Abstract
Objectives
The abortion rate varies greatly within the French overseas territories including the Caribbean island of Guadeloupe and La Réunion in the Indian Ocean. We compare women’s contraceptive paths surrounding an abortion in both territories.
Methods
The data for this study are part of a nationally representative survey of women undergoing abortion in France in 2007. The analysis included 1211women from Guadeloupe and 1531 from La Réunion.
Results
Results show differences in women’s use of contraception before the abortion by study location. Women in Guadeloupe were more likely not to have used contraception in the month they conceived (40% vs. 32%, p < 001). Among those using no contraception or less effective contraception before the abortion, 74% in Guadeloupe and 86% in La Réunion received a prescription for a very effective method such as a hormonal method or intrauterine device after the procedure. In both settings, women with no health insurance or a government health plan were 70% less likely to have received a prescription for a very effective method.
Conclusions
While this study shows a significant increase in the prescription of very effective methods, it also indicates the ineffectiveness of the health care system in closing the gap in the pre-abortion contraceptive disparities observed between Guadeloupe and La Réunion.
Keywords: Contraception, Post abortion care, Contraceptive failure, Guadeloupe, La Réunion, France, Nationally representative survey
INTRODUCTION
The abortion rate varies greatly between metropolitan France and the French overseas territories including the Caribbean island of Guadeloupe (400,500 habitants) and the island of La Réunion (800,000 habitants) in the Indian Ocean, East of Madagascar (Table 1)1. While these territories abide by the same laws as metropolitan France, benefit from the same health care system and theoretically provide the same range of sexual and reproductive health services, including affordable contraception, they all feature higher risks of negative sexual and reproductive health outcomes, including higher rates of sexually transmitted infections, teenage pregnancies and abortions2. The abortion rates in the overseas French territories are almost twice that of metropolitan France (28.6 vs. 14.7 per 1000 women per year), with substantial regional variation (Table 1)1. In all overseas territories, abortion rates for women less than 18 years old are particularly high (Table 1)1–3. These data raise concerns about the efficiency of the French health care system in providing comprehensive and effective reproductive health services tailored to the needs of the populations living in these territories.
Table 1.
Regional variation in abortion rates in the French territories
| Number of abortions per 1000 women aged 15–49 |
Number of abortions per 1000 women less than 18 years old |
|
|---|---|---|
| Metropolitan France | 14.7 | 12.3 |
| Guadeloupe | 43.5 | 30.0 |
| Martinique | 23.6 | 24.9 |
| Guiana | 38.2 | 34.7 |
| La Réunion | 21.2 | 28.3 |
Despite higher abortion rates, very little attention has been paid to the patterns of contraceptive use and the circumstances leading to contraceptive failures in these high-risk regions, which limits our understanding of the broad regional variations. Since recent comparable data on contraceptive prevalence in the French territories are not available, a first step in analysing regional variations is to compare the use of contraception surrounding an abortion and to explore how the highly medicalised French health care system responds to the specific reproductive health challenges posed in these regions. A particular interest lies in the study of peri-abortion care and specifically post-abortion contraception, as many women in the French overseas territories undergo more than one abortion: 43% of women undergoing an abortion in Guadeloupe and 28% in La Réunion have had a previous abortion4.
Using the first large nationally representative sample of women undergoing an abortion in the French territories in 2007, we explore regional differences in the use of contraception before and after an abortion in the islands of Guadeloupe, in the Caribbean, and La Réunion, in the Indian Ocean. We use this comparison as a means of assessing the ability of the French abortion care services to respond to the specific regional contraceptive needs identified in these territories.
POPULATION AND METHODS
The data for this study are part of a nationally representative survey of 11,403 women undergoing an elective abortion in France between April and September 2007. The sample was selected using a multi-step procedure. First, a random sample of 258 public or private hospitals was selected from the list of all hospitals who provided abortions in 2006 (N = 639). Hospitals were stratified by region and by caseload based on the 2006 hospital statistics. Unequal probabilities of inclusion by caseload were introduced to facilitate data collection. Among the 258 selected hospitals, six were located in Guadeloupe and six in La Réunion. We did not include hospitals from French Guiana (n = 1) and Martinique (n = 3) in this study as they would have provided too few women for the analysis (59 and 93, respectively). All women who underwent an abortion in the 12 facilities located in Guadeloupe and La Réunion or in a physician’s private practice affiliated with these facilities during the six-month study period were invited to participate in the study.
