Contraception 90 (2014) 97 – 103
Original research article
Associations of mental illness and substance use disorders with prescription contraception use among women veterans☆,☆☆,★ Lisa S. Callegari a, b,⁎, Xinhua Zhao c , Karin M. Nelson a, d , Keren Lehavot e, f , Katharine A. Bradley a, d, g , Sonya Borrero c, h a
Health Services Research and Development (HSR&D), Department of Veterans Affairs (VA) Puget Sound Health Care System, Seattle, WA b Department of Obstetrics & Gynecology, University of Washington, Seattle, WA c Center for Health Equity, Research, and Promotion, VA Pittsburgh Health Care System, Pittsburgh, PA, Seattle, WA d Department of Medicine, University of Washington, Seattle, WA e Mental Illness Research, Education and Clinic Center (MIRECC), VA Puget Sound Health Care System, Seattle, WA f Department of Psychiatry and Behavioral Sciences, University of Washington, Seattle WA g Group Health Research Institute, Seattle, WA h Department of Medicine, University of Pittsburgh, Pittsburgh, PA Received 11 November 2013; revised 25 February 2014; accepted 27 February 2014
Abstract Objective: To investigate whether mental illness and substance use disorder (SUD) are associated with having a prescription contraceptive method among women veterans. Study design: We conducted a retrospective analysis of National Veterans Administration (VA) administrative and clinical data for women veterans aged 18–45 years who made at least one primary care visit in 2008. We assessed associations between mental illness (depression, posttraumatic stress disorder, anxiety, bipolar disorder, schizophrenia and adjustment disorder) and SUD (drug/alcohol use disorder) with having a prescription contraceptive method from VA (pill, patch, ring, injection, implant and intrauterine device) using multivariable logistic regression with random effects for VA facility, adjusting for confounders. Results: Among 94,115 reproductive aged women, 36.5% had mental illness only, 0.6% had SUD only, 5.3% had both mental illness and SUD and 57.7% had neither diagnosis. In these groups, 22.1%, 14.6%, 18.2% and 17.7% (pb0.001), respectively, had documentation in 2008 of prescription contraception. After adjusting for potential confounders, women with mental illness only were as likely as women with neither diagnosis to have a prescription method and were more likely to use a highly effective prescription method (implant or intrauterine device) if using contraception [adjusted odds ratio (aOR) 1.17, 95% confidence interval (CI) = 1.08–1.27]. Women with SUD (with or without mental illness) were significantly less likely to have a prescription method than women with neither diagnosis (aOR 0.73, 95% CI = 0.57–0.95 and aOR 0.79, 95% CI = 0.73–0.86, respectively). Conclusion: Women veterans with SUD are less likely to have prescription contraception compared to other women, which may increase their risk of unintended pregnancy. Published by Elsevier Inc. Keywords: Contraceptive behavior; Veterans health; Mental disorder; Alcohol-related disorders; Substance-related disorders; Highly effective reversible contraception
☆
This study was supported by the Department of Veterans Affairs, Veteran Health Administration, Office of Research and Development, VISN 4 CHERP Pilot Project Award (PI: Sonya Borrero). L.S.C. was supported by a VA Health Services Research and Development Postdoctoral Fellowship (TPM 61-041). K.L. was supported by the VA Office of Academic Affiliations, Advanced Fellowship Program in Mental Illness Research and Treatment. ☆☆ The findings and conclusions in this report are those of the authors and do not represent the views of the Department of Veterans Affairs or the United States Government. ★ The authors have no conflicts of interest to disclose. ⁎ Corresponding author. Lisa S. Callegari, Seattle HSR&D Center, Puget Sound Veterans Administration University of Washington 1100 Olive Way, Suite 1400 Seattle, WA 98101. Tel.: +1 206 277 3129; fax: +1 206 768 5343. E-mail address: [email protected] (L.S. Callegari). http://dx.doi.org/10.1016/j.contraception.2014.02.028 0010-7824/Published by Elsevier Inc.
