Women's Health Issues 25-3 (2015) 209–215

www.whijournal.com

Original article

Reproductive Life Plan Counseling and Effective Contraceptive Use among Urban Women Utilizing Title X Services Aalap Bommaraju, MPH a,b,*, Jennifer Malat, PhD a,c, Jennifer L. Mooney, PhD a a

Department of Sociology, University of Cincinnati, Cincinnati, Ohio Cincinnati-Hamilton County Reproductive Health and Wellness Program, Cincinnati, Ohio c Kunz Center for Social Research, University of Cincinnati, Cincinnati, Ohio b

Article history: Received 5 March 2014; Received in revised form 13 February 2015; Accepted 25 February 2015

a b s t r a c t Background: Although the Centers for Disease Control and Prevention and the U.S. Office of Population Affairs recommend inclusion of reproductive life plan counseling (RLPC) in all well-woman health care visits, no studies have examined the effect of RLPC sessions on the decision to use effective contraception at publicly funded family planning sites. RLPC could be a particularly impactful intervention for disadvantaged social groups who are less likely to use the most effective contraceptive methods. Methods: Using data from 771 nonpregnant, non–pregnancy-seeking women receiving gynecological services in the CincinnatidHamilton County Reproductive Health and Wellness Program, multinomial logistic regression models compared users of nonmedical/no method with users of 1) the pill, patch, or ring, 2) depot medroxyprogesterone acetate, and 3) long-acting reversible contraception (LARC). The effect of RLPC on the use of each form of contraception, and whether it mediated the effect of race/ethnicity and education on contraceptive use, was examined while controlling for age, insurance status, and birth history. The interaction between RLPC and race/ethnicity and the interaction between RLPC and educational attainment was also assessed. Findings: RLPC was not associated with contraceptive use. The data suggested that RLPC may increase LARC use over nonmedical/no method use. RLPC did not mediate or moderate the effect of race/ethnicity or educational attainment on contraceptive use in any comparison. Conclusions: In this system of publicly funded family planning clinics, RLPC seems not to encourage effective method use, providing no support for the efficacy of the RLPC intervention. The results suggest that this intervention requires further development and evaluation. Copyright Ó 2015 by the Jacobs Institute of Women’s Health. Published by Elsevier Inc.

The United States experiences higher rates of unintended pregnancy than nearly every other nation of similar economic status (Finer & Zolna, 2011; Singh, Sedgh, & Hussain, 2010). These high rates are partly owing to high rates of unintended pregnancy among historically disadvantaged populations (Mosher, Jones, & Abma, 2012). An analysis of the 2006–2010 National Survey of Family Growth showed that never married women, younger women, and African-American women were at greater risk for unintended pregnancy than other women, in part because they used effective contraception less often (Jones,

Funding sources: None. * Correspondence to: Aalap Bommaraju, MPH, Department of Sociology, University of Cincinnati, Cincinnati, P.O. Box 210378, OH 45221-0378. Phone: 513-708-5432; fax: 513-556-0057. E-mail address: [email protected] (A. Bommaraju).

Mosher, & Daniels, 2012). Unintended pregnancy is a public health concern that disproportionately burdens individuals in marginalized social locations. With the recent increase in statelevel restrictions on abortion access, it is increasingly important that women seeking to avoid pregnancy be empowered to choose from among the most effective contraceptive methods (Medoff, 2012). Reproductive life plan counseling (RLPC) is a conversation between a medical service provider and a patient about how control of reproduction fits into a patient’s future life plans (Centers for Disease Control and Prevention [CDC], 2006; Files et al., 2011). This discussion is meant to disambiguate pregnancy intention and render contraceptive decision making a discrete action with concrete risks and benefits. Thus, the RLPC session is meant to cultivate knowledge of pregnancy intention in the patient in an effort to guide the patient toward choosing

1049-3867/$ - see front matter Copyright Ó 2015 by the Jacobs Institute of Women’s Health. Published by Elsevier Inc. http://dx.doi.org/10.1016/j.whi.2015.02.005

