Abstract
This study aims to describe factors associated with the number of past abortions obtained by New York City (NYC) abortion patients in 2010. We calculated rates of first and repeat abortion by age, race/ethnicity, and neighborhood-level poverty and the mean number of self-reported past abortions by age, race/ethnicity, neighborhood-level poverty, number of living children, education, payment method, marital status, and nativity. We used negative binomial regression to predict number of past abortions by patient characteristics. Of the 76,614 abortions reported for NYC residents in 2010, 57 % were repeat abortions. Repeat abortions comprised >50 % of total abortions among the majority of sociodemographic groups we examined. Overall, mean number of past abortions was 1.3. Mean number of past abortions was higher for women aged 30–34 years (1.77), women with ≥5 children (2.50), and black non-Hispanic women (1.52). After multivariable regression, age, race/ethnicity, and number of children were the strongest predictors of number of past abortions. This analysis demonstrates that, although socioeconomic disparities exist, all abortion patients are at high risk for repeat unintended pregnancy and abortion.
Keywords: Abortion, Induced, Pregnancy outcome, Pregnancy, Unplanned
Introduction
Approximately 95 % of elective abortions (hereafter “abortion”) in the USA result from unintended pregnancies,1 and half of abortion patients have had one or more prior abortions.2 Repeat abortions can be viewed as a proxy for repeat unintended pregnancies and may indicate persistent challenges accessing health information and services or using contraception consistently and correctly. Understanding characteristics of women having repeat unintended pregnancies ending in abortions can help determine whether certain groups have greater or unique needs for postabortion contraception services.
The most recent US population-based study of repeat abortions used data collected in 2000.2 More recent US studies are based on small, nonrepresentative samples from clinical settings.3,4 This analysis adds to the literature on repeat abortion by using 2010 data from the New York City (NYC) abortion reporting system. We compared rates of first and repeat abortions by sociodemographic group. Additionally, we calculated the proportion of 2010 reported abortions that were repeat procedures by sociodemographic characteristics and used multivariable regression modeling to relate number of past abortions to patient characteristics among 2010 abortion patients.
Material and Methods
We used data from NYC’s Department of Health and Mental Hygiene (DOHMH) abortion reporting system. NYC’s health code requires reporting of all abortions performed in NYC to DOHMH’s Bureau of Vital Statistics. NYC is a separate vital statistics reporting jurisdiction from New York state.5 The health code defines an abortion as “a purposeful interruption of an intrauterine pregnancy with the intention other than to produce a live-born infant and which does not result in a live birth.”6 The abortion report form is completed by health care providers or their designees and includes individual patient demographic and socioeconomic data, and self-reported past pregnancy outcomes, including live births, miscarriages/stillbirths, and abortions. No identifying data are collected other than patient initials, ZIP code, and birthdate. The reason for having the abortion is not reported. In comparison to estimates from the Guttmacher Institute’s Abortion Provider Census, NYC’s abortion reporting system captures approximately 88 % of abortions performed in NYC.5 On average, approximately 90,000 abortions/year (range 83,730–94,466) were reported during 2000–2010.7
DOHMH received reports of 83,730 abortions occurring during 2010; 76,614 of these were performed for NYC residents. We limited our study population to NYC residents to allow us to calculate rates using NYC population data and assign a patient neighborhood income level to each procedure using census data. The single-year restriction was made to minimize inclusion of successive abortions for a single patient. We defined repeat abortion as a procedure for which the patient reported having ≥1 past abortion. Otherwise, the procedure was defined as a first abortion. Both medication and surgical abortion procedures were included. We calculated rates of first and repeat abortions/1000 women aged 15–44 years in NYC by variables for which 2010 census denominator data were available (i.e., age, race/ethnicity, and neighborhood-level poverty). Neighborhood-level poverty was defined as the percentage of the population in a ZIP code living below the federal poverty threshold. Analysis of health outcomes by neighborhood-level poverty is a DOHMH standard practice and is described elsewhere.8 ZIP code was used as the analysis unit because it is the smallest geographic unit reported for abortion data. Each abortion procedure was assigned to one of four poverty categories, as follows: very low (<10 % of population living below federal poverty threshold), low (10 to <20 %), medium (20 to <30 %), and high (≥30 %). Additionally, we calculated the number and proportion of abortions that were repeat abortion procedures by age, race/ethnicity, neighborhood-level poverty, number of children, education, payment method, nativity, and marital status. We restricted analysis of education to women aged ≥25.