98
L.S. Callegari et al. / Contraception 90 (2014) 97–103
1. Introduction Over half (51%) of pregnancies in the United States are unintended, nearly all of which result from a combination of contraceptive nonuse and inconsistent or incorrect use of contraception [1–3]. Established risk factors for contraceptive nonuse and incorrect use include demographic, socioeconomic and reproductive characteristics [4–6]. Emerging data suggest that common mental illnesses such as depression or anxiety and substance use disorders (SUD) may also be important risk factors for poor contraceptive use [7–12]. Inadequately controlled mental illness and SUD also increase risks of adverse perinatal outcomes such as preterm birth, low birth weight and birth defects as well as adverse maternal outcomes such as postpartum depression [13–18]. Effective contraception to prevent unintended pregnancy in this population is therefore particularly critical. A growing body of evidence suggests that mental illness is associated with contraceptive nonuse, use of less effective methods and reduced adherence with compliance-based methods such as the birth control pill [7–10], although other studies have not found these associations [19,20]. Little data have been published to date on associations between diagnosed SUD and contraceptive use in adults [12,21]. Published studies in teenagers and college students have focused on associations between substance use or binge drinking and condom behavior, rather than associations between substance use or SUD and use of effective contraception [12,22–25]. Moreover, prior studies have not used national data to examine relationships between mental illness and SUD with contraceptive use. Women veterans have significantly higher rates of mental illness compared to their civilian counterparts [26,27]. A recent study found that 48% of women veterans using the VA health care system users who served in Afghanistan in Operation Enduring Freedom (OEF) or in Iraq in Operation Iraqi Freedom (OIF) screened positive for depression and 21% for posttraumatic stress disorder (PTSD) [26]. The prevalence of alcohol misuse among women veterans is also high, with 17% screening positive using the Alcohol Use Disorders Identification Test (AUDIT-C) in a national sample of OEF/OIF Veterans [28]. In addition to high rates of mental illness and SUD, women veterans have low rates of prescribed contraception compared to national estimates [29]. Given the rapidly increasing number of young women veterans entering the VA health care system [30], understanding determinants of contraceptive nonuse or use of less effective, nonprescription methods is critical for prevention of unintended pregnancies and improving birth outcomes. However, no studies have yet examined the associations between presence of mental illness or SUD and having a prescription contraceptive method in the VA. We therefore used national VA data to investigate the association of mental illness (depression, PTSD, anxiety, bipolar disorder, schizophrenia and adjustment disorder) and SUD (drug/alcohol use disorders) with having a prescription
contraceptive method [pill, patch, ring, injection, implant and intrauterine device (IUD)]. We also investigated associations between mental illness or SUD and specific types of prescription contraception categorized by level of effectiveness. We hypothesized that women with mental illness or SUD only, as well as women with both diagnoses, would be less likely to have prescription contraception and less likely to use highly effective prescription contraceptive methods (implant or IUD) compared to women without these diagnoses. 2. Materials and methods 2.1. Data sources We used secondary data from national VA administrative and clinical databases including the Pharmacy Benefits Management (PBM) Database (Version 3) and the Medical SAS inpatient and outpatient data files for fiscal year (FY) 2008 (October 1, 2007 through Sept 30, 2008). The VA uses an electronic medical record system that captures pharmacy and clinical data on all clinical care provided within VA facilities. The PBM Database was used to obtain information on prescribed contraceptive methods. The Medical SAS Datasets contain VA inpatient and outpatient clinical data and were used to obtain demographic measures, mental health and medical International Classification of Diseases, Ninth Revision, Clinical Modification (ICD-9-CM) codes, health care utilization and Current Procedural Terminology (CPT) codes for contraceptive procedures including insertions or removals of IUDs or subdermal contraceptive implants. This study was approved by the VA Pittsburgh and VA Puget Sound Institutional Review Boards. 2.2. Study setting and population The study population included all female veterans aged 18–45 who made at least one visit to a standard primary care clinic or designated women's health primary care clinic during FY2008. Women were excluded if they had evidence of prior hysterectomy, tubal sterilization, infertility or evidence of pregnancy during the study timeframe. 2.3. Outcome variables The outcome variables for the analyses were (1) any prescription contraceptive method based on pharmacy records or CPT codes for contraception-related procedures and (2) type of prescription contraceptive method used, categorized by level of effectiveness, in FY2008. A woman was considered to have a documented method of contraception if she had filled a prescription for a hormonal contraceptive at any time in FY2008 or if she had evidence of having received a contraceptive implant in the preceding 5 years or IUD in the preceding 10 years and no evidence of removal. Contraceptive methods were categorized using the World Health Organization's tiers of contraceptive effectiveness. Moderately effective contraception was defined as birth control pill, patch, ring or injection; highly effective
L.S. Callegari et al. / Contraception 90 (2014) 97–103
99
contraception was defined as contraceptive implant or IUD [31]. A previous analysis demonstrated that VA administrative data are less reliable in capturing nonprescription methods such as condoms, other over-the-counter products and behavioral methods such as fertility awareness [29]. Therefore, we focused on those prescription methods that are more reliably captured in VA data through pharmacy prescriptions and CPT codes.
association between mental illness and SUD diagnoses with effectiveness of the method used, we performed logistic regression with random effects among those women veterans with a prescription method of contraception, again adjusting for all covariates. Statistical analyses were conducted using SAS 9.2 (Cary, NC, USA) and Stata 11 (College Station, TX, USA).
