”. and 5.8% relied on barrier or natural methods (withdrawal,
We used bivariate statistics to describe women's spermicides or fertility awareness method). Less than 2% of
reports of being at risk of unintended pregnancy in the last 4 women were not using contraception while at potential risk of
weeks according to their sociodemographic characteristics (age, an unintended pregnancy, and 8.3% indicated that they had
country of birth, level of education, cohabitation status, missed a pill, had a problem with a condom or had unprotected
intercourse in the last 4 weeks. Reports of unintended pregnancy risks were also related
A total of 15.1% (13.7–16.7%) of women thought theyto pregnancy intentions and past sexual and reproductive health
could have become pregnant in the last 4 weeks withoutexperiences. Women with a history of unintended pregnancy
wanting to do so (Table 1). Such reports were higher amongwere more likely to report that they could have become
women under the age of 30 years, among those in difficultpregnant unintentionally (Table 2). This was also the case of
financial situation, and foreign-born women (Table 1). women who wanted more children or whose partner
153 C. Moreau, A. Bohet / Contraception 94 (2016) 152–159
equency of sexual intercourse or timing of last menses was not
sociated with women's report of risk. Conversely reports of
nintended pregnancy risk were strongly correlated with current
ontraceptive usage and inconsistent/errors of use. More than a
ird of women who indicated a problem with their method or
n act of unprotected intercourse in the last 4 weeks thought that
ey might have become pregnant unintentionally in the same
me period versus 13% of those who did not report such events.
ikewise, reports of unintended pregnancy risk were highly
orrelated with contraceptive practices (Table 2). More than a
ird of women who were not using a method (34.6%) thought
at they might have become pregnant unintentionally (Table 2).
ikewise, 39.5% of women using natural or barrier methods
her than condoms reported that they could have become
pregnant in the last 4 weeks without wanting to do so.
Interestingly, this fear was also present among 8.6% of IUDs
users and 7.7% of implant users.
Results from the multivariate logistic model confirmed
some of the results from the bivariate analysis (Table 4).
Specifically, the odds of reporting a risk of an unintended
pregnancy were higher among women who were in very
difficult financial situations and among women who were
foreign born. This was true regardless of the model considered.
Higher-risk perceptions were also found in women who were
not satisfied with their current sexual life, after adjusting for
pregnancy exposure. This variable was not included in the
parsimonious model however. Women's history of an
unintended pregnancy was only marginally correlated with risks
perceptions after adjusting for other characteristics. In models
adjusting for pregnancy exposure and in the parsimonious
wanted more children (Table 3). Conversely, women's attitudes model, women using no method or less effective methods of
toward terminating or continuing an unintended pregnancy were contraception were equally likely to believe they could have
not associated with women's report of unintended pregnancy become pregnant unintentionally in the last 4 weeks as
risk. compared to those using more effective methods of
Reports of unintended pregnancy risk were contraception. Adjusting for contraceptive usage,
inconsistently related to pregnancy exposure. On the one hand,
154 C. Moreau, A. Bohet / Contraception 94 (2016) 152–159
All women at potential risk of unintended pregnancya Nonuser of contraception or reporting error of use
% report pregnancy
% all women
risk
(n=2969)
% report pregnancy
% report pregnancy
risk
risk
% report pregnancy
p % women
risk
p % women
p
(n=277) p
(n=277) p
% report pregnancy p
risk
a
Women at potential risk of pregnancy in the 4 weeks prior to the survey were defined as having had heterosexual intercourse in the last 4 weeks and
All women at potential risk of unintended pregnancya Nonuser of contraception or reporting error of use
% report pregnancy
% women (n=2969) % report
risk
pregnancy risk % report pregnancy
p % women risk
(n=277) p
% report pregnancy p
risk p
% report pregnancy
risk
Cohabitation status of relationship No partner 7.6 14.8 .46 7.3 33.8 .89 Noncohabitating partner 18.2 17.1 22.7 35.1 Cohabitating partner 74.2 16.6 70.0 38.1
Duration of relationship
≤1 year or no partner 13.2 21.2 .002 19.1 44.2 .31 N1 year 86.8 14.1 80.9 35.4 Sexual orientation
Heterosexual 97.7 14.9 .13 96.2 45.3 .61 Bisexual 2.3 21.7 3.8 36.8 Number of births
0 36.0 14.8 .45 39.2 36.7 .47 1 15.3 17.3 15.2 46.2 N1 48.7 14.5 45.6 34.4 History of unintended pregnancy
Yes 33.0 18.1 .006 36.3 44.5 .09 No 67.0 13.6 63.7 32.9 STI in last 5 years
Yes 5.8 20.4 .08 6.5 63.2 .03 No 94.2 14.7 93.5 35.3 Gynecological follow-up in the
last year Yes 92.4 14.9 .09 86.7 38.0 .66 No 7.6 19.9 13.3 33.7 Satisfied with current sexual life
Yes 95.3 14.8 .08 95.5 36.7 .52 No 4.7 21.0 4.5 46.0 Number of acts of sexual intercourse in last 4 weeks b5 acts 29.0 16.4 .10 29.0 41.0 .68 5–9 acts 27.4 12.5
30.0 33.5 10+ 43.4 15.9 41.1 37.1 Last menses