Each woman was assigned a sampling weight that was inversely proportional to the probability of the facility being selected in the sample. The facility-level weights were adjusted for the non-responding facilities according to their geographical location, caseload and public or private status. A further adjustment was introduced to reflect the characteristics of women undergoing an abortion in France (age, abortion technique, and type of facility) based on the national abortion statistics provided by hospital records. All analyses are weighted to take the complex sampling design into account.
Questionnaires
Data were collected at the time of the abortion (the day of the surgical procedure or the day they received mifepristone, for those who had a medical abortion) by means of two questionnaires. The first one, collecting medical information on gestational age, the procedure, contraceptive counselling and prescription, was completed by the healthcare provider who performed the abortion. At the same time, women completed a self-administered questionnaire they returned in a sealed envelope before they left the hospital. They provided information on their socio-demographic background, their contraceptive use at the time of conception and the reasons for contraceptive failure, as well as on contraception counselling and prescription during the course of the abortion care. These two questionnaires were related by a common anonymous identifying number in order to link the medical and socio-demographic information for an individual woman.
Study population
The 12 hospitals contributed a total of 2998 completed questionnaires: 1336 women from Guadeloupe and 1662 from La Réunion. The estimated response rates were 75% in Guadeloupe and 78% in the island of La Réunion. From the initial sample of 2998, we excluded women if they reported their pregnancy was intended (n = 124) or if their pregnancy was terminated for medical reasons (n = 56). We further excluded women for whom the use of contraception at the time of the abortion was unknown (n = 76). Our final study population (N = 2742) comprised 1211 women from Guadeloupe and 1531 from La Réunion. For 2454 of these women information was available from both questionnaires (self-administered and medical) and for 288, we had information from the women’s questionnaire only.
Analysis
We first examined the demographic (age, parity, country of birth) and socio-economic circumstances (income, level of education, health insurance status, cohabitating status, professional situation) associated with women’s use of contraception at the time of conception by study location. The level of non-response on most of the variables was less than 3%. However, 24.6% of cases were missing information about income, 10% about previous live births and 12% about previous abortions. These last two items were available only from the medical questionnaire, which was not completed for 288 women in our study population. Missing values were imputed by fitting regression models using a multiple imputation process5.
Women’s use of contraception was assessed by asking what method of contraception they had used last, if they had stopped using it in the month they conceived and why they thought they had become pregnant. They were also asked specifically if they had used emergency contraception (EC) to try to avoid the current pregnancy.
We compared women’s post-abortion contraceptive prescription by study region and used a hierarchical algorithm, selecting the most effective method if they reported more than one. We further compared women’s and health care professionals’ responses about post-abortion prescription of contraception, among the 2223 women for whom this information was available from both the women and medical questionnaires. However, the wording of the questions differed between questionnaires: women were asked if they were prescribed a contraceptive method while the physicians were asked if they had prescribed or recommended a method.
We complemented this analysis by comparing women’s individual contraceptive paths surrounding the abortion according to their geographical location using two points in time (at the time of conception and the post-abortion prescription). Women’s self-reported use of contraception (based on the women’s questionnaires) was described in three categories: very effective methods (pill, intrauterine device [IUD], implant, injectables, patch, vaginal ring, sterilisation), less effective methods (condom, spermicides, withdrawal, fertility awareness), and no contraception. The term ‘IUD’ referred to both copper-IUDs and levonorgestrel-intrauterine systems.
Finally we investigated factors associated with the prescription of very effective methods after the abortion. Variables with p-values smaller than 0.25 in the univariate analysis were included in the multivariate logistic regression model. Analyses were conducted using Stata software version 10 SE which takes into account the clustering and weighting of the sample. The study received the approval of the Commission Nationale de l’Informatique et des Libertés (CNIL) in Paris which deals with ethical and anonymity issues.
RESULTS
Description of the study population
The sociodemographic characteristics of the women are presented in Table 2. Most women in both settings were in their 20s, were single and had two or more children at the time of the abortion. Forty-five percent of women had not graduated from high school, while only 16.5% had some higher education. Women in La Réunion were younger, and more likely to be unemployed. Women in Guadeloupe were more likely to be poor (less than 600 €/month), to report a history of induced abortion and to be foreign-born.