We calculated mean number of self-reported past abortions by each covariate and the unadjusted relative increase in past abortions for each level of each covariate in comparison with reference values. Reference categories were as follows: aged 25–29 years, high poverty, black non-Hispanic, one child, high school graduate, Medicaid payment, US-born, and unmarried. We developed a negative binomial regression model, adjusting for all covariates, with outcome defined as number of past abortions. We used the model to estimate mean number of previous abortions for women with each covariate pattern. We then calculated the adjusted ratio of mean number of past abortions for each level of each covariate in comparison with the reference value for that covariate.
The education covariate had 21 % missing values. All other covariates had from <1 to 5 % missing values. The outcome—number of past abortions—was missing for 2267 observations (3 %). For the descriptive analysis, procedures with missing data for the covariates of interest were removed from the analysis. For the multivariable analysis, missing values for all potential predictors and the outcome were imputed by using multiple imputation techniques9 with IVEware software (University of Michigan, Ann Arbor, Michigan). Five imputations were performed, and each imputed data set was analyzed by using SAS PROC GENMOD. Resulting estimates and standard errors were pooled by using PROC MIANALYZE. Because of the high proportion of missing values for education, a sensitivity analysis was conducted omitting education. Since resulting estimates were within 3 % of estimates obtained with education included, only results with the full model (i.e., including education) are presented. All statistical analyses were conducted by using SAS (version 9.2, SAS Institute, Inc., Cary, North Carolina). This study underwent human subjects review at the Centers for Disease Control and Prevention and was determined to be a nonresearch, public health surveillance activity.
Results
In 2010, repeat abortions comprised 57 % of the 74,347 abortions for which data on number of previous abortions were available. This proportion in NYC has ranged from 54 to 59 % during the previous 10 years (Table 1). Among 2010 abortion patients, the percentage of abortions that were repeat abortions increased with age, number of children, and neighborhood-level poverty (Table 2). A higher proportion of abortions were repeat procedures among Puerto Rican and black non-Hispanic women, compared with other racial/ethnic groups. The proportion of repeat procedures was also higher among US-born women, women with a high school education, and procedures paid for by Medicaid or other third-party insurance.
TABLE 1.
Number and proportion of reported abortions that were first and repeat abortions, New York City, 2001–2010
| Year | First abortion | Repeat abortion |
|---|---|---|
| 2001 | 37,280 (41 %) | 52,898 (59 %) |
| 2002 | 38,332 (42 %) | 51,902 (58 %) |
| 2003 | 37,784 (42 %) | 51,579 (58 %) |
| 2004 | 38,951 (43 %) | 51,149 (57 %) |
| 2005 | 39,500 (45 %) | 48,609 (55 %) |
| 2006 | 38,041 (44 %) | 48,335 (56 %) |
| 2007 | 40,281 (46 %) | 48,139 (54 %) |
| 2008 | 36,372 (42 %) | 50,221 (58 %) |
| 2009 | 35,768 (42 %) | 48,625 (58 %) |
| 2010 | 34,461 (42 %) | 46,821 (58 %) |
Includes all abortions performed in NYC, including those for non-NYC residents
TABLE 2.