In the last 7 days 24.7 18.1 .19 30.0 38.2 .35 Between 7 and 14 days 18.0 15.2 16.8 32.8 Between 14 and 21 days 20.0 14.3 32.1 34.9 More than 21 days 25.0 12.8
17.7 39.1 Don't know/missing 12.4 18.3 3.3 72.4 Current use of contraception
No contraception 1.5 34.6 b.001 LARC methods 26.5 8.5 Hormonal methods 1.5 15.7 Pill 52.2 12.7
Condom 12.4 25.2 Other natural or barrier method 5.8 39.5 Errors of use/unprotected sex in
last 4 weeks Yes 8.3 37.5 b.001 No 91.7 13.0
a
Women at potential risk of pregnancy in the 4 weeks prior to the survey were defined as having had heterosexual intercourse in the last 4 weeks and
Table 3 Associations between women's attitudes and beliefs regarding pregnancy risks, couple's pregnancy intentions and reports of unintended pregnancy
risk
All women at potential risk of unintended pregnancya Nonuser of contraception or reporting error of use
% report pregnancy
% women
risk
(n=2969)
% report pregnancy
% report pregnancy
risk
risk
% report pregnancy
p % women
risk
p % women
p
(n=277) p
(n=277) p
% report pregnancy p
risk
a
Women at potential risk of pregnancy in the 4 weeks prior to the survey were defined as having had heterosexual intercourse in the last 4 weeks and
were not
pregnant, sterile or trying to conceive.
tudy, the greater proportion of women who report risk among
hose who reported a history of unintended pregnancy suggest
hat past reproductive experiences inform current reports of risk.
While many women do not recognize their risk of
pregnancy when exposed, others worry about pregnancy in the
bsence of exposure. This was true of 8% of LARC users in our
tudy. The discrepancy between exposure and reports of risk is
both substantial and intriguing calling for a better understanding
of the ways women and men construct and reevaluate their
perceptions of risk based on prior and current experiences. Lack
of reproductive control or the notion that
“pregnancy just happens” regardless of preventive behaviors,
described in a study among low-income women in Pittsburgh,
may explain some of our findings [16].
Perceptions of pregnancy risk need also be interpreted in
the context of women's relational social and economic
circumstances. The lack of information about women's
relational context did not allow an exploration of partnership
influences on pregnancy risk perception with the exception of
partner's pregnancy intentions, found to have an independent
effect on women's perceptions of pregnancy risk. Financial
difficulty, being foreign born and lower educational level,
which are markers of social and economic disadvantage in
France, were strongly correlated with higher
Analysis conducted among women at potential risk of pregnancy in the 4 weeks prior to the survey who were defined as having had heterosexual intercourse in the
last 4 weeks and were not pregnant, sterile or trying to conceive.
perceptions of pregnancy risk in our study, irrespective of eir predictive association with reproductive outcomes.
pregnancy exposure. Gatny and colleagues reported similar
results in a study of pregnancy scares among young women in
the State of Michigan showing African American women and Conclusions
those with lower GPA [17] were more likely to experience such
events. In this study, the authors identified “pregnancy scare” as This study indicates significant levels of unintended
any “probable pregnancy” in a given week that is not confirmed regnancy risk perceptions in the general population in France,
by a pregnancy test among women who did not intend to ostly informed by contraceptive practices and errors of use.
become pregnant. This construct is distinct from our measure of et, results reveal large discrepancies between pregnancy
unintended pregnancy risk perceptions, although in both cases, xposure and perceptions of pregnancy suscep- tibility calling
it captures women who perceive themselves as being at risk of a r greater attention to perceptions of pregnancy risk as a driver
pregnancy they did not wish for. In the Michigan study, the f unprotected sexual intercourse. In particular, a majority
authors found that one in five unintended pregnancies identified omen who have an unmet need for contraception or report
during the 2-year follow-up period of the study was preceded by rors of use do not think they can become pregnant
a “pregnancy scare”, while 40% of young women who reported ccidentally, which calls for better SRH educational programs
a pregnancy scare experienced an unintended pregnancy in improve pregnancy awareness. On the other hand, targeted
subsequent months [17]. Zabin and colleagues also reported that terventions toward women who
adolescents with negative pregnancy test were at high risk of perceive themselves at risk may contribute toward reducing
experiencing unintended pregnancies in subsequent months, unintended pregnancies given the frequency of such
suggesting that women who identify an instance where they perceptions.
think they might have become pregnant without wanting to do
so may be an appropriate target group for intervention [18]. Acknowledgments
Subsequent studies should explore the link between pregnancy
risk perceptions and pregnancy scares and the association of risk We thank all women who participated in the FECOND study.