210

A. Bommaraju et al. / Women's Health Issues 25-3 (2015) 209–215

the most effective contraceptive method for themselves. Although both the CDC and the U.S. Office of Population Affairs recommend that RLPC sessions should be a component of every well-woman visit, evaluation of the effectiveness of this intervention is lacking (Gavin et al., 2014). In this article, we assess the effectiveness of RLPC among patients covered by Title X services in one Midwestern, U.S. city. Specifically, we examined whether the use of effective forms of contraception is associated with having received RLPC and whether RLPC decreases the education gap or the gap between Black and Latina women and White women in effective contraceptive use. Because federal and state agencies recommend RLPC, our analyses address an important policy question. Effective Reversible Contraceptive Methods There are currently numerous effective reversible contraceptive methods available to women in the United States. These methods can be categorized into three broad classes according to estimates of their typical use efficacy and mode of administration: 1) hormonal oral contraceptives, transdermal patches, and vaginal ringsdrequiring daily, weekly, or monthly maintenance, 2) depot medroxyprogesterone acetate (DMPA), an injectable contraceptive that must be administered every 3 months, and 3) long-acting reversible contraception (LARC), including the intrauterine device (IUD) and the subdermal implant, which last between 3 and 10 years after insertion, depending on the device. LARCs are the most effective reversible contraceptive methods available currently for women. These methods have typical use efficacy rates that are similar to female sterilization, and they can be easily reversed with rapid return to fertility (Hatcher et al., 2011). In addition, LARC methods have far fewer contraindications than many other types of contraceptive methods (American College of Obstetricians and Gynecologists, 2011). Population estimates of the use of effective contraceptive methods, including all three classes mentioned, differ by race/ ethnicity and socioeconomic status. Jacobs and Stanfors’ (2013) examination of the 2006–2010 National Survey of Family Growth found that Black, and to a lesser degree Latina, women were less likely than White women to utilize effective hormonal contraception. Examining factors affecting contraceptive use from the 2006–2010 National Survey of Family Growth, Dehlendorf et al. (2014b) found that women with at least some college education had significantly higher odds of choosing any effective contraceptive method over no contraceptive method when compared with women who had less than a high school diploma. Additionally, Jones et al. (2012) found that a lower percentage of women at risk of unintended pregnancy with less than a high school diploma chose IUDs when compared with more highly educated women. It is important to note that racial/ ethnic differences exist in patterns of contraceptive use among methods of relatively equivalent efficacy as well. For example, Dehlendorf et al., (2011) found that Black and Latina clients in California’s family planning program for low-income women were less likely to use oral contraceptives, vaginal rings, and LARC methods than White women, but more likely to choose patches and DMPA. Despite variation in patterns of individual method use, this body of research suggests that disadvantaged racial/ethnic status and low educational attainment is generally associated with women using nonmedical methods or no contraception at all. Racial and socioeconomic status differences in contraceptive use patterns can be attributed to social, cultural, and economic

processes that influence contraceptive choice. First, knowledge about the availability, safety, and reliability of effective contraceptive methods depends on a potential user’s relationships with other women who have used or continue to use these methods and the potential user’s level of social support for the use of these methods (Sangi-Haghpeykar, Ali, Posner, & Poindexter, 2006; Teal & Romer, 2013). Second, as with other medical interactions (Shim, 2010), contraceptive counseling over effective method use requires negotiation between the provider and patient to develop shared understandings about efficacy and uses for these complex technologies. Thus, provider–patient interactions play a critical role in an individual’s decision to use an effective contraceptive method (Dehlendorf, Kimport, Levy, & Steinauer, 2014a; Kavanaugh, Frowirth, Jerman, Popkin, & Ethier, 2013). Finally, because effective contraceptive methods can have prohibitively high up-front costs, economic capital is an important factor in considering certain methods over others (Dennis & Grossman, 2012; Trussell, Hassan, Lowin, Law, & Filonenko, 2015). Public health interventions designed to increase usage of effective contraceptive methods focus on two domains: reducing economic barriers and cultivating knowledge of the benefits of effective contraceptive use through patient education. Interventions that reduce the cost of contraception have been successful in increasing use of effective contraceptive methods (Foster et al., 2011; Secura et al., 2014). The helpfulness of efforts to cultivate knowledge of contraceptive methods is less clear. One route to improve knowledge of contraceptive options is through the process of RLPC in the provider–patient interaction. During RLPC, providers attempt to take a life-course perspective toward the reproductive health of their patient. The patient is asked to think about pregnancy intention, future life goals, and the role of reproduction in those life goals. The results of this conversation allow for the provider to suggest medical and nonmedical interventions to either facilitate a healthy pregnancy or prevent pregnancy from occurring. If the patient is not seeking pregnancy, the provider is expected to discuss various contraceptive options starting from the most effective methods that do not have medical contraindications (usually LARCs) and going down the list of methods based on 1-year typical use efficacy estimates. The effectiveness of RLPC is unknown. Social factors may impede or improve the effectiveness of RLPC. First, the race and class of a patient has an effect on what happens in the doctor–patient interaction. Racial/ethnic minorities and less educated patients often report worse experiences than White and highly educated patients in health care interactions (van Ryn et al., 2011). In addition, patients are less likely to adopt medical technologies, particularly newer ones, when they lack social and cultural knowledge of the services being offered (Phelan, Link, & Tehranifar, 2010). Additionally, given the unjust racial history of public family planning efforts in the United States (Roberts, 1997), contraceptive counseling interactions with health care providers may be interpreted differently depending on the race/ethnicity of the patient or provider. We are aware of no study that describes how RLPC from health care providers influences effective contraceptive use, whether it has similar effects for all racial/ethnic and education groups, and whether it decreases racial/ethnic and educational differences in contraceptive use. Our sample from a publicly funded health system that minimizes the cost of contraceptive access allows an analysis that focuses on the effect of counseling free from of economic barriers.