2.4. Independent variables
3. Results
The primary predictor of interest was mental illness or SUD at any point during the study period or the FY preceding the study. Mental illness and SUD were identified based on presence of two or more outpatient or one or more inpatient ICD-9-CM codes for psychiatric or mental health diagnoses in FY2007 and FY2008 [32]. Specific mental health disorders identified using ICD-9-CM codes included depression (296.20–296.25, 296.30–296.35, 300.4, 311), PTSD (309.81), anxiety (300.00–300.09, 300.20–300.29, 300.3), bipolar disorder (296.0, 296.4–296.8), schizophrenia (295.0–295.9) and adjustment disorders (309.0–309.9, excluding 309.81). SUD identified included alcohol use disorders (303.0–303.99, 305.00–305.03) and drug use disorders (292.01–292.99, 304.0–304.99, 305.2–305.99). Additional covariates included in the analysis were age, race/ethnicity, marital status, individual annual income, religion, presence of non-VA insurance, geographic region, OEF/OIF Veteran or not and medical comorbidities which could affect contraceptive choice (hypertension, history of venous thromboembolism, coronary artery disease, breast cancer, stroke, tobacco use among women age 35 or older, diabetes with complications, liver disease and migraines with aura). We also examined health care utilization variables including whether primary care was received at a standard primary care clinic or a women's health clinic, the number of visits to a primary care clinic, whether the patient was seen in a hospital-based versus communitybased outpatient clinic and whether or not the patient was seen in a gynecology clinic. 2.5. Statistical analysis We compared demographic and clinical characteristics of our sample by presence or absence of mental illness or SUD diagnoses (mental illness only, SUD only, both SUD and mental illness diagnoses and neither mental illness nor SUD diagnoses). We also compared having a prescription contraceptive method and specific method types by presence or absence of mental illness or SUD. Chi-squared tests were used to compare variables, all of which were coded as categorical variables. To test whether mental illness or SUD diagnoses were associated with having a prescription contraceptive method, we performed logistic regression using random effects to account for nonindependence of outcomes among women seen at the same VA facility and adjusting for all covariates. To test whether there was an
A total of 103,950 women veterans aged 18–45 made at least one visit to a primary care clinic or women's health clinic in FY2008. We excluded 3846 (3.7%) women who had evidence of prior hysterectomy or tubal sterilization at VA and 1928 (1.9%) women with infertility diagnoses. In addition, we excluded 4061 (3.9%) women who had evidence of pregnancy during FY2008, resulting in a final cohort of 94,115 women. Sample demographic and clinical characteristics are shown in Table 1. In our sample, 36.5% had mental illness only, 0.6% had SUD only, 5.3% had both diagnoses and 57.7% had neither diagnosis. Among women with SUD with or without mental illness, 38.4% had an alcohol use disorder diagnosis only, 28.5% had a drug abuse diagnosis only and 33.1% had both (data not shown). Among women with mental illness only, the most common ICD-9-CM codes included depression (71.1%), PTSD (32.0%), anxiety (29.8%) and bipolar disorder (12.0%). Among women with both mental illness and SUD, 75.9% had an ICD-9-CM code for depression, 48.4% for PTSD, 35.2% for anxiety and 32.5% for bipolar disorder. Overall, 19.3% of the study sample had documentation in FY2008 of having a contraceptive method: 22.1% of women with mental illness only, 14.6% of women with SUD only, 18.2% of women with both diagnoses and 17.7% of women with neither diagnosis (Table 2). More women with mental illness only or both mental illness and SUD used a highly effective method compared to women with neither diagnosis in unadjusted analyses. Table 3 presents adjusted associations of the four mental illness and SUD categories with having prescription contraception and use of highly effective contraception. The first model examines associations with the outcome of any prescription contraceptive use among our entire sample (N= 94,115). Women with mental illness only did not differ significantly from women with neither diagnosis [adjusted odds ratio (aOR) = 1.04, 95% confidence interval (CI) = 1.00–1.08, p=0.06], although a nonsignificant trend was observed. Women with SUD only had 27% lower odds of having prescription contraception (aOR = 0.73, 95% CI = 0.57–0.95, p=0.02), and women with both diagnoses had 21% lower odds of having prescription contraception (aOR = 0.79, 95% CI = 0.73–0.86, pb0.001), compared to women with neither diagnosis. The second model examines associations with use of a highly effective method among
100
L.S. Callegari et al. / Contraception 90 (2014) 97–103
Table 1 Demographic and clinical characteristics of the study sample by presence of mental illness and substance use disorder (SUD), N=94,115
Table 1 (continued) Mental illness & SUD Neither Mental SUD only Both illness only N= 542 N= 4,956 N=54,259 N=34,358 (0.58%) % (5.3%) % (57.7%) % (36.5%) %
Mental illness & SUD Mental SUD only Both Neither illness only N=542 N= 4,956 N= 54,259 N=34,358 (0.58%) % (5.3%) % (57.7%) % (36.5%) % Demographics Age 18–25 26–35 36–45 Race/Ethnicity Hispanic Non-Hispanic White Non-Hispanic Black Other Unknown/Missing Marital status Married Div/sep/Widowed Never married Unknown Adjusted annual income ≤ US$20K Has non-VA insurance Religion Protestant Catholic Other Unaffiliated Primary care clinic type Standard primary care clinic only Women's health clinic only Both Number of visits 1–2 3–5 6+ Seen in gynecology clinic Hospital-based clinic Geographic region a Midwest Northeast South West OEF/OIF Medical comorbidities b Types of mental health diagnoses Depression PTSD Anxiety Bipolar Schizophrenia Adjustment disorder Count of types of mental health diagnoses 0 1 2
12.9 37.6 49.5
11.1 25.5 63.5
10.1 29.4 60.5
17.6 38.4 44.0
6.4 47.2 22.5 2.9 20.9
4.6 42.8 31.0 3.3 18.3
5.2 54.4 21.9 2.8 15.6
6.3 34.3 25.1 2.9 31.4
33.5 28.6 36.9 1.0 37.0
16.1 34.1 49.4 0.4 23.8
20.4 36.2 42.9 0.5 32.6
33.3 22.5 36.9 7.3 36.5
33.2
17.9
26.3
29.2
47.9 17.8 10.2 24.1
53.3 17.2 9.6 19.9
47.6 18.8 10.6 23.0
45.1 16.1 8.9 29.8
52.1
49.3
46.2
60.2
11.9
16.2
12.4
14.3
36.0
34.5
41.4
25.6
39.1 38.5 22.3 20.8
47.0 36.0 17.0 19.7
33.3 36.4 30.3 21.5
63.5 28.2 8.2 16.5
92.0
94.5
93.5
88.8
17.8 10.8 49.8 21.6 71.9 25.1
19.0 10.0 47.2 23.8 58.1 43.4
21.4 13.6 41.5 23.4 63.0 41.7
16.4 9.4 52.8 21.5 74.7 17.2
71.1 32.0 29.8 12.0 3.0 13.4
0.0 0.0 0.0 0.0 0.0 0.0
75.9 48.4 35.2 32.5 8.0 13.3
0.0 0.0 0.0 0.0 0.0 0.0
0.0 54.2 32.6
100.0 0.0 0.0
0.0 29.9 38.4
100.0 0.0 0.0
3 4+ Drug use disorder Alcohol use disorder
11.0 2.2 0.0 0.0
0.0 0.0 54.1 68.5
22.0 9.7 62.4 71.8
0.0 0.0 0.0 0.0
Chi-squared tests were used to test group differences; all variables differed significantly with pb0.001 by mental illness and SUD status except adjustment disorder (p=0.86) and alcohol use disorder (p=0.10). a Geographic region is based on US Census regions. b Comorbidities include medical conditions that could impact the use of estrogen-containing methods: hypertension, history of venous thromboembolic disease, coronary artery disease, breast cancer, stroke, tobacco use among women aged 35 years old or older, diabetes with complications, liver disease and migraines with aura.