Table 2.
Socio-demographic characteristics of women having an abortion in the French overseas territories and type of abortion technique and abortion provider
| Women’s socio-demographic characteristics Total |
Guadeloupe % n = 1211 |
La Réunion % n = 1531 |
Total % 2742 |
p |
|---|---|---|---|---|
| Age | ||||
| <18 | 7.5 | 11.8 | 9.8 | <0.0001 |
| 18–19 | 10.9 | 11.4 | 11.2 | |
| 20–24 | 22.7 | 25.6 | 24.2 | |
| 25–29 | 16.8 | 18.7 | 17.8 | |
| 30–34 | 20.4 | 15.2 | 17.6 | |
| 35–39 | 14.0 | 11.7 | 12.8 | |
| 40 and over | 7.8 | 5.6 | 6.7 | |
| Living in a couple | ||||
| Yes | 29.0 | 33.5 | 31.4 | 0.02 |
| No | 71.0 | 66.5 | 68.6 | |
| Number of children ever born | ||||
| None | 30.9 | 34.8 | 32.9 | 0.14 |
| One | 24.8 | 24.9 | 24.9 | |
| Two or more | 44.3 | 40.3 | 42.2 | |
| Previous abortion(s) | ||||
| Yes | 52.0 | 37.4 | 44.3 | <0.0001 |
| No | 48.0 | 62.6 | 55.7 | |
| Place of birth | ||||
| Metropolitan France | 16.4 | 27.5 | 22.3 | <0.0001 |
| French territories (Caribbean/Polynesia/La Réunion) | 68.6 | 66.5 | 67.5 | |
| Foreign born | 15.0 | 6.0 | 10.2 | |
| Level of education | ||||
| Less than high school | 41.8 | 47.7 | 44.9 | 0.14 |
| Professional training high school diploma | 8.7 | 7.6 | 8.1 | |
| Classic high school diploma | 10.3 | 8.7 | 9.5 | |
| ≤2 years after high school | 8.7 | 9.4 | 9.0 | |
| >2 years after high school | 7.1 | 6.0 | 6.5 | |
| Still attending school and <2 years after high school | 23.5 | 20.7 | 22.1 | |
| Income/person living in household | ||||
| Less than 600 €/month | 31.3 | 23.8 | 27.3 | 0.001 |
| 600–1199 €/month | 38.7 | 42.4 | 40.7 | |
| 1200–2399 €/month | 23.1 | 24.3 | 23.7 | |
| 2400 €/month or higher | 6.9 | 9.5 | 8.3 | |
| Health insurance | ||||
| No insurance or government medical aid | 16.3 | 15.6 | 15.9 | 0.84 |
| Social security | 27.6 | 27.3 | 27.4 | |
| Additional private insurance | 56.1 | 57.2 | 56.7 | |
| Professional situation | ||||
| Working | 34.5 | 31.0 | 32.6 | 0.001 |
| Unemployed | 33.9 | 41.6 | 37.9 | |
| Student | 22.9 | 20.0 | 21.4 | |
| Housewife or other | 8.8 | 7.4 | 8.1 | |
| Characteristics of the abortion | % | % | ||
| Abortion technique | ||||
| Medical | 47.4 | 53.0 | 50.4 | 0.007 |
| Surgical | 52.6 | 47.0 | 49.6 | |
| Health care setting | ||||
| Public hospital | 76.6 | 57.5 | 66.5 | <0.0001 |
| Private clinic | 20.6 | 18.2 | 19.4 | |
| Physician’s private practice | 2.5 | 24.3 | 14.2 |
Source: DREES – enquête IVG – 2007 (Data from the French Ministry of Health; the source does not refer to a publication.)
A higher proportion of women received abortion care in public hospitals in Guadeloupe, while women in La Réunion were more likely to have gone to a physician’s private practice for a medical abortion. Medical abortion was more frequent in La Réunion than in Guadeloupe.
Contraceptive situation at the time of conception
Results show differences in women’s use of contraception before the abortion by study location (Table 3). Women in Guadeloupe were more likely not to have used contraception in the month they became pregnant and less likely to report a failure using very effective methods. These differences remained after adjusting for women’s socio-demographic characteristics (data not shown). In both settings, incorrect or inconsistent use accounted for the majority of contraceptive failures: 97% of pill users had missed one or more pills or had temporarily stopped taking the pill during the month they became pregnant and 86% of condom users attributed the pregnancy to a condom slippage or breakage. Twelve percent of women indicated they had used EC to try to avoid the current pregnancy, with no difference by study location (p = 0.17).