A. Bommaraju et al. / Women's Health Issues 25-3 (2015) 209–215

Methods Study Site Data for this study are from the CincinnatidHamilton County Reproductive Health and Wellness Program (RHWP) during the period March 2012 through February 2013. The RHWP is a grantfunded program that provides low-cost reproductive health services to low-income men and women in Cincinnati and Hamilton County, Ohio. Operating out of the Cincinnati Health Department’s primary care health centers, the RHWP provides the full gamut of contraceptive methods for every gynecological visit with a woman who is capable of becoming pregnant. These health centers serve individuals in the county who are unable to access private health services owing to their insurance status or ability to pay out of pocket. The Ohio Department of Health funds the RHWP through the Department of Health and Human Services’ Title X Family Planning program. The RHWP issues discounts for services, including costs of contraceptive devices, based on a sliding scale that assesses the individual’s income and family size as a percent of the current federal poverty guidelines. This sliding scale allows for those at or below the federal poverty line to receive free services. Individuals with incomes >100% of the federal poverty line qualify for incremental discounts until their income eclipses 250% of the federal poverty line. Income verification is required for other health department services, but the RHWP accepts self-declared incomes. During the period under study, the Cincinnati Health Department health centers did not formally collect fees assessed to individuals who were required to pay partial or full amounts if they lacked proof of income, were unable to pay, or were unwilling to pay at the time of service. Thus, in this study, barriers to contraceptive use due to financial reasons were very low. Participants Data for the study were collected from RHWP patients by clerical and clinical staff members at each clinical site from March 2012 through February 2013. Data were recorded in the patient’s electronic medical record and then transmitted to a secondary database. This secondary database contained clinical encounter information for every patient receiving Title X services at the RHWP. It was used to construct the deidentified dataset used for this research. The composition of this dataset was limited in its scope because it was collected for grant reporting purposes. As such, the data used for this study were not disaggregated by health center or provider. Both the University of Cincinnati Institutional Review Board and the Cincinnati Health Department Institutional Review Board approved this study. Individuals were included in this study if they 1) received services from the RHWP during the study period, 2) had their initial visit to the RHWP on or before January 11, 2013, 3) were female, 4) were 16 years old or older, 5) had at least 7 weeks to have a follow-up appointment with the RHWP to receive their contraceptive method of choice, and 6) self-identified as Black, White, or Latina. Individuals were excluded if they 1) were seeking pregnancy or were pregnant when they received services through the RHWP at any time during the study period or 2) were not physically capable of becoming pregnant at the time of their visit. Patients who were under 18 years old during the study period did not require parental permission to receive services from the RHWP because of Title X regulations