the subset of women veterans with a contraceptive method (N= 18,115). Compared to women with neither diagnosis, women with mental illness only had 17% higher odds of using a highly effective method (aOR = 1.17, 95% CI = 1.08–1.27, pb0.001) and women with both mental illness and SUD had 30% higher odds of using a highly effective method (aOR = 1.30, 95% CI = 1.10–1.54, p=0.002).
4. Discussion Little is known about the associations between mental illness, SUD and contraceptive use in adult women. In this cross-sectional analysis of over 90,000 women veterans, women with SUD either alone or in combination with mental illness were significantly less likely to have a prescription contraceptive method compared to women with neither mental illness nor SUD. Women with mental illness only, in contrast, were as likely as women with neither diagnosis to have a prescription contraceptive method. Among women veterans using contraception, those with mental illness were more likely to use a highly effective method (IUD or implant) compared to those with neither diagnosis. While no previous studies to our knowledge examine the relationship between diagnosed SUD and contraceptive use among adults, our finding that women with SUD were less likely to have a prescription method of contraception during the study period is generally consistent with prior literature linking use of alcohol and drugs to high-risk sexual behaviors such as inconsistent condom use in teens and young adults [12,23–25]. Several older studies that included associations between alcohol use and birth control in teens and young adults, however, reached conflicting conclusions [12,21]. Our findings provide preliminary evidence that women with diagnosed SUD may be at increased risk of unintended pregnancy due to nonuse of contraception or reliance on less effective, nonprescription methods.
L.S. Callegari et al. / Contraception 90 (2014) 97–103 Table 2 Contraceptive method type among women veterans with and without mental illness and SUD, N= 94,115 Mental illness & SUD Mental SUD only % Both % Neither % illness only % Any effective method 22.1 Highly effective 5.2 IUD 5.1 Implant 0.1 Moderately effective a 16.8 OCP 12.5 Patch 0.4 Ring 1.2 Injection 3.2
14.6 4.1 3.9 0.2 10.5 6.3 0.2 0.6 3.7
18.2 4.9 4.8 0.1 13.2 8.8 0.2 1.1 3.9
17.7 3.6 3.5 0.1 14.0 10.8 0.3 1.0 2.2
OCP=oral contraceptive pill. Chi-squared tests were used to test group differences; all variables differed significantly with pb0.001 by mental illness and SUD status. a Five hundred three (0.53%) women used more than one contraceptive method during the study period.
Previous studies have linked psychological symptoms of depression and stress to inconsistent contraceptive use or nonuse [10,11], discontinuation of contraception [9] and use of less effective methods [7,8] and hypothesized that these symptoms may predispose women to riskier contraceptive behavior. Two recent analyses, however, did not detect these associations [19,20], suggesting that there is variation by context and patient population. The first study among postpartum women in Brazil found no association between perinatal psychiatric disorders and subsequent use of less effective contraceptive methods postpartum [19]; the second found that low-income women with more psychological distress prior to an abortion were more likely to choose more effective rather than less effective postabortion methods [20]. In our sample, women veterans with mental illness diagnoses were as likely to have a prescription contraceptive method and more likely to use a highly effective method than women veterans without these diagnoses. Our study differs from previous analyses in that we defined mental illness using diagnosis codes in the year prior to or during the study period, rather than using current symptoms. In addition, our study occurred in the unique setting of the VA health care system, where the emphasis on identifying and treating mental illness and close integration of mental health services into primary care may mitigate any effects of mental illness on contraceptive use. Although women with mental illness were slightly more likely to have non-VA insurance compared to women without mental illness, women with mental health needs may still preferentially use the VA system for contraception rather than seeking care outside the VA [25], which could bias towards higher rates of documented contraception. Among women veterans who had a documented method of contraception from VA, those with mental illness, with or without coexisting SUD, were more likely to use a highly effective, compliance-free method compared to those
101