Table 3.
Contraceptive use by women undergoing an abortion in the French overseas territories, during the month they became pregnant
| Contraceptive use in the month of conception | Guadeloupe % |
La Réunion % |
Total % |
p* |
|---|---|---|---|---|
| No contraception | 40.4 | 32.2 | 36.1 | 0.0003 |
| Intrauterine device/implant | 1.3 | 1.8 | 1.6 | |
| Patch/vaginal ring | 3.0 | 3.2 | 3.1 | |
| Pill | 21.0 | 27.9 | 24.6 | |
| Condom | 12.9 | 15.6 | 14.3 | |
| Spermicides | 0.2 | 0.3 | 0.2 | |
| Withdrawal | 11.0 | 11.7 | 11.3 | |
| Fertility awareness | 9.0 | 6.3 | 7.6 | |
| Sterilisation | 0.0 | 0.1 | 0.0 | |
| Emergency contraception | 1.3 | 1.0 | 1.2 |
Source: DREES – enquête IVG – 2007. (Data from the French Ministry of Health; the source does not refer to a publication.)
p-value testing the difference in pre-abortion contraception in three categories (no contraception, less effective methods and very effective methods) by study location.
Results show both similarities and differences in the factors associated with women’s pre-abortion use of contraceptive by study location. In both settings, older women were at greater risk of not using contraception in the month they conceived. Women less than 18 years of age were less likely to report a failure using very effective methods and conversely more likely to report a failure using barrier methods (mostly condoms). This was also the case for more highly educated women and those still attending school in La Réunion. Finally, single women in La Réunion were less likely to have used a method of contraception in the month they conceived than other women. We found no difference in use of contraception by income, health insurance status and history of induced abortion.
Contraceptive paths before and after abortion
Half of the women in both settings planned to switch from less effective (including no contraception) to very effective methods after the procedure (Table 5). However, a higher proportion of women in La Réunion remained on very effective methods before and after the abortion (29.5% vs.19.8% in Guadeloupe). The increased odds of having a very effective method before and after the abortion in La Réunion remained significant after controlling for socio-demographic characteristics (Odds Ratio [OR] = 1.7; [95% CI 1.3–2.1], p < 0.0001).
Table 5.
Women’s contraceptive paths before and after the abortion (N = 2,609)
| Contraceptive use at the time of conception | ||||
|---|---|---|---|---|
| No contraception | Less effective contraception |
Very effective contraception |
||
| Post-abortion contraceptive prescription | % | Total ** | ||
| Guadeloupe | ||||
| No contraceptive prescription | 10.3 | 8.0 | 5.8 | 24.1 |
| Less effective contraception | 0.1 | 0.4 | 0.0 | 0.5 |
| Very effective contraception | 29.9 | 25.7 | 19.8 | 75.4 |
| Total* | 40.3 | 34.2 | 25.6 | 100 |
| La Réunion | ||||
| No contraceptive prescription | 4.7 | 4.2 | 3.7 | 12.6 |
| Less effective contraception | 0.2 | 1.0 | 0.3 | 1.5 |
| Very effective contraception | 27.4 | 29.2 | 29.5 | 85.9 |
| Total* | 32.1 | 34.5 | 33.4 | 100 |
Source: DREES – enquête IVG – 2007 (Data from the French Ministry of Health; the source does not refer to a publication.).
Distribution of women’s contraceptive situation at the time the pregnancy leading to the abortion started. Results are slightly different from those of Table 1 as they are restricted to women who also provide information about post-abortion contraceptive prescription.
Distribution of women’s post-abortion contraceptive prescription.