211

concerning minor confidentiality. Out of 1,422 unique female patients in the RHWP during the study period, 771 were retained for the sample using these criteria. Measures Providers and nurses were trained on RLPC through mandatory completion of a standardized, web-based Title X orientation program (Cardea, 2014). Unfortunately, this training tool was removed from the internet when Title X guidelines were updated in April of 2014, after the study period closed. This training was supplemented with specialized training on RLPC using examples of RLPC worksheets drawn from the CDC and the Delaware Healthy Mothers and Infants Consortium (CDC, 2012; Noyes & Savin, 2012). Providers were trained to open a dialogue with patients about their future life plans and the impact of pregnancy and parenthood on those plans. Although shared decision making was emphasized in this training, specific techniques of motivational interviewing were not part of the training. Providers were trained to discuss contraceptive options in order of typical use efficacy estimates depending on medical eligibility criteria (CDC, 2010; Steiner et al., 2006). This RLPC training protocol was necessarily specific to the RHWP because, at the time of the study, the Ohio Department of Health did not require Title X–funded sites to use a standardized tool to guide the provision of RLPC. The RHWP required no formal documentation of RLPC sessions; providers were only required to report whether or not such a counseling session occurred during the patient encounter. Answering this question was required before closing the encounter in the electronic medical record. Although RLPC was expected to occur for most clients, RLPC did not occur if providers lacked time for this intervention or if patients refused to engage in conversations concerning their reproductive life plans. RLPC was measured through a binary categorical variable indicating whether or not the patient ever had a RLPC session during any RHWP clinical visit over the course of the study period. If they had an RLPC session before the study period, it was not recorded. However, the RHWP had only recently begun seeing clients at the beginning of the study period. Thus, it is unlikely that any clients had received RLPC before the initiation of this study. The dependent variable of interest was contraceptive choice at the end of the study period. The variable was recoded into four categories: (1) nonmedical/no contraceptive method, including no method, natural family planning methods, and barrier methods, (2) hormonal pills, patches, and rings; (3) DMPA, and (4) LARC methods, including IUDs, intrauterine systems, and hormonal implants. These categories were created based on the differential 1-year typical use efficacy of each type of contraception (Hatcher et al., 2011). Owing to its nature as a semi–longacting form of contraception, DMPA was categorized on its own in this schema despite having similar typical use efficacy to oral, patch, and ring contraceptives. Other variables of interest were race and educational attainment. Patients were asked to self-report their race as one of seven categories and to report whether they were of Hispanic origin. The variable used here recoded these variables into nonHispanic Black, non-Hispanic White, and Hispanic. Hispanic patients are referred to as Latina throughout this study. Other racial groups were excluded because their numbers were too small for analysis. Educational attainment was recoded into a dichotomous variable indicating whether or not the patient had a high school diploma or GED equivalent.

212

A. Bommaraju et al. / Women's Health Issues 25-3 (2015) 209–215

participants (41.8%) received an RLPC session during their term in the RHWP. Table 1 also shows the bivariate associations between contraceptive method use and the independent variables. RLPC was not associated with type of contraception, although the pattern of results suggests that those who received RLPC were less likely to use no medical method of contraception. The association between race and contraceptive choice was significant (p < .001), with Black women being more likely to use no method and less likely to use a LARC method. Latinas were the least likely to use nonmedical/no method. Finally, there were small but significant differences in education level among users of the different methods (p < .001). Pill/patch/ring users had higher educational attainment than the other groups, with DMPA users containing the fewest numbers of women with a high school diploma or GED. Table 2 contains the results from two multinomial logistic regression models. Model 1 shows the effect of race/ethnicity and education on contraceptive choice, when controlling for age, insurance status, history of a recent birth, and the existence of living children who were born to the patient. Black racial identification was only a significant indicator of lower odds of LARC use in the comparison between nonmedical/no method users and LARC users (OR, 0.379; 95% CI, 0.195–0.738). The model indicated no difference between Latinas and Whites (p > .10). Having a high school diploma or GED was significantly associated with lower odds of choosing DMPA over nonmedical/no method (OR, 0.533; 95% CI, 0.322–0.883). In Model 2, we added RLPC to the multinomial logistic regression. The change in the -2 log likelihood, based on a test that includes all of the variables in Model 2, indicated that the effect of RLPC was not significant (p ¼ .209). The results suggest that there may be an effect of RLPC in the comparison between LARC users and nonmedical/no method users (OR, 1.635; 95% CI, 1.025–2.608), but the results of the model comparisons do not allow us to conclude that the effect is significant. Because RLPC is not associated with contraceptive choice, when controlling for the other variables in the model, we conclude that RLPC does not mediate the relationship between race/ethnicity and contraceptive use or between education and contraceptive use. Because RPLC may have had different effects based on race/ethnicity or education that masked an overall effect, we tested for an interaction between RLPC and these variables. The tests were not significant (results available upon request). Consequently, we conclude that the effect of RLPC on

Several control variables associated with contraceptive choice were also included in the logistic regression model. Age was coded as a mean-centered, continuous numerical variable. Insurance status, a dichotomous variable, indicated whether or not the patient had health insurance during their last RHWP visit. Whether the patient had any living children born to them and whether or not the patient had a birth within the last 18 months of their last visit were measured as separate binary categorical variables. Analytic Strategy Using SPSS Version 22 (IBM Corp, 2013), multinomial logistic regression was used to estimate the models comparing nonmedical/no method use to the other methods. We used nested, y-standardized regression models to test whether RLPC mediated race/ethnicity and education differences in contraceptive choice (Mood, 2010). Model 1 included the indicators for race and education, as well as the other control variables. Model 2 introduced the indicator of RLPC. To test for the significance of the variables of interest, we examined the change that occurred in the -2 log likelihood if a particular coefficient was removed from the model. As shown, RLPC was not a significant predictor, which indicates that there was not a mediation effect. Therefore, for simplicity, Table 2 does not present y-standardized coefficients. Rather, it includes the adjusted odds ratios (with nonmedical/no method functioning as the reference group), 95% CIs, and p values. Two additional models tested interaction effects between RLPC and race/ethnicity and between RLPC and educational attainment, but were not included because the results were not significant. Results Table 1 presents the distribution of variables used in our analysis. The final sample included 771 participants. The distribution of ending contraceptive method choice within the population was relatively uniform: 170 women (22.1%) were in the nonmedical/no method group; 207 (26.8%) used pills, patches, or vaginal rings; 213 (27.6%) used DMPA; and 181 (23.5%) used LARC methods. Nearly three-quarters of the population (73.7%) selfidentified as Black, 14.5% of participants were Latina, and 11.8% were White. About one-third of the sample did not have a high school diploma or GED equivalent. Roughly two-fifths of the Table 1 Distribution of Factors Affecting Contraceptive Use among Contraceptive User Groups Factor or Covariate

