The introduction of medical termination of pregnancy by means of the use of mifepristone and prostaglandins has changed abortion practices dramatically in several countries. In 2006, 69% of all abortions in Finland,1 56% in Sweden,2 and 30% in the United Kingdom3 were performed using the medical method. In 2005 the corresponding figure in the United States was 13%.4 Despite these changes, overall induced abortion rates changed only minimally in several European countries between 1996 and 2003, the period during which many European countries brought medical termination of pregnancy into general use. During the corresponding period, the induced abortion rate in the United States declined by 8%.5
A systematic review has shown medical termination of pregnancy to be safe and effective.6 In a previous epidemiologic study, no difference in the incidences of spontaneous abortion, ectopic pregnancy, preterm delivery, or low birth weight was found after medical rather than surgical termination of pregnancy.7
Because medical termination of pregnancy is well-tolerated among women,8–10 the question of whether it is associated with an increased risk of repeat abortion has been raised. Several studies have assessed the characteristics of women seeking repeat abortion.11–13 Women seeking repeat termination of pregnancy tend to be older than those having their first termination of pregnancy, because older women have had a longer period of exposure to the risk of having an unintended pregnancy.14 Age, however, is thought to be more of a confounder than a true risk factor.13 In addition, parity has emerged as a risk factor for repeat abortion, although, like age, it is more of a confounding factor.11 Histories of physical or sexual abuse12 and alcohol and drug abuse13 have also been associated with repeat abortion in cross-sectional studies. In a recent prospective study, young age, parity, a history of previous abortion, and smoking emerged as risk factors for repeat abortion after medical termination of pregnancy.15
Contraceptive choices of women seeking repeat abortion differ from those of first-time aborters; women seeking repeat abortions have been reported more often to have used some method of contraception,16 such as oral contraceptives12 or depot medroxyprogesterone acetate.13 However, in a Danish study, women seeking a third abortion tended to use less efficient contraceptive methods (and none at all) than women undergoing their first or second termination of pregnancy.17
The role of the abortion method on the overall abortion rate and the risk of repeat abortion has not been assessed previously. The purpose of the present study was to compare the frequency of and risk factors for repeat abortions after surgical compared with medical termination of pregnancy using large population-based national health registries with high-quality data. Identification of women who are at an increased risk of repeat termination of pregnancy would be of major public health importance, because it may be used in preventive health care interventions to reduce the risk of repeat abortion.
MATERIALS AND METHODS
The setting was a population-based cohort study. The cohort consisted of 40,360 women with induced abortion in Finland between January 1, 2000, and December 31, 2006. The inclusion criteria were index abortion (the first abortion in this cohort) occurring between January 1, 2000, and December 31, 2005, duration of gestation 63 days or less, and no simultaneous sterilization at the time of index abortion. The duration of gestation in index abortions was limited to 9 weeks (63 days), because medical abortion is used mostly in pregnancies up to that time.18 Medical abortion using mifepristone and prostaglandin was introduced in Finland in 2000, and since then the two methods (ie, surgical and medical) have been offered as alternatives in all hospitals performing termination of pregnancy. The subjects were followed up until the first repeat abortion, or until December 31, 2006. Thus, the minimum follow-up time was at least 1 year. To be considered as a repeat abortion, the interval between the index and subsequent abortion had to be 30 days or more. The duration of gestation was not limited in repeat abortions. Deaths (n=102) in the cohort were taken into account in calculation of follow-up time. The data in the Abortion Registry was linked with that in the Cause-of-Death Registry of Statistics Finland to determine the time of follow-up for the subjects who died during the follow-up period in this cohort. Data linkage within the Abortion Registry and between the two data sources was achieved by using the women's unique personal identification numbers. Before the analysis, personal identification numbers were removed from the data.
The method of induced abortion was divided into two categories. The medical method included induced abortions performed with mifepristone alone, or with a combination of mifepristone and misoprostol or other prostaglandins. The surgical method included induced abortions performed using either dilation and curettage or vacuum aspiration. The subjects were divided into two study cohorts (medical or surgical termination of pregnancy) based on the method used in the index abortion.
Social status was defined by using the stated occupation or the highest educational level found in the Abortion Register. Coding was based on national standards published by Statistics Finland.19–22 The groups were divided into five categories: upper white-collar workers, lower white-collar workers, blue-collar workers, students (level of education not defined), and others (all other groups, such as entrepreneurs, farmers, unemployed, retired, housewives or missing data). The type of residence was defined by the data on the municipality of residence, available in the Abortion Register. The municipalities were divided into three categories: urban, densely populated, and rural areas, according to national standards.19–21
In Finland, according to the current Act on Induced Abortions, termination of pregnancy can be allowed up to 20 weeks of gestation (24 weeks in cases of a medical condition of the fetus) for social, medical, or ethical reasons. Social reasons include considerable strain on living or other conditions, age below 17 or above 40 years at conception, or women who have given birth to at least four children. Medical reasons include mental deficiency, severe illness or handicap of the fetus, pregnancy being a risk to the woman's life or health, woman's sickness, physical defect or infirmity, or the mother/father not being able to take care of the child. Ethical reasons are rape, incest, or other reason mentioned in the Penal Law.