Post-abortion contraceptive prescription
As a result, women’s post-abortion contraceptive situation differed substantially by study location (Table 6). In Guadeloupe, a quarter of the women indicated they had not received a prescription for a contraceptive method in the course of the abortion care, compared with 12.6% of women in La Réunion. This difference remained after adjusting for social and demographic characteristics (OR for not receiving a prescription in Guadeloupe vs. La Réunion = 2.3 [95% CI 1.7–3.1], p < 0.0001). At the same time, only 2.4% of healthcare professionals in Guadeloupe and 4.1% in La Réunion indicated they had not prescribed or recommended a method of contraception (p = 0.09). Differences in the responses of women and healthcare providers were most visible among the 364 women who reported no post-abortion contraceptive prescription and who had a completed medical questionnaire. Healthcare professionals indicated they had recommended or prescribed a method for 96% of these women in Guadeloupe and for 80% of these women in La Réunion (p = 0.02). The 42 providers who reported they had neither issued a prescription nor recommended a method declared they were planning to do so during the next visit (n = 24) or assumed the woman’s regular doctor would provide the prescription (n = 18).
Table 6.
Women’s responses regarding contraceptive prescription after the abortion (N = 2,609)
| Post-abortion contraceptive prescription |
Guadeloupe La Réunion | ||
|---|---|---|---|
| % | p | ||
| No contraception | 24.1 | 12.6 | <0.0001 |
| Intrauterine device | 23.3 | 24.2 | |
| Implant/injections | 4.3 | 7.8 | |
| Patch/ring | 1.6 | 0.6 | |
| Pill | 46.2 | 53.4 | |
| Barrier (including condoms)/natural methods | 0.5 | 1.2 | |
| Emergency contraception | 0.0 | 0.3 | |
Source: DREES – enquête IVG – 2007 (Data from the French Ministry of Health; the source does not refer to a publication.)
According to their responses, half of the women received a prescription for the pill and a quarter received a prescription for an IUD. In some cases, the IUD may have been inserted at the time of the surgical procedure, although the information was not provided in the questionnaire. According to the healthcare professional’s responses however, a majority of providers recommended or prescribed an IUD in Guadeloupe (52.8% vs. 25.6% in La Réunion) while a majority recommended or prescribed the pill in La Réunion (57.5% vs. 36.8% in Guadeloupe).
The determinants of post-abortion contraceptive prescription varied by study location for only one factor: the type of abortion facility (Table 7). The prescription of a very effective method after the abortion varied by women’s health insurance status, although the association did not reach significance in La Réunion (p = 0.07). Women in Guadeloupe who had no health insurance or only a government health plan were 70% less likely to receive a prescription for a very effective method. However, the effect of health insurance status in Guadeloupe was only significant among women who were foreign born. Half of these women (55.8%) had no insurance or relied on government health insurance (as compared to 9.2% of other women in the same sample). Among women who were foreign born, those who had no insurance or a government health plan were three times less likely to receive a prescription for a very effective method compared with those who had a regular health insurance (OR = 3.2 [95% CI 1.6–6.6], p = 0.002). The same was not true for women who were born in the French territories or Metropolitan France (OR = 1.2 [95% CI 0.7–2.0], p = 0.51). These results also indicate that for women who have regular health insurance, we found no difference in post-abortion contraceptive prescriptions by country of birth (OR = 1.2 [95% CI 0.6–2.2], p = 0.59).
Table 7.
Factors associated with receiving a prescription for very effective contraception after abortion
| Guadeloupe | La Réunion | |||||||
|---|---|---|---|---|---|---|---|---|
| % very effective method prescription |
adjusted OR |
95% CI | p* | % very effective method prescription |
adjusted OR |
95% CI | p* | |
| Health insurance | ||||||||
| No insurance or government medical aid | 64.7 | 1 | 0.007 | 81.3 | 1 | 0.07 | ||
| Social security with or without additional private insurance | 77.5 | 1.7 | 1.2–2.5 | 86.8 | 1.6 | 1.0–2.7 | ||
| Health care setting | ||||||||
| Public hospital | 74.3 | 91.8 | 1 | <0.0001 | ||||
| Private hospital | 80.1 | 70.4 | 0.2 | 0.1–0.4 | ||||
| Physician’s private practice | 72.6 | 83.6 | 0.6 | 0.4–1.1 | ||||
Source: DREES – enquête IVG – 2007 (Data from the French Ministry of Health; the source does not refer to a publication.)
The Table only includes variables that remained significant in the multivariate analysis. Variables for which p-values were less than 0.25 in the univariate model were included in the multivariate models (country of origin, income, health insurance, abortion technique and health care setting). We also introduced ‘age’ as a control variable, although the p-value was greater than 0.25 in the univariate analysis for La Réunion. Variables for which p-values were 0.25 or more (living in a couple, number of children ever born, previous abortion, level of education, professional situation) were not included in the multivariate models.