Nonmedical/No Method Users

Pill/Patch/Ring Users

DMPA Users

LARC Users

n or Mean (% in Row or SD) RLPC Race*** Black White Latina High school diploma/GED*** Age*** Insured Recent birth*** Has living children*** Total

59 (18.3%) 140 18 12 131 30.14 78 12 101 170

(24.6%) (19.8%) (10.7%) (25.7%) (7.89) (20.3%) (13.0%) (19.6%) (22.1%)

86 (26.7%) 157 26 24 137 25.98 97 14 110 207

(27.6%) (28.6%) (21.4%) (26.9%) (6.32) (25.2%) (15.2%) (21.3%) (26.8%)

92 (28.6%) 159 16 38 119 26.41 121 33 161 213

(28.0%) (17.6%) (33.9%) (23.4%) (6.42) (31.4%) (35.9%) (31.2%) (27.6%)

85 (26.4%) 112 31 38 122 28.29 89 33 144 181

(19.7%) (34.1%) (33.9%) (24.0%) (6.90) (23.1%) (35.9%) (27.9%) (23.5%)

Total, n or Mean (% of Total or SD) 322 (41.8%) 568 91 112 509 27.58 385 92 516 771

(73.7%) (11.8%) (14.5%) (66.0%) (7.03) (49.9%) (11.9%) (66.9%) (100%)

Abbreviations: DMPA, depot medroxyprogesterone acetate; LARC, long-acting reversible contraception; RLPC, reproductive life plan counseling. p < .001; **p < .01; *p < .05. Significant differences in “Age” reported using multinomial logistic regression (for comparison between nonmedical/no method users with pill/patch/ring users p < .001, with DMPA users p < .001, and with LARC users p < .05). Significant differences with all other variables reported using the c2 test.

***

A. Bommaraju et al. / Women's Health Issues 25-3 (2015) 209–215

213

Table 2 Adjusted Odds Ratios for Multinomial Logistic Regression of Factors Affecting Contraceptive Method Use Characteristic

Model 1 OR*,y

Model 2 95% CI Low

Nonmedical/no method users and pill/patch/ring users Black 0.690 0.352 Latina 1.141 0.416 High school diploma/GED 0.740 0.444 RLPC – – Nonmedical/no method users and DMPA users Black 0.990 0.468 Latina 2.159 0.766 High school diploma/GED 0.533 0.322 RLPC – – Nonmedical/no method users and LARC users Black 0.379 0.195 Latina 1.549 0.582 HS Diploma/GED 1.037 0.599 RLPC – – Total no. of cases 771 2 D model c (dF) –

p

OR*,y

High

95% CI

p

Low

High

1.351 3.128 1.233

.279 .798 .247 –

0.683 1.178 0.728 1.269

0.348 0.428 0.436 0.813

1.341 3.243 1.216 1.979

.268 .751 .226 .294

2.092 6.086 0.883

.979 .145 .015 –

0.980 2.272 0.520 1.414

0.463 0.802 0.313 0.900

2.078 6.434 0.864 2.221

.959 .122 .012 .133

0.738 4.124 1.795

.004 .381 .898 –

0.375 1.660 0.990 1.635 771 4.538 (3)

0.192 0.620 0.570 1.025

0.734 4.446 1.719 2.608

.004 .313 .972 .039

Abbreviations: DMPA, depot medroxyprogesterone acetate; LARC, long-acting reversible contraception; RLPC, reproductive life plan counseling. * Controlling for age, insurance status, history of a recent birth, and the existence of living children who were born to the participant. y Nonmedical/no method users are the reference category.