Abortion procedures and number of past abortions, by patient characteristics, New York City, 2010
| Patient characteristics | Total number of abortions reported | Number (%) of abortions that were repeat procedures |
|---|---|---|
| Age (years) (0.4 % missing) | ||
| <15 | 395 | 50 (13 %) |
| 15–17 | 4165 | 913 (22 %) |
| 18–19 | 6837 | 2585 (38 %) |
| 20–24 | 22,184 | 12,385 (56 %) |
| 25–29 | 18,513 | 12,217 (66 %) |
| 30–34 | 12,357 | 8576 (69 %) |
| 35–39 | 6922 | 4839 (70 %) |
| ≥40 | 2698 | 1823 (68 %) |
| Race/ethnicity (5 % missing) | ||
| White non-Hispanic | 7,362 | 3322 (45 %) |
| Asian/Pacific Islander | 4,007 | 1736 (43 %) |
| Other Hispanic | 14,966 | 8205 (55 %) |
| Puerto Rican | 9,666 | 6278 (65 %) |
| Black non-Hispanic | 35,021 | 22,966 (66 %) |
| Neighborhood-level povertya (2 % missing) | ||
| Very low <10 % | 6,510 | 3,284 (50 %) |
| Low 10–19.9 % | 20,248 | 11,327 (56 %) |
| Medium 20–29.9 % | 17,767 | 10,429 (59 %) |
| High ≥30 % | 28,412 | 17,821 (63 %) |
| Number of living children (3 % missing) | ||
| 0 | 30,681 | 10,163 (33 %) |
| 1 | 19,436 | 13,944 (72 %) |
| 2 | 13,652 | 10,644 (78 %) |
| 3 | 5575 | 4426 (79 %) |
| 4 | 2028 | 1624 (80 %) |
| ≥ 5 | 1131 | 926 (82 %) |
| Education level (21 % missing) | ||
| Below high school | 10,801 | 5,664 (52 %) |
| High school | 29,220 | 18,847 (65 %) |
| Some college | 11,689 | 7400 (63 %) |
| College graduate | 7337 | 3776 (51 %) |
| Payment method (4 % missing) | ||
| Self | 16,103 | 8303 (52 %) |
| Medicaid | 32,252 | 19,889 (62 %) |
| Other third party payer | 23,037 | 14,430 (63 %) |
| Nativity (4 % missing) | ||
| USA | 51,575 | 31,428 (61 %) |
| Outside USA | 19,491 | 10,928 (56 %) |
| Marital status (4 % missing) | ||
| Married | 9616 | 5789 (60 %) |
| Unmarried | 62,141 | 37,012 (60 %) |
| Overall total | 74,347 | 43,567 (58 %) |
aNeighborhood-level poverty was defined as the percentage of the population in a ZIP code who live below the federal poverty threshold
Repeat abortion rates increased with patient age from 0.6/1000 for women aged <15 years to 37.5/1000 for women aged 20–24 years and decreased with age thereafter. Black non-Hispanics had a four times higher first abortion rate (27.6) and nine times higher repeat abortion rate (52.5) than white non-Hispanics (6.8 and 5.6, respectively). First and repeat abortion rates increased as neighborhood-level poverty increased, from 9.2 and 9.4 in very low poverty areas to 22.1 and 37.1 in high poverty areas, respectively (Table 3).
TABLE 3.
Rate of first and repeat abortion, by selected patient characteristics, New York City, 2010
| Patient characteristics | First abortion | Repeat abortion | ||
|---|---|---|---|---|
| No.a | Rate per 1000 population | No. | Rate per 1000 population | |
| Age (years) | ||||
| <15 | 345 | 3.9 | 50 | 0.6 |
| 15–17 | 3252 | 21.1 | 913 | 5.9 |
| 18–19 | 4252 | 39.2 | 2585 | 23.8 |
| 20–24 | 9799 | 29.7 | 12,385 | 37.5 |
| 25–29 | 6296 | 16.5 | 12,217 | 32.0 |
| 30–34 | 3781 | 11.0 | 8576 | 25.0 |
| 35–39 | 2083 | 6.9 | 4839 | 16.1 |
| ≥40 | 875 | 3.0 | 1823 | 6.2 |
| Race/ethnicity | ||||
| White non-Hispanic | 4040 | 6.8 | 3322 | 5.6 |
| Asian/Pacific Islander | 2271 | 8.6 | 1736 | 6.6 |
| Hispanicb | 10,149 | 18.1 | 14,483 | 25.9 |
| Black non-Hispanic | 12,055 | 27.6 | 22,966 | 52.5 |
| Neighborhood-level povertyc | ||||
| Very low <10 % | 3226 | 9.2 | 3284 | 9.4 |
| Low 10 %–19.9 % | 8921 | 14.2 | 11,327 | 18.1 |
| Medium 20–29.9 % | 7338 | 16.2 | 10,429 | 23.0 |
| High ≥30 % | 10,591 | 22.1 | 17,821 | 37.1 |
Numerator includes all abortion procedures reported for 2010 and is not deduplicated; thus, individual women might have had multiple procedures
aTotals do not match due to varying amounts of missing data for each variable of interest
bRate calculated for Hispanic women because census denominators are unavailable for Puerto Rican and other Hispanic women
cNeighborhood-level poverty was defined as the percentage of the population in a ZIP code who live below the federal poverty threshold
Table 4 summarizes the unadjusted mean number of past abortions and unadjusted and adjusted ratio of mean number of past abortions for each sociodemographic characteristic. On multivariable regression, all covariates of interest were significantly associated with self-reported number of past abortions. After multivariable adjustment, predicted number of past abortions increased with age. For women aged 35–39 years, the adjusted ratio of means (ARM) was 1.05 (95 % CI, 1.02–1.08), compared with women aged 25–29 years, indicating that mean number of past abortions among women aged 35–39 years was 5 % higher than among women aged 25–29 years after adjustment for other covariates. Similarly, the number of past abortions predicted by our model increased with increasing number of children. ARMs ranged from 0.50 (95 % CI, 0.49–0.51) for no children to 1.38 (95 % CI, 1.30–1.46) for ≥5 children, compared with 1 child. ARMs were highest for Puerto Ricans (0.99; 95 % CI, 0.96–1.01) and lowest for Asians and Pacific Islanders (0.67; 95 % CI, 0.64, 0.70) compared with black non-Hispanics. ARMs across neighborhood-level poverty, education, payment method, nativity, and marital status groups were of smaller magnitude.