without these diagnoses. These findings are encouraging in light of a recent study by Borrero et al. which found that adherence to compliance-based methods among women veterans using VA is low [33]. One possible explanation for our findings is that providers in VA may encourage use of compliance-free methods such as IUDs (rates of implant use were low) differentially in women with mental illness or SUD, as these conditions have been linked to lower compliance with health maintenance [34]. We also cannot exclude the possibility that residual confounding influenced our findings. While we adjusted for marital status, we were unable to adjust for other reproductive factors associated with use of highly effective methods such as parity, age at first sex and number of partners [35]. Because our data are cross-sectional, we could not determine whether mental illness or SUD occurred before or after the initiation of contraceptive methods. We are therefore only able to evaluate associations between mental illness and SUD with contraceptive use, rather than infer causal relationships. Additional longitudinal studies in VA and other settings are needed to further assess these relationships and their variation by context and patient population. Prospective designs are also needed to investigate potential mechanisms for observed associations. Additional limitations of our study include that we were not able to determine with certainty from the available clinical data whether women were at risk of unintended pregnancy. For example, while we were able to exclude those women with documented sterilization, hysterectomy or infertility in the VA, we may have included women in our sample with surgery or treatment outside VA. In addition, we could not exclude women in same-sex relationships or women not sexually active during the study period, as this information is not available in VA clinical data. Rates of mental illness and SUD are higher among women veterans in
Table 3 Adjusted comparisons of prescription contraceptive use by mental illness and SUD among women veterans Any prescription contraceptive method a N= 94,1150
Highly effective method a,b N= 18,115
aOR (95% CI) p value aOR (95% CI) p value Mental illness & SUD diagnoses None Reference Mental illness only 1.04 (1.00–1.08) SUD only 0.73 (0.57–0.95) Mental illness +SUD 0.79 (0.73–0.86)
Reference 1.17 (1.08–1.27) 0.02 1.26 (0.75–2.13) b0.001 1.30 (1.10–1.54) 0.06
b0.001 0.39 0.002
aOR = adjusted odds ratio. a Adjusted for age, race/ethnicity, marital status, income, religion, nonVA insurance, clinic type and number of visits, gynecologist visit, geography, military conflict and medical comorbidities. b Among the subset of women veterans with a prescription contraceptive method.
102
L.S. Callegari et al. / Contraception 90 (2014) 97–103
sexual minority relationships [36], and it is unclear how our results would be altered if these women were excluded from our analyses. We were also unable to exclude women who were trying to get pregnant; however, studies using national data have demonstrated that at any given time less than 5% of reproductive-aged women are seeking to become pregnant [37]. Lastly, our results may not be generalizable to women veterans who do not use VA or to civilian women. In summary, this study of women Veterans found that mental illness and SUD diagnoses are associated with having a prescription method of contraception in the VA healthcare system. Women with SUD appear to be least likely to have a contraceptive method, placing them at elevated risk of unplanned pregnancy. This finding is particularly concerning given the substantially increased risk of both fetal and maternal morbidity among women with SUD. Prospective research is needed to confirm these associations. If confirmed, preventive interventions to increase uptake of effective contraceptive methods will be critical among women veterans and other populations at high risk of SUD.
[15]
[16]
[17]
[18]
[19]
[20]
[21]
[22]
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): S43–8. [2] Trussell J. Contraceptive failure in the United States. Contraception 2011;83:397–404. [3] Jones RK, Darroch JE, Henshaw SK. Contraceptive use among U.S. women having abortions in 2000-2001. Perspect Sex Reprod Health 2002;34:294–303. [4] Frost JJ, Singh S, Finer LB. Factors associated with contraceptive use and nonuse, United States, 2004. Perspect Sex Reprod Health 2007;39:90–9. [5] Wu J, Meldrum S, Dozier A, Stanwood N, Fiscella K. Contraceptive nonuse among US women at risk for unplanned pregnancy. Contraception 2008;78:284–9. [6] Frost JJ, Singh S, Finer LB. U.S. women's one-year contraceptive use patterns, 2004. Perspect Sex Reprod Health 2007;39:48–55. [7] Hall KS, Moreau C, Trussell J, Barber J. Role of young women's depression and stress symptoms in their weekly use and nonuse of contraceptive methods. J Adolesc Health 2013;53:241–8. [8] Garbers S, Correa N, Tobier N, Blust S, Chiasson MA. Association between symptoms of depression and contraceptive method choices among low-income women at urban reproductive health centers. Matern Child Health J 2010;14:102–9. [9] Hall KS, White KO, Rickert VI, Reame N, Westhoff C. Influence of depressed mood and psychological stress symptoms on perceived oral contraceptive side effects and discontinuation in young minority women. Contraception 2012;86:518–25. [10] Berenson AB, Breitkopf CR, Wu ZH. Reproductive correlates of depressive symptoms among low-income minority women. Obstet Gynecol 2003;102:1310–7. [11] Stidham Hall K, Moreau C, Trussell J, Barber J. Young women's consistency of contraceptive use - does depression or stress matter? Contraception 2013;88:641–9. [12] Cooper ML. Alcohol use and risky sexual behavior among college students and youth: evaluating the evidence. J Stud Alcohol Suppl 2002:101–17. [13] Dunkel Schetter C, Tanner L. Anxiety, depression and stress in pregnancy: implications for mothers, children, research, and practice. Curr Opin Psychiatry 2012;25:141–8. [14] Grote NK, Bridge JA, Gavin AR, Melville JL, Iyengar S, Katon WJ. A meta-analysis of depression during pregnancy and the risk of preterm