effective contraceptive use does not vary by race/ethnicity or by educational attainment. Discussion This study analyzed patterns of contraceptive use to discern the effect of RLPC sessions on effective contraceptive use. Further, we assessed whether RLPC reduced race/ethnicity or educational attainment differences in the use of effective contraception. Finally, we evaluated whether the effect of RLPC varied by race/ethnicity or educational attainment. These results suggest that the provision of RLPC is not associated with using effective contraception. The women in this sample used medical contraception at a higher rate (78%) than women in national samples (65%; Dehlendorf et al., 2014b). Further, nearly one-quarter of women (24.2%) had selected a LARC method. This is more than double the national rate of LARC use (Finer, Jerman, & Kavanaugh, 2012). The high rate of LARC use found in this clinical population is likely a product of the structurally induced need for fertility control among low-income women receiving health care from public health institutions. Because this sample already had a high rate of medical contraception use, with high LARC use in particular, our test of the effectiveness of RLPC may have been conservative. If women who were most open to effective contraceptive use had already adopted a method, RLPC would have had to influence women who were least likely to adopt a medical method of contraception. It is possible that RLPC may be effective in a population with lower rates of medical contraceptive use. On the other hand, the results could mean RLPC is not an effective intervention to increase contraceptive use. It is possible that some providers do not buy into RLPC and its time demands, resulting in RLPC not being implemented in its ideal form. Other research has found that, when providers guide contraceptive counseling sessions, they often do not focus on patients’ concerns, preferences, and previous experiences (Dehlendorf et al., 2014a). It is also possible that provider characteristicsdincluding racial concordance with the patient or providers’ inclination to insert an IUD rather than administer a pill,

patch, or ringdmight affect the health care interaction and limit RLPC effectiveness. More information about patient–provider interactional dynamics during the RLPC session would allow for a better understanding of RLPC effectiveness. Future evaluation of RLPC should pair a large-scale medical records review, as we have done here, with observation of clinician training in RLPC and clinical practice. Such a study would allow for a deeper understanding of the effects of providers’ contraceptive preferences on RLPC provision. It would also help to elucidate variation in RLPC delivery across providers and between physicians, midlevel providers, and lay counselors. We hypothesized that RLPC would help to close the gap in effective contraceptive use between White women and women of color. However, given that we did not find the intervention effective, we did not observe that it narrowed the racial/ethnic gap in effective contraceptive use. We also speculated that RLPC may function differentially depending on educational attainment and race/ethnicity; however, the analyses did not support that proposition either. Past patterns suggest that efforts to increase service and technology use are often most effective among those with more social advantage (Phelan et al., 2010). Thus, interventions like RLPC, which are designed to reduce inequality, can actually increase it because of differential utilization. Additionally, owing to the negative relationship between Black women and population control efforts, it is possible that RLPC could have further exacerbated racial differences in use of effective contraceptive methods, particularly LARCs. Thus, not finding a differential effect of RLPC on contraceptive use could be considered positive news. A final unexpected finding is that a high school degree increased the odds of not using DMPA over nonmedical/no method. DMPA has been associated with numerous side effects subsequent to long-term use including loss of bone mineral density and excessive weight gain (American College of Obstetricians and Gynecologists, 2014). As such, it is possible that an aversion to these side effects is more salient among women with higher educational attainment. Additionally, provider bias toward recommending DMPA to lower socioeconomic status women may play a role as well. However, no study has

214

A. Bommaraju et al. / Women's Health Issues 25-3 (2015) 209–215

attempted to uncover mechanisms that might account for sociodemographic differences in DMPA use. Implications for Practice and/or Policy Our analysis suggests that RLPC provision did not encourage effective contraceptive use in this system of clinics, but further research is necessary before generalizable conclusions about RLPC efficacy can be developed. To evaluate systematically the efficacy of this intervention, reproductive health programs utilizing RLPC should standardize their RLPC training protocols and require detailed chart documentation of domains covered during each RLPC session. Furthermore, observations of RLPC sessions are also needed to evaluate whether RLPC quality is maintained across patient subgroups. Finally, because it is currently unclear if health professionals or lay counselors are better suited to deliver the RLPC session, comparative analyses of RLPC sessions are important for determining best practices. Limitations There are some limitations to this study. First, the data did not include clinical contraindications that may have influenced contraceptive choice. This limitation may have affected the conclusions because individuals in the nonmedical/no method group may be left with no medical contraceptive options because of contraindications, rather than because RLPC did not influence their decision. Second, the data did not include sociobehavioral determinants of contraceptive choice, such as patients’ sexual orientation, sexual activity, number of sexual partners, marital status, attitudes toward contraception, and pregnancy ambivalence. Such information could have been used to understand method use more fully. Third, the data were collected from women in one city and may not generalize to other locations. Cincinnati is a new destination city for Latinos, which is similar to other Midwest and Southern cities, but different from other cities with established Latino populations. Additionally, the RHWP population may also be nongeneralizable because of its significantly high rate of LARC use compared with national samples. Finally, our measure of RLPC is limited owing to its susceptibility to misclassification bias. It is possible that providers may have checked the RLPC box even when they had not performed the intervention because of time constraints in the busy clinical setting. This possibility would result in the benefits of RLPC being underestimated in this study. Despite these limitations, this study provides an important test of the effectiveness of RLPC. Conclusion With the implementation of the Affordable Care Act, and despite constitutional challenges to the contraceptive mandate, we are hopeful that many of the economic barriers to contraceptive access will be alleviated over time. Thus, interventions to resolve racial and educational disparities in effective contraceptive use must begin to focus on improving and evaluating behavioral interventions such as RLPC. Particular attention should be paid to how such programs can reduce disparities in contraceptive use. Acknowledgments Aalap Bommaraju had full access to all the data in the study and takes responsibility for the integrity of the data and the