TABLE 4.
Unadjusted mean number of past abortions and unadjusted and adjusted ratio of mean number of past abortions for 2010 abortion patients, by patient characteristics, New York City, 2010
| Patient characteristics | Unadjusted mean number of past abortions | Unadjusted ratio of means | Adjusted ratio of means (95 % CI) |
|---|---|---|---|
| Age (years) | |||
| <15 | 0.18 | 0.12 | 0.19 (0.14, 0.26) |
| 15–17 | 0.29 | 0.19 | 0.29 (0.27, 0.31) |
| 18–19 | 0.56 | 0.36 | 0.49 (0.47, 0.51) |
| 20–24 | 1.06 | 0.68 | 0.78 (0.76, 0.80) |
| 25–29 (ref.) | 1.56 | 1.00 | |
| 30–34 | 1.77 | 1.13 | 1.08 (1.05, 1.10) |
| 35–39 | 1.72 | 1.10 | 1.05 (1.02, 1.08) |
| ≥40 | 1.61 | 1.03 | 1.03 (0.99, 1.08) |
| Race/ethnicity | |||
| White non-Hispanic | 0.82 | 0.54 | 0.71 (0.68, 0.73) |
| Asian/Pacific Islander | 0.77 | 0.51 | 0.67 (0.64, 0.70) |
| Other Hispanic | 1.10 | 0.72 | 0.82 (0.80, 0.84) |
| Puerto Rican | 1.56 | 1.03 | 0.99 (0.96, 1.01) |
| Black non-Hispanic (ref.) | 1.52 | 1.00 | |
| Neighborhood-level povertya | |||
| Very low <10 % | 0.97 | 0.66 | 0.89 (0.86, 0.92) |
| Low 10–19.9 % | 1.21 | 0.82 | 0.99 (0.97, 1.01) |
| Medium 20–29.9 % | 1.30 | 0.88 | 0.97 (0.95, 0.99) |
| High ≥30 % (ref.) | 1.47 | 1.00 | |
| Number of living children | |||
| 0 | 0.56 | 0.36 | 0.50 (0.49, 0.51) |
| 1 (ref.) | 1.56 | 1.00 | |
| 2 | 1.93 | 1.24 | 1.15 (1.13, 1.18) |
| 3 | 2.11 | 1.35 | 1.22 (1.19, 1.26) |
| 4 | 2.21 | 1.42 | 1.24 (1.19, 1.30) |
| ≥5 | 2.50 | 1.60 | 1.38 (1.30, 1.46) |
| Education levelb | |||
| Below high school | 2.01 | 1.05 | 0.94 (0.92, 0.96) |
| High school (ref.) | 1.92 | 1.00 | |
| Some college | 1.48 | 0.77 | 0.93 (0.91, 0.95) |
| College graduate | 0.91 | 0.47 | 0.84 (0.79, 0.88) |
| Payment method | |||
| Self | 1.00 | 0.72 | 0.89 (0.87, 0.91) |
| Medicaid (ref.) | 1.39 | 1.00 | |
| Other third party payer | 1.44 | 1.04 | 1.02 (1.00, 1.04) |
| Nativity | |||
| USA (ref.) | 1.39 | 1.00 | |
| Outside USA | 1.12 | 0.81 | 0.78 (0.77, 0.80) |
| Marital status | |||
| Married | 1.31 | 0.99 | 0.86 (0.84, 0.88) |
| Unmarried (ref.) | 1.32 | 1.00 | |
aNeighborhood-level poverty was defined as the percentage of the population in a ZIP code who live below the federal poverty threshold
bFor education level, we restricted bivariate analysis to women aged ≥25 years
cResults from multivariate negative binomial regression modeling number of past abortions with other covariates held at reference values
Discussion
This is the first published analysis of predictors of repeat abortions based on a US population-based sample in over a decade and only the second such analysis ever published.2 Additionally, our multivariable analysis enabled us to explore relationships among sociodemographic characteristics and repeat abortion among women who have had at least one abortion. Among 2010 abortion patients, our analysis revealed smaller socioeconomic disparities than those observed by comparing citywide rates of first and repeat abortion. This demonstrates that, although socioeconomic disparities exist, the majority of abortion patients are at high risk for repeat unintended pregnancy and abortion.