In La Réunion, the prescription of a very effective method also depended on the type of healthcare setting. Women who had their abortion in a private hospital or a physician’s private practice were less likely to receive a prescription for a very effective method than women who had their abortion in a public hospital. Finally, post-abortion prescription of a very effective method was not dependent on parity, income, professional situation, or level of education (data not shown).
DISCUSSION
This study is the first to explore peri-abortion contraceptive practices in a large representative sample of women living in the French overseas territories. Consistent with the conclusions of previous research from industrialised countries, including metropolitan France6–8, most abortions in these overseas territories follow a contraceptive failure. Nearly a third of the women described a failure using very effective methods (mostly missed pills) and another third using barrier or natural methods. The comparison between French overseas territories, however, shows wide regional disparities in pre-abortion use of contraception, revealing a higher-risk profile among women presenting for an abortion in Guadeloupe.
These regional variations, which were not explained by socio-demographic differentials, may reflect overall differences in contraceptive prevalence, which ultimately could account for the disparities in abortion rates observed between the two regions. Indeed, as several authors point out, contraceptive use is a key factor in explaining differentials in abortion rates, especially in territories presenting with similar total fertility rates (the TFR in La Réunion and Guadeloupe is 2.49,10) and governed by the same laws and policies regarding abortion11. In this context, variation in abortion rates may reveal inter-regional differences in the quality of contraceptive services and sex education or their inability to respond to the specific needs of the populations they serve12.
Our results suggest that the French abortion care system seems unable to reduce contraceptive disparities as there were substantial differences found in post-abortion contraception prescription by study location. In both cases the abortion process was associated with an increase in very effective methods prescribed and consequently also in the theoretical effectiveness of contraceptive coverage. This increase, also reported in the study conducted in metropolitan France13, as well as in other large studies of women undergoing an abortion in the United Kingdom, Switzerland and the United States7,8,14, was evident in both Guadeloupe and La Réunion. The same proportion of women in Guadeloupe and la Réunion who were not using contraception or using less effective methods were prescribed a more effective method after the procedure. As a result, the preexisting regional gap in the use of contraception persisted after the abortion procedure. A quarter of women in Guadeloupe and 12.6% in La Réunion reported receiving no prescription for a contraceptive method after the procedure. For a small fraction of these women, healthcare providers indicated they had not recommended or prescribed a contraceptive method as they intended to do so during the next visit or assumed the woman’s regular doctor would provide contraception. The difference between women and healthcare providers’ responses reflects response errors or a gap between counselling and prescribing; in either case the high proportion of women who considered having no prescription for a very effective method after the abortion is alarming, especially in Guadeloupe. This is particularly true as some women may not use the method they were prescribed. We were unable to assess actual use of contraception after the abortion, because of the high refusal rate for the follow-up interview (75%), tested in the pilot study. Post-abortion contraception prescription imperfectly reflects actual use of contraception but is a prerequisite for the use of very effective methods and therefore an important factor in explaining women’s post-abortion contraceptive practices.
In both settings, health insurance status was associated with women’s post-abortion contraceptive prescription. Women with the lowest level of health insurance were less likely to receive a prescription for very effective methods, a finding which brings into question the reality of universal access to the most effective methods in these territories. This association may either reflect a cost issue or a specific issue for migrant women, or both. Cost is unlikely to be a barrier to contraceptive prescription for most French women; indeed, 65% or more of the expense for most prescribed methods is reimbursed by the national health insurance plan. The only exceptions are the patch, the ring and third generation pills. For non-resident women the cost of contraception and the availability of the methods in their own country may be a barrier to prescription. This seems particularly true for foreign born women in Guadeloupe; it draws attention to specific issues related to migrant women in the Caribbean where Guadeloupe is one of the few territories to provide unrestricted access to safe first trimester abortions on request. A qualitative study conducted across five countries in the Caribbean indicates women frequently travel across borders to escape stigma and punishment in a context of restrictive access to abortion in the region15. This migration process raises specific concerns about obstacles to optimal care, which are also evident in the present study. Unfortunately, information about women’s place of residence was not available, which does not allow the distinction between women who were born abroad and who resided legally, illegally or who did not reside in the French territories. However, given these women underwent an abortion in the French health care system, they should have been provided the same contraceptive options as others, if contraception is considered part of abortion care.