accuracy of the data analysis. The authors thank the Cincinnati Health Department for its participation in this study, David Maume for his contributions to the data analysis, and our anonymous reviewers for their valuable suggestions to improve the quality of this work. References American College of Obstetricians and Gynecologists (ACOG). (2011). Longacting reversible contraception: Implants and intrauterine devices. Practice bulletin no. 121. Obstetrics & Gynecology, 118(1), 184–196. American College of Obstetricians and Gynecologists (ACOG). (2014). Depot medroxyprogesterone acetate and bone effects. Committee opinion no. 602. Obstetrics & Gynecology, 123(6), 1398–1402. Cardea. (2014). Title X orientation. Available: http://www.cardeaservices.org/ resourcecenter/title-x-orientation. Accessed December 1, 2014. Centers for Disease Control and Prevention (CDC). (2006). Recommendations to improve preconception health and health care – United States: A report of the CDC/ATSDR Preconception Care Work Group and the Select Panel on Preconception Care. Morbidity and Mortality Weekly Report, 55(RR-6), 1–23. Centers for Disease Control and Prevention (CDC). (2010). U.S. medical eligibility criteria for contraceptive use, 2010. Morbidity and Mortality Weekly Report, 59(RR-4), 1–85. Centers for Disease Control and Prevention (CDC). (2012). Preconception health and health care: My reproductive life plan. Available: http://www.cdc. gov/preconception/documents/reproductivelifeplan-worksheet.pdf. Accessed December 1, 2014. Dennis, A., & Grossman, D. (2012). Barriers to contraception and interest in overthe-counter access among low-income women: A qualitative study. Perspectives on Sexual and Reproductive Health, 44(2), 84–91. Dehlendorf, C., Foster, D. G., Thiel de Bocanegra, H., Brindis, C., Bradsberry, M., & Darney, P. (2011). Race, ethnicity, and differences in contraception among low-income women: Methods received by family PACT clients, California, 2001-2007. Perspectives on Sexual and Reproductive Health, 43(3), 181–187. Dehlendorf, C., Kimport, K., Levy, K., & Steinauer, J. (2014a). A qualitative analysis of approaches to contraceptive counseling. Perspectives on Sexual and Reproductive Health, 46(4), 233–240. Dehlendorf, C., Park, S. Y., Emeremni, C. A., Comer, D., Vincett, K., & Borrero, S. (2014b). Racial/ethnic disparities in contraceptive use: Variation by age and women’s reproductive experiences. American Journal of Obstetrics & Gynecology, 210(6), 526.e1–526.e9. Files, J. A., Frey, K. A., David, P. S., Hunt, K. S., Noble, B. N., & Mayer, A. P. (2011). Developing a reproductive life plan. Journal of Midwifery & Women’s Health, 56(5), 468–474. Finer, L. B., Jerman, J., & Kavanaugh, M. L. (2012). Changes in use of long-acting contraceptive methods in the United States, 2007–2009. Fertility and Sterility, 98(4), 893–897. Finer, L. B., & Zolna, M. R. (2011). Unintended pregnancy in the United States: Incidence and disparities, 2006. Contraception, 84(5), 478–485. Foster, D. G., Biggs, M. A., Rostovtseva, D., Thiel de Bocanegra, H., Darney, P. D., & Brindis, C. D. (2011). Estimating the fertility effect of expansions of publicly funded family planning services in California. Women’s Health Issues, 21(6), 418–424. Gavin, L., Moskosky, S., Carter, M., Curtis, K., Glass, E., Godfrey, E., . Zapata, L. (2014). Providing quality family planning services: Recommendations of CDC and the U.S. Office of Population Affairs. Morbidity and Mortality Weekly Report, 63(RR-4), 1–54. Hatcher, R. A., Trussell, J., Nelson, A. L., Cates, W., Kowal, D., & Policar, M. (2011). Contraceptive technology: Twentieth revised edition. New York: Ardent Media, Inc. IBM Corp. Released 2013. IBM SPSS statistics for Windows, version 22.0. Armonk, NY: IBM Corp. Jacobs, J., & Stanfors, M. (2013). Racial and ethnic differences in U.S. women’s choice of reversible contraceptives, 1995-2010. Perspectives on Sexual and Reproductive Health, 45(3), 139–147. Jones, J., Mosher, W., & Daniels, K. (2012). Current contraceptive use in the United States, 2006–2010, and changes in patterns of use since 1995. (National Health Statistics Reports No. 60). Hyattsville, MD: National Center for Health Statistics. Kavanaugh, M. L., Frohwirth, L., Jerman, J., Popkin, R., & Ethier, K. (2013). Longacting reversible contraception for adolescents and young adults: Patient and provider perspectives. Journal of Pediatric & Adolescent Gynecology, 26(2), 86–95. Medoff, M. H. (2012). Restrictive abortion laws, antiabortion attitudes and women’s contraceptive use. Social Science Research, 41(1), 160–169. Mood, C. (2010). Logistic regression: Why we cannot do what we think we can do, and what we can do about it. European Sociological Review, 26(1), 67–82. Mosher, W. D., Jones, J., & Abma, J. C. (2012). Intended and unintended births in the United States: 1982–2010. (National Health Statistics Reports No. 55). Hyattsville, MD: National Center for Health Statistics. Noyes, S. S., & Savin, M. K. (2012). My life, my plan: Delaware’s reproductive life planning for teens. Journal of Obstetric, Gynecologic, & Neonatal Nursing, 41(s1), S3.