Approximately 40,000 (57 %) of abortions performed in NYC in 2010 were repeat abortions. Repeat abortions were a substantial proportion of total abortions in each sociodemographic group we examined, constituting >50 % of total abortions for the majority of groups. Nationally, an estimated 45 % of abortions performed in 2010 were repeat abortions.10 Similarly, the Guttmacher Institute estimated that 48 % of abortions performed in the USA during 2000–2001 were repeat procedures.2 The higher proportion of repeat abortions in NYC might reflect demographic differences or greater access to abortion services. However, comparison data are unavailable for other demographically diverse urban areas.
Among 2010 NYC abortion patients, socioeconomic status variables (i.e., education, Medicaid payment, and neighborhood-level poverty) had a smaller relationship with number of past abortions than age, number of children, and race/ethnicity. Similarly, a California study in an urban hospital indicated no association between poverty, education, or insurance coverage and repeat abortions,3 and the Guttmacher Institute’s 2000–2001 national Abortion Patient Survey identified no association between individual poverty status and repeat abortions.
Older age has been associated with repeat abortions in this and other cross-sectional studies,2–4 but prospective studies have reported decreased incidence of future abortion in association with older age.11–13 The association appears in cross-sectional studies because older age represents a longer period at risk for unplanned pregnancy. Additionally, repeat abortion has been associated with cohabitation and greater coital frequency,14 both markers for increased pregnancy risk.
Repeat abortions among women with more children might reflect an increased likelihood of pregnancy termination after a woman has achieved or surpassed her desired family size. Lower numbers of past abortions among foreign-born women, although consistent with previous research,3 are insufficiently understood and might relate to lower rates of unintended pregnancy, limited access to abortion services, or unwillingness to disclose previous abortions.
The relationship between race/ethnicity and number of past abortions was stronger than that of socioeconomic factors in our analysis. We included Asians/Pacific Islanders and Puerto Ricans, who had not been studied previously at the population level. Black non-Hispanic and Puerto Rican abortion patients had approximately 40 % more past abortions on average than white non-Hispanic women. At the population level, black non-Hispanics had a nine times higher repeat abortion rate than white non-Hispanics. Racial/ethnic disparities also exist for repeat unintended childbearing: black women are more likely to have repeat unintended births than women of other racial/ethnic groups.15 Thus, black women are at higher risk for repeat unintended pregnancy, regardless of pregnancy outcome. These differences might reflect underlying disparities in access to or use of family planning services. Cultural attitudes toward abortion,16 contraception, and the health care system, lack of control over contraceptive decision making, or ambivalence about pregnancy intentions might also contribute.17
This study provides information solely on patient demographics and does not address women’s contraceptive preferences or practices. However, others have reported that although repeat abortion patients are more likely to have been using contraception than women undergoing a first abortion,14,18 inconsistent use or nonuse of contraception remains common.4,19 Women discontinue contraceptives or use them inconsistently for various reasons, including side effects, cost, disruptive life events, and misperceptions regarding pregnancy risk.20
Abortion providers have an opportunity to reduce future unintended pregnancies by providing contraception counseling and services on the day of the procedure. The majority of US abortion providers offer contraception counseling (96 %) or oral contraceptive pills (99 %) at the abortion visit,21 but contraceptive counseling, alone without providing a contraceptive method, is ineffective at preventing future unplanned pregnancy and abortion.22,23 The American College of Obstetricians and Gynecologists recommends offering long-acting reversible contraceptives (LARC) as first-line contraception for most women, and adopting same day insertion protocols for surgical abortions.24–26 However, fewer than one third of US providers offer immediate postabortion LARC.21 NYC’s Bureau of Vital Statistics began collecting data on postabortion contraception in 2011. Data from the latter half of 2012 show that contraception was not offered at 29 % of abortion visits. Contraception was offered but declined at 13 % of visits. Among the 58 % of visits where contraception was offered and accepted, 18 % of patients received a long-acting method (IUD, implant or injectable) (NYC Vital Statistics unpublished data). Data are not available on the proportion of patients who were offered but declined LARC or who underwent medication abortions and were ineligible for same-day LARC insertion. Barriers to LARC provision include higher initial costs to patients, unfavorable insurance reimbursement policies (e.g., reduced compensation for LARC insertion performed during the abortion visit) and in some cases, lack of staff trained to perform LARC insertion.27,28