A National Registry on induced abortions and sterilizations has been maintained since 1977 by the National Research and Development Centre for Welfare and Health (the STAKES database).1 In accordance with the current legislation, the physician performing termination of pregnancy is obliged to report the case to the National Registry within 1 month, using a specific data collection form. Data on induced abortions are collected from all hospitals and clinics in which induced abortions are performed in Finland.
The Finnish Abortion Registry contains data on women having induced abortions. The data include background information such as pregnancy history, occupation, residence, municipality, and marital status. Data concerning the current pregnancy, such as the method of contraception used before the pregnancy, gestational duration at termination of pregnancy, indication for abortion, the method of termination of pregnancy, and planned future contraception are registered.
The present study was conducted after approval from the institutional review board (Ethics Committee) of the Northern Ostrobothnia Hospital District. In addition, Ministry of Social Affairs and Health and Statistics Finland gave their permissions to use the confidential person-level data in the nationwide health registers in this study. Legal approval from the Data Protection Ombudsman was obtained before the analyses, as required by the national data protection legislation.
Differences between the groups were assessed by using Student t tests, the χ2 test, and the two-sample test of proportions. Kaplan–Meier curves were constructed for the medical and surgical groups and compared by using the log rank test. Cox regression analyses were performed to identify the risk factors for repeat abortions, and the estimated risks are presented as hazard ratios (HRs) with 95% confidence intervals (95% CIs). Variables that showed a statistically significant association with repeat abortion in univariable analysis were further entered in multivariable analysis. The statistical analyses were performed by using SPSS 15.0 for Windows (SPSS Inc., Chicago, IL) and Stata 5.0 (StataCorp LP, College Station, TX) statistical software.
Figure 1 shows the proportions (%) of medical and surgical abortions between 2000 and 2005 in the study cohort. In addition, the general annual abortion rate (per 1,000 women aged 15–49 years) in Finland is shown.1 The use of medical abortion increased from 8.5% in 2000 to 73.9% in 2005. The annual abortion rate varied between 8.9 and 9.4/1,000 women aged 15–49 years.
The total number of women fulfilling the inclusion criteria in 2000–2005 was 40,360. Of these, 19,841 had medical abortion and 20,519 had surgical abortion. The follow-up time (mean±standard deviation) for the cohort undergoing medical abortion was 3.0±1.5 years, and it was 4.3±1.9 years for women undergoing surgical abortion (P<.001). The overall mean follow-up time for the whole cohort was 3.7±1.8 years.
The sociodemographic characteristics of the two cohorts are shown in Table 1. When comparing the two cohorts, the groups differed statistically in several respects. Notably, the duration (mean±standard deviation) of gestation was shorter at the time of abortion among the women in the medical compared with the surgical group (7.2±1.1 compared with 7.9±1.0 weeks, P<.001). More subjects in the medical group were of an unknown social status (37%) compared with those in the surgical group (22%).
A subgroup analysis was performed among primigravid women. The primigravid women were younger (21.3±5.3 compared with 26.6±7.6 years, P<.001), more often single (84.9% compared with 63.1%, P<.001), and more often students (46.4% compared with 26.8%, P<.001) when compared with the entire cohort (data not shown).
Overall, 14.2% of the subjects requested a repeat abortion within the follow-up time. Table 2 shows the proportion of women with repeat abortion per follow-up year in relation to the index abortion. The incidence of repeat abortion per follow-up year per 1,000 women differed significantly only for the year 2001 (34.9/1,000 in the medical group compared with 39.5/1,000 in the surgical group; P=.04). In addition, the total numbers of repeat abortions per follow-up year per 1,000 women differed between the two cohorts, being 40.4/1,000 in the medical group and 37.9/1,000 in the surgical group (P=.01).
The time interval to repeat abortion in relation to the index abortion was recorded. The time intervals were similar in the two cohorts. For example, the mean times to repeat abortion were 32.1 and 32.3 months among women undergoing medical and surgical abortion, respectively, in 2000. In 2005, the corresponding time intervals were 10.5 and 10.0 months. In Kaplan–Meier analysis, the cumulative risk of repeat abortions in the medical and surgical groups in relation to the index abortion did not differ statistically significantly (data not shown).