[23] [24]
[25]
[26]
[27]
[28]
[29]
[30]
[31]
[32]
[33]
[34]
birth, low birth weight, and intrauterine growth restriction. Arch Gen Psychiatry 2010;67:1012–24. Seng JS, Low LK, Sperlich M, Ronis DL, Liberzon I. Post-traumatic stress disorder, child abuse history, birthweight and gestational age: a prospective cohort study. BJOG 2011;118:1329–39. Rogal SS, Poschman K, Belanger K, et al. Effects of posttraumatic stress disorder on pregnancy outcomes. J Affect Disord 2007;102: 137–43. Floyd RL, O'Connor MJ, Sokol RJ, Bertrand J, Cordero JF. Recognition and prevention of fetal alcohol syndrome. Obstet Gynecol 2005;106:1059–64. Witt WP, Wisk LE, Cheng ER, et al. Poor prepregnancy and antepartum mental health predicts postpartum mental health problems among US women: a nationally representative population-based study. Womens Health Issues 2011;21:304–13. Faisal-Cury A, Menezes PR, Huang H. The relationship between perinatal psychiatric disorders and contraception use among postpartum women. Contraception 2013;88:498–502. Steinberg JR, Tschann JM, Henderson JT, Drey EA, Steinauer JE, Harper CC. Psychological distress and post-abortion contraceptive method effectiveness level chosen at an urban clinic. Contraception 2013;88:717–24. Santelli J, Carter M, Orr M, Dittus P. Trends in sexual risk behaviors, by nonsexual risk behavior involvement, U.S. high school students, 1991-2007. J Adolesc Health 2009;44:372–9. Scott-Sheldon LA, Carey MP, Vanable PA, Senn TE, CouryDoniger P, Urban MA. Alcohol consumption, drug use, and condom use among STD clinic patients. J Stud Alcohol Drugs 2009;70:762–70. Leigh BC. Alcohol and condom use: a meta-analysis of event-level studies. Sex Transm Dis 2002;29:476–82. Peipert JF, Zhao Q, Meints L, Peipert BJ, Redding CA, Allsworth JE. Adherence to dual-method contraceptive use. Contraception 2011;84:252–8. Halpern-Felsher BL, Millstein SG, Ellen JM. Relationship of alcohol use and risky sexual behavior: a review and analysis of findings. J Adolesc Health 1996;19:331–6. Haskell SG, Gordon KS, Mattocks K, et al. Gender differences in rates of depression, PTSD, pain, obesity, and military sexual trauma among Connecticut War Veterans of Iraq and Afghanistan. J Womens Health (Larchmt) 2010;19:267–71. Lehavot K, Hoerster KD, Nelson KM, Jakupcak M, Simpson TL. Health indicators for military, veteran, and civilian women. Am J Prev Med 2012;42:473–80. Eisen SV, Schultz MR, Vogt D, et al. Mental and physical health status and alcohol and drug use following return from deployment to Iraq or Afghanistan. Am J Public Health 2012;102(Suppl 1):S66–73. Borrero S, Mor MK, Zhao X, McNeil M, Ibrahim S, Hayes P. Contraceptive care in the VA health care system. Contraception 2012;85(6):580–8. Frayne SM, Phibbs CS, Friedman SA, Berg EA, Ananth L, Iqbal S, et al. Sourcebook: women veterans in the Veterans Health Administration. Volume 1. Sociodemographic characteristics and use of VHA care. Washington, DC: Veterans Health Administration; 2010. Steiner MJ, Trussell J, Mehta N, Condon S, Subramaniam S, Bourne D. Communicating contraceptive effectiveness: a randomized controlled trial to inform a World Health Organization family planning handbook. Am J Obstet Gynecol 2006;195:85–91. Mattocks KM, Skanderson M, Goulet JL, et al. Pregnancy and mental health among women veterans returning from Iraq and Afghanistan. J Womens Health (Larchmt) 2010;19:2159–66. Borrero S, Zhao X, Mor MK, Schwarz EB, Good CB, Gellad WF. Adherence to hormonal contraception among women veterans: differences by race/ethnicity and contraceptive supply. Am J Obstet Gynecol 2013;209:103e1–11. DiMatteo MR, Lepper HS, Croghan TW. Depression is a risk factor for noncompliance with medical treatment: meta-analysis of the effects of
L.S. Callegari et al. / Contraception 90 (2014) 97–103 anxiety and depression on patient adherence. Arch Intern Med 2000;160:2101–7. [35] Kavanaugh ML, Jerman J, Hubacher D, Kost K, Finer LB. Characteristics of women in the United States who use long-acting reversible contraceptive methods. Obstet Gynecol 2011;117:1349–57.
103
[36] Mattocks KM, Sadler A, Yano EM, et al. Sexual victimization, health status, and VA healthcare utilization among lesbian and bisexual OEF/ OIF veterans. J Gen Intern Med 2013;28(Suppl 2):S604–8. [37] Mosher WD, Jones J. Use of contraception in the United States: 19822008. Vital Health Stat 23 2010:1–4.