A. Bommaraju et al. / Women's Health Issues 25-3 (2015) 209–215 Phelan, J. C., Link, B. G., & Tehranifar, P. (2010). Social conditions as fundamental causes of health inequalities: Theory, evidence, and policy implications. Journal of Health and Social Behavior, 51(S), S28–S40. Roberts, D. (1997). Killing the Black body: Race, reproduction and the meaning of liberty. New York: Pantheon Books. Sangi-Haghpeykar, H., Ali, N., Posner, S., & Poindexter, A. N. (2006). Disparities in contraceptive knowledge, attitude and use between Hispanic and nonHispanic whites. Contraception, 74(2), 125–132. Secura, G. M., Madden, T., McNicholas, C., Mullersman, J., Buckel, C. M., Zhao, Q., . Peipert, J. E. (2014). Provision of no-cost, long-acting contraception and teenage pregnancy. New England Journal of Medicine, 371(14), 1316–1323. Shim, J. K. (2010). Cultural health capital: A theoretical approach to understanding health care interaction and the dynamics of unequal treatment. Journal of Health and Social Behavior, 51(1), 1–15. Singh, S., Sedgh, G., & Hussain, R. (2010). Unintended pregnancy: Worldwide levels, trends, and outcomes. Studies in Family Planning, 41(4), 241–250. Steiner, M. J., Trussell, J., Mehta, N., Condon, S., Subramaniam, S., & Bourne, D. (2006). Communicating contraceptive effectiveness: A randomized controlled trial to inform a world health organization family planning handbook. American Journal of Obstetrics and Gynecology, 195(1), 85–91. Teal, S. B., & Romer, E. S. (2013). Awareness of long-acting reversible contraception among teens and young adults. Journal of Adolescent Health, 52(4 Suppl), S35–S39.

215

Trussell, J., Hassan, F., Lowin, J., Law, A., & Filonenko, A. (2015). Achieving costneutrality with long-acting reversible contraceptive methods. Contraception, 91, 49–56. van Ryn, M., Burgess, D. J., Dovidio, J. F., Phelan, S. M., Saha, S., Malat, J., . Perry, S. (2011). The impact of racism on clinician cognition, behavior, and clinical decision-making. Du Bois Review, 8(1), 199–218.

Author Descriptions Aalap Bommaraju, MPH, specializes in the evaluation of public health interventions targeted at improving the provision of reproductive health services to vulnerable populations.

Jennifer Malat, PhD, has primarily focused her research on how race affects doctor–patient encounters in ways that reduce health care quality. Her recent research has also examined how social forces affect the health of whites, blacks, and Latinos.

Jennifer L. Mooney, PhD, has extensive experience in the implementation of preventive reproductive health programs. She also has a long history of research expertise in the areas of substance abuse, crime, gender, and race.