Thus, our results indicate the need for a careful assessment of the availability and accessibility of contraceptives for all women regardless of their social circumstances and geographical origin. They also point out to differences in contraceptive prescriptions by type of health care facility. Indeed, women from La Réunion who had their abortion in the private sector were less likely to receive a prescription for a highly effective method, revealing heterogeneity in the French health system’s management of post-abortion contraceptive prescription.
CONCLUSION
While this study shows a significant increase in contraceptive uptake after an abortion, it also indicates the ineffectiveness of the abortion care services in closing the gap in the pre-existing contraceptive disparities observed between the French territories of Guadeloupe and La Réunion. In other words, despite providing equal quality of care for abortions for individuals with different needs, the French health care system fails to promote health equity and to reduce regional sexual and reproductive health inequalities. In both settings, the lack of regular health insurance was a barrier to receiving a prescription for very effective methods, which questions the reality of universal access to the most effective methods in these territories. These results should be brought to the attention of policy makers and health care professionals so they can reflect on new strategies to address the specific contraceptive needs of women in these territories.
Table 4.
Socio-demographic factors associated with women’s contraceptive situation at the time of conception
| Guadeloupe | La Réunion | ||||||||
|---|---|---|---|---|---|---|---|---|---|
| No method |
Barrier or natural |
Very effective method |
No method |
Barrier or natural |
Very effective method |
||||
| Age | % | p* | % | p* | p** interaction | ||||
| Total | 40.4 | 34.3 | 25.3 | 32.2 | 34.8 | 33.0 | |||
| <20 | 35.0 | 54.3 | 10.8 | 0.05 | 33.9 | 39.1 | 27.0 | 0.07 | 0.002 |
| 20–24 | 40.0 | 31.7 | 28.4 | 31.7 | 29.5 | 38.7 | |||
| 25–29 | 42.5 | 29.1 | 28.4 | 32.0 | 32.1 | 35.9 | |||
| 30–34 | 38.4 | 28.2 | 33.5 | 27.7 | 41.3 | 31.0 | |||
| 35–39 | 42.7 | 30.4 | 26.9 | 32.6 | 36.8 | 30.6 | |||
| 40 and over | 50.6 | 29.6 | 19.8 | 38.5 | 28.0 | 33.5 | |||
| Living in a couple | |||||||||
| Yes | 37.6 | 29.9 | 32.4 | 0.08 | 24.4 | 36.4 | 39.1 | 0.001 | 0.05 |
| No | 41.5 | 36.2 | 22.4 | 36.0 | 34.1 | 29.9 | |||
| Children ever born | |||||||||
| Yes | 41.9 | 28.3 | 29.8 | 0.12 | 30.2 | 32.3 | 37.5 | 0.02 | 0.02 |
| No | 37.0 | 47.8 | 15.2 | 35.9 | 39.7 | 24.4 | |||
| Health insurance | |||||||||
| No insurance or government insurance | 45.1 | 29.4 | 25.5 | 32.9 | 31.7 | 35.4 | 0.17 | 0.04 | |
| Social security | 37.5 | 37.8 | 24.7 | 37.5 | 28.9 | 33.6 | |||
| Additional private insurance | 40.4 | 34.1 | 25.6 | 29.4 | 38.5 | 32.1 | |||
| Professional situation | |||||||||
| Working | 43.8 | 33.3 | 22.9 | 0.04 | 28.7 | 40.1 | 31.3 | 0.17 | |
| Unemployed | 39.7 | 26.7 | 33.6 | 34.1 | 26.1 | 39.8 | |||
| Student | 37.8 | 49.4 | 12.7 | 32.2 | 45.9 | 22.0 | |||
| Housewife or other | 36.5 | 28.5 | 35.0 | 35.6 | 32.7 | 31.8 | |||
Source: DREES – enquête IVG – 2007 (Data from the French Ministry of Health; the source does not refer to a publication.)
p-values from a polytomous multivariate regression model exploring the associations between women’s characteristics and the type of contraception used in the month they conceived (defined as a three-category variable: no contraception / less effective methods / very effective methods). Only variables with a p-value < 0.25 in the univariate analysis were included in the multivariate analysis.
p-values for the test of interaction exploring differences in effects of women’s social and demographic characteristics on their pre-abortion contraceptive profile by study location.