Because the risk of repeat abortion is significantly higher among black non-Hispanic and Puerto Rican women, attention should be paid to barriers to their use of postabortion contraception. Research shows that black and Hispanic women have greater misconceptions about contraception and concerns about contraceptive safety.29 Moreover, the history of coercive family planning policies in the USA30 and ongoing experiences of race-based discrimination when accessing family planning services31 has led to mistrust of the health care system and negative attitudes toward contraception.32 To address these barriers, postabortion contraceptive counseling and services should provide accurate information about all contraceptive choices while addressing questions about contraceptive safety and concerns arising from patient mistrust. Research is needed regarding current practices in postabortion contraception and factors that influence providers’ decisions to offer various contraceptive methods. Additionally, research is needed on factors that influence women’s choice of contraceptive method and decisions to continue or discontinue contraception.
In contrast to most published studies of repeat abortion, our study population is large, demographically diverse and population-based. We were able to calculate population rates of repeat abortion by age, race/ethnicity, and neighborhood-level poverty. Additionally, we newly describe repeat abortion epidemiology among Asian and Puerto Rican NYC residents.
Limitations
One limitation of this study was that past pregnancy outcome data relied on patient self-report. Underreporting of prior abortions among women accessing abortion services has been estimated at 9 %. Nonwhite women and women with a lower educational level were more likely to underreport.33 If this occurred in our population, we may have underestimated the effects of race/ethnicity and education on number of abortions. Moreover, our unit of analysis is the abortion event, not the woman, because patient-level identifiers are unavailable. Thus, if a patient had more than one abortion during 2010, we could not account for overrepresentation of her characteristics. Information is limited regarding spacing of repeat abortions in the USA. In a 1977 clinic-based study in Atlanta, Georgia, approximately half of repeat abortion patients had had an abortion in the previous 12 months.18 Further, using abortion as a proxy for unintended pregnancy underestimates unintended pregnancy incidence because approximately 37 % of births also result from unintended pregnancy.34 Additionally, we had a relatively high proportion of missing values for education, requiring the use of multiple imputation, which although shown to produce less bias and greater precision than complete case analysis, may still introduce bias.35 Finally, we could not exclude therapeutic abortions, i.e., abortions performed because of a woman’s health or fetal abnormalities. Therapeutic abortions, which comprise approximately 5 % of all abortions,1 are probably more common among older age groups. Including them in this analysis may have caused us to overestimate the relationship between age and number of abortions.
Conclusions
Repeat abortions have comprised over 50 % of total abortions in NYC for 10 years. We demonstrate racial/ethnic and economic disparities in repeat abortion rates and show that the majority of abortion patients are at high risk for repeat unintended pregnancy and abortion. Unfortunately, many women remain without access to postabortion contraception services despite the fact that such services decrease the incidence of repeat abortion. Addressing barriers like high initial costs, unfavorable insurance reimbursement policies, and lack of trained providers would extend the availability of postabortion contraception services to more women.
Acknowledgments
The authors thank Julie Magri, MD, MPH, Centers for Disease Control and Prevention and Cristina Yunzal-Butler, PhD, New York City Department of Health and Mental Hygiene.
Funding
No financial disclosures were reported by the authors of this paper.
Footnotes
Note
The findings and conclusions in this report are those of the author(s) and do not necessarily represent the official position of the Centers for Disease Control and Prevention.