Original research article
Associations of mental illness and substance use disorders with prescription contraception use among women veterans☆,☆☆,★ Lisa S. Callegari a, b,⁎, Xinhua Zhao c , Karin M. Nelson a, d , Keren Lehavot e, f , Katharine A. Bradley a, d, g , Sonya Borrero c, h a
Health Services Research and Development (HSR&D), Department of Veterans Affairs (VA) Puget Sound Health Care System, Seattle, WA b Department of Obstetrics & Gynecology, University of Washington, Seattle, WA c Center for Health Equity, Research, and Promotion, VA Pittsburgh Health Care System, Pittsburgh, PA, Seattle, WA d Department of Medicine, University of Washington, Seattle, WA e Mental Illness Research, Education and Clinic Center (MIRECC), VA Puget Sound Health Care System, Seattle, WA f Department of Psychiatry and Behavioral Sciences, University of Washington, Seattle WA g Group Health Research Institute, Seattle, WA h Department of Medicine, University of Pittsburgh, Pittsburgh, PA Received 11 November 2013; revised 25 February 2014; accepted 27 February 2014
Abstract Objective: To investigate whether mental illness and substance use disorder (SUD) are associated with having a prescription contraceptive method among women veterans. Study design: We conducted a retrospective analysis of National Veterans Administration (VA) administrative and clinical data for women veterans aged 18–45 years who made at least one primary care visit in 2008. We assessed associations between mental illness (depression, posttraumatic stress disorder, anxiety, bipolar disorder, schizophrenia and adjustment disorder) and SUD (drug/alcohol use disorder) with having a prescription contraceptive method from VA (pill, patch, ring, injection, implant and intrauterine device) using multivariable logistic regression with random effects for VA facility, adjusting for confounders. Results: Among 94,115 reproductive aged women, 36.5% had mental illness only, 0.6% had SUD only, 5.3% had both mental illness and SUD and 57.7% had neither diagnosis. In these groups, 22.1%, 14.6%, 18.2% and 17.7% (pb0.001), respectively, had documentation in 2008 of prescription contraception. After adjusting for potential confounders, women with mental illness only were as likely as women with neither diagnosis to have a prescription method and were more likely to use a highly effective prescription method (implant or intrauterine device) if using contraception [adjusted odds ratio (aOR) 1.17, 95% confidence interval (CI) = 1.08–1.27]. Women with SUD (with or without mental illness) were significantly less likely to have a prescription method than women with neither diagnosis (aOR 0.73, 95% CI = 0.57–0.95 and aOR 0.79, 95% CI = 0.73–0.86, respectively). Conclusion: Women veterans with SUD are less likely to have prescription contraception compared to other women, which may increase their risk of unintended pregnancy. Published by Elsevier Inc. Keywords: Contraceptive behavior; Veterans health; Mental disorder; Alcohol-related disorders; Substance-related disorders; Highly effective reversible contraception
☆
This study was supported by the Department of Veterans Affairs, Veteran Health Administration, Office of Research and Development, VISN 4 CHERP Pilot Project Award (PI: Sonya Borrero). L.S.C. was supported by a VA Health Services Research and Development Postdoctoral Fellowship (TPM 61-041). K.L. was supported by the VA Office of Academic Affiliations, Advanced Fellowship Program in Mental Illness Research and Treatment. ☆☆ The findings and conclusions in this report are those of the authors and do not represent the views of the Department of Veterans Affairs or the United States Government. ★ The authors have no conflicts of interest to disclose. ⁎ Corresponding author. Lisa S. Callegari, Seattle HSR&D Center, Puget Sound Veterans Administration University of Washington 1100 Olive Way, Suite 1400 Seattle, WA 98101. Tel.: +1 206 277 3129; fax: +1 206 768 5343. E-mail address: [email protected] (L.S. Callegari). http://dx.doi.org/10.1016/j.contraception.2014.02.028 0010-7824/Published by Elsevier Inc.
98
L.S. Callegari et al. / Contraception 90 (2014) 97–103
1. Introduction Over half (51%) of pregnancies in the United States are unintended, nearly all of which result from a combination of contraceptive nonuse and inconsistent or incorrect use of contraception [1–3]. Established risk factors for contraceptive nonuse and incorrect use include demographic, socioeconomic and reproductive characteristics [4–6]. Emerging data suggest that common mental illnesses such as depression or anxiety and substance use disorders (SUD) may also be important risk factors for poor contraceptive use [7–12]. Inadequately controlled mental illness and SUD also increase risks of adverse perinatal outcomes such as preterm birth, low birth weight and birth defects as well as adverse maternal outcomes such as postpartum depression [13–18]. Effective contraception to prevent unintended pregnancy in this population is therefore particularly critical. A growing body of evidence suggests that mental illness is associated with contraceptive nonuse, use of less effective methods and reduced adherence with compliance-based methods such as the birth control pill [7–10], although other studies have not found these associations [19,20]. Little data have been published to date on associations between diagnosed SUD and contraceptive use in adults [12,21]. Published studies in teenagers and college students have focused on associations between substance use or binge drinking and condom behavior, rather than associations between substance use or SUD and use of effective contraception [12,22–25]. Moreover, prior studies have not used national data to examine relationships between mental illness and SUD with contraceptive use. Women veterans have significantly higher rates of mental illness compared to their civilian counterparts [26,27]. A recent study found that 48% of women veterans using the VA health care system users who served in Afghanistan in Operation Enduring Freedom (OEF) or in Iraq in Operation Iraqi Freedom (OIF) screened positive for depression and 21% for posttraumatic stress disorder (PTSD) [26]. The prevalence of alcohol misuse among women veterans is also high, with 17% screening positive using the Alcohol Use Disorders Identification Test (AUDIT-C) in a national sample of OEF/OIF Veterans [28]. In addition to high rates of mental illness and SUD, women veterans have low rates of prescribed contraception compared to national estimates [29]. Given the rapidly increasing number of young women veterans entering the VA health care system [30], understanding determinants of contraceptive nonuse or use of less effective, nonprescription methods is critical for prevention of unintended pregnancies and improving birth outcomes. However, no studies have yet examined the associations between presence of mental illness or SUD and having a prescription contraceptive method in the VA. We therefore used national VA data to investigate the association of mental illness (depression, PTSD, anxiety, bipolar disorder, schizophrenia and adjustment disorder) and SUD (drug/alcohol use disorders) with having a prescription