ACKNOWLEDGEMENTS
The national abortion survey was conducted by the ‘Direction de la recherche, des études, de l’évaluation et des statistiques’ of the French Ministry of Health and the statistical analysis was funded by a grant from the ‘Departement Général de la Santé’ of the Ministry of Health in France.
Footnotes
Declaration of interest: The authors report no conflicts of interest. The authors alone are responsible for the content and the writing of the paper.
REFERENCES
- 1.Vilain A. Les interruptions volontaires de grossesse en 2007. [Accessed 4 February 2010];Etudes et Résultats. 2009 n°713. from: http://www.sante.gouv.fr/drees/etude-resultat/er-pdf/er713.pdf.
- 2.Clair M. Interruptions volontaires de grossesses en Guadeloupe, Qui? Où? Comment? Info Santé. 2004;13:13–18. [Google Scholar]
- 3.Observatoire Régional de la Santé de la Réunion. Contraception et interruptions volontaires de grossesses à la Réunion. [Accessed 4 February 2010];2008 from: http://www.urml-reunion.net/chikungunya/2008/TDB_IVG-contraception2008.pdf. [Google Scholar]
- 4.Le Corre M, Thomson E. Les IVG en 1998. Etudes et résultats. 2000;69:1–4. [Google Scholar]
- 5.Carlin JB, Galati JC, Royston P. A new framework for managing and analyzing multiply imputed data in Stata. Stata Journal. 2008;8:49–67. [Google Scholar]
- 6.Bajos N, Leridon H, Goulard H, et al. Contraception: From accessibility to efficiency. Hum Reprod. 2003;18:994–999. doi: 10.1093/humrep/deg215. [DOI] [PubMed] [Google Scholar]
- 7.Jones RK, Darroch JE, Henshaw SK. Contraceptive use among US women having abortions in 2000–2001. Perspect Sex Reprod Health. 2002;34:294–303. [PubMed] [Google Scholar]
- 8.Schunmann C, Glasier A. Specialist contraceptive counselling and provision after termination of pregnancy improves uptake of long-acting methods but does not prevent repeat abortion: A randomized trial. Hum Reprod. 2006;21:2296–2303. doi: 10.1093/humrep/del168. [DOI] [PubMed] [Google Scholar]
- 9.INSEE Antilles Guyane. Premiers résultats. Bilan démographique de la Guadeloupe en 2006. [Accessed 4 February 2010];2007 Dec;(no. 23) from: http://www.insee.fr/fr/insee_regions/guadeloupe/publi/PR_BDMO06_ga.html. [Google Scholar]
- 10.INSEE Réunion. Economie de la Réunion n°131. [Accessed 4 February 2010];2007 Dec; from: http://www.insee.fr/fr/insee_regions/reunion/prodser/pub_elec/revue/revue131/revue131_demographie.html. [Google Scholar]
- 11.Marston C, Cleland J. Relationships between contraception and abortion: A review of the evidence. Int Fam Plann Perspect. 2003;29:6–13. doi: 10.1363/ifpp.29.006.03. http://www.agi-usa.org/ pubs/journals/2900603.html.-_blank. [DOI] [PubMed] [Google Scholar]
- 12.Ketting E. Global overview of abortion. Planned Parenthood Challenges. 1993;1:27–29. [PubMed] [Google Scholar]
- 13.Bajos N, Lamarche Vadel A, Gilbert F, et al. Contraception at the time of abortion: High-risk period or high-risk women? Human Reprod. 2006;21:2862–2867. doi: 10.1093/humrep/del268. [DOI] [PubMed] [Google Scholar]
- 14.Bianchi-Demicheli F, Perrin E, Bianchi PG, et al. Contraceptive practice before and after termination of pregnancy: A prospective study. Contraception. 2003;67:107–113. doi: 10.1016/s0010-7824(02)00475-4. [DOI] [PubMed] [Google Scholar]
- 15.Pheterson G, Azize Y. Abortion practice in the Northeast Caribbean: ‘just write down stomach pain’. Reprod Health Matters. 2005;13:44–53. doi: 10.1016/S0968-8080(05)26201-8. [DOI] [PubMed] [Google Scholar]