References
- 1.Finer LB, Zolna MR. Shifts in intended and unintended pregnancies in the United States, 2001–2008. Am J Public Health. 2014;104(Suppl 1):S43–S48. doi: 10.2105/AJPH.2013.301416. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Jones R, Singh S, Finer L, Frohwirth L. Repeat abortion in the United States. New York, NY: Guttmacher Institute; 2006. Available at: http://www.guttmacher.org/pubs/2006/11/21/or29.pdf. Accessed September 10, 2014
- 3.Prager SW, Steinauer JE, Foster DG, Darney PD, Drey EA. Risk factors for repeat elective abortion. Am J Obstet Gynecol. 2007;197:575.e1-575.e6. [DOI] [PubMed]
- 4.Bleil ME, Adler NE, Pasch LA, Sternfeld B, Reijo-Pera RA, Cedars, MI. Adverse childhood experiences and repeat induced abortion. Am J Obstet Gynecol. 2011;204:122.e1-6. doi:10.1016/j.ajog.2010.09.029. [DOI] [PMC free article] [PubMed]
- 5.Toprani A, Madsen A, Das T, Gambatese M, Greene C, Begier E. Evaluating New York City's Abortion Reporting System: insights for public health data collection systems. J Public Health Manag Pract. 2014;20:392–400. doi: 10.1097/PHH.0b013e31829c88b8. [DOI] [PubMed] [Google Scholar]
- 6.Rules of the City of New York, Title 24, Article 203. Available at: http://rules.cityofnewyork.us/codified-rules. Accessed September 10, 2014.
- 7.New York City Department of Health and Mental Hygiene. Annual summary of vital statistics. Available at: http://www.nyc.gov/html/doh/html/data/vs-summary.shtml. Accessed September 10, 2014.
- 8.Toprani A, Hadler JL. Selecting and Applying a Standard Area-based Socioeconomic Status Measure for Public Health Data: Analysis for New York City. New York City Department of Health and Mental Hygiene Epidemiology Research Report 2013; 1–12. Available at: http://www.nyc.gov/html/doh/downloads/pdf/epi/epiresearch-SES-measure.pdf Accessed September 10, 2014.
- 9.Little RJ, Rubin DB. Statistical Analysis with Missing Data. Hoboken, NJ: Wiley; 2002. [Google Scholar]
- 10.Centers for Disease Control and Prevention. Abortion Surveillance—United States, 2010. MMWR Surveill Sum. 2013;62(No. SS-8):1–44. [PubMed]
- 11.Heikinheimo O, Gissler M, Suhonen S. Age, parity, history of abortion and contraceptive choices affect the risk of repeat abortion. Contraception. 2008;78:149–154. doi: 10.1016/j.contraception.2008.03.013. [DOI] [PubMed] [Google Scholar]
- 12.Niinimaki M, Pouta A, Bloigu A, et al. Frequency and risk factors for repeat abortions after surgical compared with medical termination of pregnancy. Obstet Gynecol. 2009;113:845–852. doi: 10.1097/AOG.0b013e31819cae06. [DOI] [PubMed] [Google Scholar]
- 13.Osler M, Morgall JM, Jensen B. Repeat abortion in Denmark. Dan Med Bull. 1992;39:89–91. [PubMed] [Google Scholar]
- 14.Berger C, Gold D, Andres D. Repeat abortion: is it a problem? Fam Plann Perspect. 1984;16:70–75. doi: 10.2307/2134932. [DOI] [PubMed] [Google Scholar]
- 15.Wildsmith E, Guzzo K, Hayford S. Repeat unintended, unwanted and seriously mistimed childbearing in the United States. Perspect Sex Reprod Health. 2010;42:14–22. doi: 10.1363/4201410. [DOI] [PubMed] [Google Scholar]
- 16.Strickler J, Danigelis NL. Changing frameworks in attitudes toward abortion. Sociological Forum. 2002;17:187–201. doi: 10.1023/A:1016033012225. [DOI] [Google Scholar]
- 17.Kendall C, Afable-Munsuz A, Speizer I, Avery A, Schmidt N, Santelli J. Understanding pregnancy in a population of inner-city women in New Orleans—results of qualitative research. Soc Sci Med. 2005;60:297–311. doi: 10.1016/j.socscimed.2004.05.007. [DOI] [PubMed] [Google Scholar]
- 18.Leach J. The repeat abortion patient. Fam Plann Perspect. 1977;9:37–39. doi: 10.2307/2134295. [DOI] [PubMed] [Google Scholar]
- 19.Jones RK, Darroch JE, Henshaw SK. Contraceptive use among US women having abortions in 2000–2001. Perspect Sex Reprod Health. 2002;34:294–303. doi: 10.2307/3097748. [DOI] [PubMed] [Google Scholar]