To define the risk factors for repeat abortions, univariable and multivariable analyses were performed for the medical and surgical groups and for both groups together (Table 3, Fig. 2). The risk factors for repeat abortion were largely similar after both medical and surgical abortion. In multivariable analyses, advancing age was associated with decreasing risk, and previous abortion(s) was associated with an increased risk of repeat abortion in both groups. Moreover, being single or parous increased the risk of repeat abortion in the medical cohort. In the surgical abortion cohort, the risk was also increased among women belonging to the social class “others.” When compared with the urban environment, living in a rural area was associated with a decreased risk of repeat abortion among women undergoing medical abortion.
When compared with planned use of combined oral contraceptives for postabortion contraception, the risk of repeat abortion in multivariable analysis was decreased if sterilization was planned for future contraception in both cohorts. Figure 2 shows the hazard ratios (and 95% CIs) for repeat abortion derived from multivariable analysis of both groups combined. The risk of repeat abortion increased with parity, previous abortion(s), having the socioeconomic status “others,” or cohabiting or being single (compared with married women). However, the risk of repeat abortion decreased with advancing age and was lower among those living in a densely populated or rural area (compared with an urban environment) and when the levonorgestrel-releasing intrauterine system, a copper-releasing intrauterine device or sterilization was planned for future contraception (compared with planned use of combined oral contraceptives). As derived from univariable analysis, the method of the index abortion (medical compared with surgical) was not associated with an altered risk of repeat abortion (HR 0.98, 95% CI 0.93–1.04).
When primigravid women were analyzed separately, advancing age was associated with a decreased risk of repeat abortion. Method of the index abortion, social or marital status, type of residence, or planned method of contraception did not have an effect on the risk of repeat abortion (data not shown). For example, when compared with the age group of younger than 20 years, the HRs (and 95% CIs), derived from multivariable analyses were 0.44 (0.35–0.56) and 0.43 (0.34–0.54) among women aged 25–29 years undergoing medical and surgical abortion, respectively. Among primigravid women undergoing medical abortion, the risk of repeat abortion was increased (1.74, 1.09–2.78) among blue-collar workers. Of the other variables analyzed, marital status, type of residence, gestational duration, and planned contraception did not have an effect on the risk of repeat abortion in either cohort. When both cohorts of primigravid women were combined, only older age was associated with a significantly altered risk of another abortion.
In the present study, the risk factors for repeat abortion were mostly identical after both medical and surgical abortion. Importantly, we found the risk of repeat abortion not to be affected by the method of abortion.
The strength of this study is the large cohort, covering practically all abortions performed in Finland during the study period. In registry-based studies, coverage and quality of the data are of utmost importance. The validity of the information in the Finnish Abortion Registry has been studied previously by comparing consecutive samples of information with the medical records in the hospitals. The completeness of the registry was found to be reliable; 95% of the information in the registry was identical to that in the medical records.23 Moreover, it was recently shown that 99% of abortions are included in the registry.15 Thus, we find the external validity of this study to be high.
In Finland, medical termination of pregnancy has become the dominant method of abortion, whereas the use of surgical methods diminished during the study period. The follow-up time was significantly shorter for subjects who underwent medical termination of pregnancy, hence the proportion of women undergoing repeat abortion was calculated per thousand women per follow-up year. The time of follow-up was also taken into account in the univariable and multivariable analyses. An important question is, if the women choosing medical methods are different from those choosing surgical methods, would the study be biased? Women undergoing medical and surgical termination of pregnancy differed subtly, but statistically significantly, in several respects. It is likely that these subtle differences (such as younger age) explain the somewhat higher rate of abortion incidence per 1,000 women-years among the cohort undergoing medical abortion. However, the difference between the two methods disappeared in the Kaplan-Meier and univariable analyses. Clinically meaningful differences between the two cohorts emerged as regards the duration of gestation at the time of termination of pregnancy. Because surgical termination of pregnancy is preferably performed after the sixth week of gestation,18 this difference reflects more differences in clinical practice than in women seeking abortion.
The risk factors for repeat abortion were similar to those described in previous studies: young age, parity, and history of induced abortion(s).13,15 Among primigravid women, advancing age was the only factor associated with a significantly altered risk of repeat abortion. In a previous study based on self-reporting, lower socioeconomic status, a low level of education, and the relationship status correlated with an increased risk of repeat abortion.12 Similarly, in the present study, being a blue-collar worker or classified in the group of low socioeconomic status (such as unemployed) was associated with an increased risk of repeat abortion. Being single or cohabiting also emerged as risk factors. As regards the type of residence, the risk was diminished among subjects living in rural or densely populated areas when compared with women living in an urban environment.