perceptions with subsequent unintended pregnancies. The FECOND study was supported by a grant from the French
This study has a number of limitations, including the Ministry of Health, a grant from the French National Agency of
cross-sectional design of the survey, which limits the Research (#ANR-08-BLAN-0286-01; PIs N. Bajos, C Moreau)
interpretability of some of the associations including the link and funding from National Institute of Health and Medical
between risks perceptions and contraceptive behaviors. The Research and the National Institute for Demographic Research.
construct of pregnancy intentions is also multidimen- sional
[18]. Some individuals may desire to have children but because References
of life circumstances may not intend to have any [19]. Others
may not intend to have children but would be happy if they
were faced with an unplanned pregnancy [20,21]. These [1] Bajos N, Bohet A, Le Guen M, Moreau C, the Fecond team. Contraception
in France: new context, new practices? Popul Soc 2012;492:1–4. [2] Bajos N,
dimensions are not captured in the current analysis but could
Aline Bohet A. Effectiveness of family planning policies: the abortion paradox.
further elucidate the construct of pregnancy risk perceptions and
PLoS One 2014;9(3):e91539, http://dx.doi.org/ 10.1371/journal.pone.0091539. patterns, 2004. Perspect Sex Reprod Health 2007;39(1):48–55. [10] Legleye S,
[3] Moreau C, Trussell J, Desfreres J, Bajos N. Patterns of contraceptive use Bohet A, Razafindratsima N, Bajos N, Moreau C. Improving survey
before and after an abortion: results from a nationally representative survey of partcipation. Public Opin Q 2013;77:666–95. [11] Sonfield A, Hasstedt K, Gold
women undergoing an abortion in France. Contraception 2010;82:337–44. [4] RB. Moving forward: family planning in the era of health reform. New York:
Goulard H, Moreau C, Gilbert F, Cocon Group, Bajos N. Contraceptive Guttmacher Institute; 2014. [12] Rosenstock IM. Historical origins of the health
failures and determinants of emergency contraception use. Contraception belief model. Health
2006;74:208–13. [5] Raymond EG, Trussell J, Polis C. Population effect of Educ Monogr 1974;2:328–35. [13] Nettleman MD, Chung H,
increased access to emergency contraceptive pills: a systematic review. ObstetBrewer J, Ayoola A, Reed PL. Reasons for unprotected intercourse: analysis of
Gynecol 2007;109:181–8. [6] Polis CB, Grimes DA, Schaffer K, Blanchard K,the PRAMS survey. Contraception 2007;75:361–6. [14] Lakha F, Glasier A.
Glasier A, Harper CC. Advance provision of emergency contraception for Unintended pregnancy and use of emergency contraception among a large
pregnancy prevention. Cochrane Database of Systematic Reviews 2007 cohort of women attending for antenatal care or abortion in Scotland. Lancet
2010(2):CD005497, http://dx.doi.org/10.1002/14651858.CD005497.pub2. [7] 2006;368:1782–7. [15] R ahman M, Berenson A, Herrera S. Perceived
Williamson LM, Buston K, Sweeting H. Young women's perceptions of susceptibility to pregnancy and its association with safer sex, contraceptive
pregnancy risk and use of emergency contraception: findings from a qualitativeadherence and subsequent pregnancy among adolescent and young adult
study. Contraception 2009;79:310–5. [8] Moreau C, Bouyer J, Goulard H, women. Contraception 2013;87:437–42. [16] Borrero S, Nikolajski C, Steinberg
Bajos N. The remaining barriers to the use of emergency contraception J, Freedman L, Akers A, Ibrahim S, et al. “It just happens”: a qualitative study
perception of pregnancy risk by women undergoing induced abortions. exploring low-income women's perspectives on pregnancy intention and
Contraception 2005;71(3):202–7. [9] Frost JJ, Singh S, Finer LB. U.S. planning. Contra- ception 2015;91:150–6.
women's one-year contraceptive use
158 C. Moreau, A. Bohet / Contraception 94 (2016) 152–159
[17] Gatny H, Kusunoki Y, Barber J. Pregnancy scares and subsequent [20] Miller WB. Childbearing motivations, desires, and intentions: a theoretical
unintended pregnancy. Demogr Res 2014;31:1229–42. [18] Z abin framework. Genet Soc Gen Psychol Monogr 1994;120(2):223–58. [21] Aiken
LS, Emerson MR, Ringers PA, Sedivy V. Adolescents with negative pregnancy AR, Dillaway C, Mevs-Korff N. A blessing I can't afford: factors underlying the
test results. An accessible at-risk group. JAMA 1996;275:113–7. [19] Santelli paradox of happiness about unintended pregnancy. Soc Sci Med
JS, Lindberg LD, Orr MG, Finer LB, Speizer I. Toward a multidimensional 2015;132:149–55, http://dx.doi.org/ 10.1016/j.socscimed.2015.03.038 [Epub
measure of pregnancy intentions: evidence from the United States. Stud Fam 2015 Mar 19].
Plan 2009 Jun;40(2):87–00.
159 C. Moreau, A. Bohet / Contraception 94 (2016) 152–159