- 20.Frost JJ, Darroch JE, Remez L. Improving Contraceptive Use. New York, NY: Guttmacher Institute; 2008. [PubMed] [Google Scholar]
- 21.Kavanaugh ML, Jones RK, Finer LB. How commonly do US abortion clinics offer contraceptive services? Contraception. 2010;82(4):331–336. doi: 10.1016/j.contraception.2010.04.010. [DOI] [PubMed] [Google Scholar]
- 22.Ferreira AL, Lemos A, Figueiroa JN, de Souza AI. Effectiveness of contraceptive counselling of women following an abortion: a systematic review and meta-analysis. Eur J Contracept Reprod Health Care. 2009;14(1):1–9. doi: 10.1080/13625180802549970. [DOI] [PubMed] [Google Scholar]
- 23.Halpern V, Lopez LM, Grimes DA, Stockton LL, Gallo MF. Strategies to improve adherence and acceptability of hormonal methods of contraception. Cochrane Database Syst Rev. 2013;10:CD004317. doi: 10.1002/14651858.CD004317.pub4. [DOI] [PubMed] [Google Scholar]
- 24.Okusanya BO, Oduwole O, Effa EE. Immediate postabortal insertion of intrauterine devices. Cochrane Database Syst Rev. 2014;7:CD001777. doi: 10.1002/14651858.CD001777.pub4. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Madden T, Westhoff C. Rates of follow-up and repeat pregnancy in the 12 months after first-trimester induced abortion. Obstet Gynecol. 2009;113:663–668. doi: 10.1097/AOG.0b013e318195dd1e. [DOI] [PubMed] [Google Scholar]
- 26.American College of Obstetricians and Gynecologists (ACOG). ACOG Committee Opinion No 450. Increasing Use of Contraceptive Implants and Intrauterine Devices To Reduce Unintended Pregnancy. Obstet Gynecol 2009;114:1434–8. doi:10.1097/AOG.0b013e3181c6f965. [DOI] [PubMed]
- 27.Morse J, Freedman L, Speidel JJ, Thompson KM, Stratton L, Harper CC. Postabortion contraception: qualitative interviews on counseling and provision of long-acting reversible contraceptive methods. Perspect Sex Reprod Health. 2012;44(2):100–106. doi: 10.1363/4410012. [DOI] [PubMed] [Google Scholar]
- 28.Thompson KM, Speidel JJ, Saporta V, Waxman NJ, Harper CC. Contraceptive policies affect post-abortion provision of long-acting reversible contraception. Contraception. 2011;83(1):41–47. doi: 10.1016/j.contraception.2010.06.008. [DOI] [PubMed] [Google Scholar]
- 29.Dehlendorf C, Harris LH, Weitz TA. Disparities in abortion rates: a public health approach. Am J Public Health. 2013;103:1772–1779. doi: 10.2105/AJPH.2013.301339. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Stern AM. Sterilized in the name of public health: race, immigration, and reproductive control in modern California. Am J Public Health. 2005;95:1128–1138. doi: 10.2105/AJPH.2004.041608. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Becker D, Tsui AO. Reproductive health service preferences and perceptions of quality among low-income women: racial, ethnic and language group differences. Perspect Sex Reprod Health. 2008;40:202–211. doi: 10.1363/4020208. [DOI] [PubMed] [Google Scholar]
- 32.Bird ST, Bogart LM. Conspiracy beliefs about HIV/AIDS and birth control among African Americans: implications for the prevention of HIV, other STIs, and unintended pregnancy. J Soc Issues. 2005;61:109–126. doi: 10.1111/j.0022-4537.2005.00396.x. [DOI] [PubMed] [Google Scholar]
- 33.Udry JR, Gaughan M, Schwingl PJ, Van den Berg BJ. A medical record linkage analysis of abortion underreporting. Fam Plann Perspect. 1996;28:228–231. doi: 10.2307/2135842. [DOI] [PubMed] [Google Scholar]
- 34.Mosher WD, Jones J, Abma JC. Intended and Unintended Births in the United States: 1982–2010. National Health Statistics Reports; no 55. Hyattsville, MD: National Center for Health Statistics; 2012. [PubMed] [Google Scholar]
- 35.Greenland S, Finkle WD. A critical look at methods for handling missing covariates in epidemiologic regression analyses. Am J Epidemiol. 1995;142:1255–1264. doi: 10.1093/oxfordjournals.aje.a117592. [DOI] [PubMed] [Google Scholar]