The effect of contraceptive choice on the risk of another abortion is less clear. In a recent randomized study, specialist contraceptive counseling and provision after termination of pregnancy improved the rate of initial uptake of long-acting methods but did not have an effect on the rate of repeat abortion.24 In a previous study, we found that verified use of intrauterine contraception after medical termination of pregnancy was associated with a lower risk of repeat abortion.15 Similarly, in the present study, planned use of intrauterine contraception or sterilization was associated with a lower risk of repeat abortion. However, the present study design allowed us to assess only planned, but not verified, use of postabortion contraception.
The risk of repeat abortion is explained mainly by various sociodemographic characteristics. Young women, especially those with a low socioeconomic and/or an unstable relationship status, are an important target group as regards contraceptive counseling, as are women with a history of previous deliveries and/or induced abortions. However, the method of abortion used is not associated with an altered risk of repeat abortion.
4. Jones RK, Zolna MR, Henshaw SK, Finer LB. Abortion in the United States: incidence and access to services, 2005. Perspect Sex Reprod Health 2008;40:6–16.
5. Sedgh G, Henshaw SK, Singh S, Bankole A, Drescher J. Legal abortion worldwide: incidence and recent trends. Int Fam Plan Perspect 2007;33:106–16.
6. Kulier R, Gulmezoglu AM, Hofmeyr GJ, Cheng LN, Campana A. Medical methods for first trimester abortion. Cochrane Database of Systematic Reviews 2004, Issue 1. Art. No.: CD002855. DOI: 10.1002/14651858.CD002855.pub3.
7. Virk J, Zhang J, Olsen J. Medical abortion and the risk of subsequent adverse pregnancy outcomes. N Engl J Med 2007;357:648–53.
8. Ho PC. Women's perceptions on medical abortion. Contraception 2006;74:11–5.
9. Honkanen H, von Hertzen H. Users' perspectives on medical abortion in Finland. Contraception 2002;65:419–23.
10. Jensen JT, Harvey SM, Beckman LJ. Acceptability of suction curettage and mifepristone abortion in the United States: a prospective comparison study. Am J Obstet Gynecol 2000;182:1292–9.
11. St John H, Critchley H, Glasier A. Can we identify women at risk of more than one termination of pregnancy? Contraception 2005;71:31–4.
12. Fisher WA, Singh SS, Shuper PA, Carey M, Otchet F, MacLean-Brine D, et al. Characteristics of women undergoing repeat induced abortion. CMAJ 2005;172:637–41.
13. Prager SW, Steinauer JE, Foster DG, Darney PD, Drey EA. Risk factors for repeat elective abortion. Am J Obstet Gynecol 2007;197:575.e1–6.
14. Jones RK, Singh S, Finer LB, Frohwirth LF. Repeat abortion in the United States. Occasional Report, No. 29. New York (NY): Guttmacher Institute; 2006.
15. Heikinheimo O, Gissler M, Suhonen S. Age, parity, history of abortion and contraceptive choices affect the risk of repeat abortion. Contraception 2008;78:149–54.
16. Westfall JM, Kallail KJ. Repeat abortion and use of primary care health services. Fam Plann Perspect 1995;27:162–5.
17. Osler M, David HP, Morgall JM. Multiple induced abortions: Danish experience. Patient Educ Couns 1997;31:83–9.
18. Finnish Gynecological Association's task force. Abortion [in Finish]. Duodecim 2001;117:2084–94.
19. Statistics Finland. Classification of occupations. Handbook 14. Helsinki (Finland): Statistics Finland; 1987.
20. Statistics Finland. Classification of socioeconomic groups. Handbook 17. Helsinki (Finland): Statistics Finland; 1989.
21. Koulutusluokitus 31.12.1991 (Classification of education. 31 December 1991 [in Finnish]): Handbook 1. Helsinki (Finland): Statistics Finland; 1992.
22. Gissler M, Meriläinen J, Vuori E, Hemminki E. Register based monitoring shows decreasing socioeconomic differences in Finnish perinatal health. J Epidemiol Community Health 2003;57:433–9.
23. Gissler M, Ulander VM, Hemminki E, Rasimus A. Declining induced abortion rate in Finland: data quality of the Finnish abortion register. Int J Epidemiol 1996;25:376–80.
24. Schunmann C, Glasier A. Specialist contraceptive counseling and provision after termination of pregnancy improves uptake of long-acting methods but does not prevent repeat abortion: a randomized trial. Hum Reprod 2006;21:2296–303.
© 2009 by The American College of Obstetricians and Gynecologists.