contraceptive method [pill, patch, ring, injection, implant and intrauterine device (IUD)]. We also investigated associations between mental illness or SUD and specific types of prescription contraception categorized by level of effectiveness. We hypothesized that women with mental illness or SUD only, as well as women with both diagnoses, would be less likely to have prescription contraception and less likely to use highly effective prescription contraceptive methods (implant or IUD) compared to women without these diagnoses. 2. Materials and methods 2.1. Data sources We used secondary data from national VA administrative and clinical databases including the Pharmacy Benefits Management (PBM) Database (Version 3) and the Medical SAS inpatient and outpatient data files for fiscal year (FY) 2008 (October 1, 2007 through Sept 30, 2008). The VA uses an electronic medical record system that captures pharmacy and clinical data on all clinical care provided within VA facilities. The PBM Database was used to obtain information on prescribed contraceptive methods. The Medical SAS Datasets contain VA inpatient and outpatient clinical data and were used to obtain demographic measures, mental health and medical International Classification of Diseases, Ninth Revision, Clinical Modification (ICD-9-CM) codes, health care utilization and Current Procedural Terminology (CPT) codes for contraceptive procedures including insertions or removals of IUDs or subdermal contraceptive implants. This study was approved by the VA Pittsburgh and VA Puget Sound Institutional Review Boards. 2.2. Study setting and population The study population included all female veterans aged 18–45 who made at least one visit to a standard primary care clinic or designated women's health primary care clinic during FY2008. Women were excluded if they had evidence of prior hysterectomy, tubal sterilization, infertility or evidence of pregnancy during the study timeframe. 2.3. Outcome variables The outcome variables for the analyses were (1) any prescription contraceptive method based on pharmacy records or CPT codes for contraception-related procedures and (2) type of prescription contraceptive method used, categorized by level of effectiveness, in FY2008. A woman was considered to have a documented method of contraception if she had filled a prescription for a hormonal contraceptive at any time in FY2008 or if she had evidence of having received a contraceptive implant in the preceding 5 years or IUD in the preceding 10 years and no evidence of removal. Contraceptive methods were categorized using the World Health Organization's tiers of contraceptive effectiveness. Moderately effective contraception was defined as birth control pill, patch, ring or injection; highly effective
L.S. Callegari et al. / Contraception 90 (2014) 97–103
99
contraception was defined as contraceptive implant or IUD [31]. A previous analysis demonstrated that VA administrative data are less reliable in capturing nonprescription methods such as condoms, other over-the-counter products and behavioral methods such as fertility awareness [29]. Therefore, we focused on those prescription methods that are more reliably captured in VA data through pharmacy prescriptions and CPT codes.
association between mental illness and SUD diagnoses with effectiveness of the method used, we performed logistic regression with random effects among those women veterans with a prescription method of contraception, again adjusting for all covariates. Statistical analyses were conducted using SAS 9.2 (Cary, NC, USA) and Stata 11 (College Station, TX, USA).
2.4. Independent variables
3. Results
The primary predictor of interest was mental illness or SUD at any point during the study period or the FY preceding the study. Mental illness and SUD were identified based on presence of two or more outpatient or one or more inpatient ICD-9-CM codes for psychiatric or mental health diagnoses in FY2007 and FY2008 [32]. Specific mental health disorders identified using ICD-9-CM codes included depression (296.20–296.25, 296.30–296.35, 300.4, 311), PTSD (309.81), anxiety (300.00–300.09, 300.20–300.29, 300.3), bipolar disorder (296.0, 296.4–296.8), schizophrenia (295.0–295.9) and adjustment disorders (309.0–309.9, excluding 309.81). SUD identified included alcohol use disorders (303.0–303.99, 305.00–305.03) and drug use disorders (292.01–292.99, 304.0–304.99, 305.2–305.99). Additional covariates included in the analysis were age, race/ethnicity, marital status, individual annual income, religion, presence of non-VA insurance, geographic region, OEF/OIF Veteran or not and medical comorbidities which could affect contraceptive choice (hypertension, history of venous thromboembolism, coronary artery disease, breast cancer, stroke, tobacco use among women age 35 or older, diabetes with complications, liver disease and migraines with aura). We also examined health care utilization variables including whether primary care was received at a standard primary care clinic or a women's health clinic, the number of visits to a primary care clinic, whether the patient was seen in a hospital-based versus communitybased outpatient clinic and whether or not the patient was seen in a gynecology clinic. 2.5. Statistical analysis We compared demographic and clinical characteristics of our sample by presence or absence of mental illness or SUD diagnoses (mental illness only, SUD only, both SUD and mental illness diagnoses and neither mental illness nor SUD diagnoses). We also compared having a prescription contraceptive method and specific method types by presence or absence of mental illness or SUD. Chi-squared tests were used to compare variables, all of which were coded as categorical variables. To test whether mental illness or SUD diagnoses were associated with having a prescription contraceptive method, we performed logistic regression using random effects to account for nonindependence of outcomes among women seen at the same VA facility and adjusting for all covariates. To test whether there was an
A total of 103,950 women veterans aged 18–45 made at least one visit to a primary care clinic or women's health clinic in FY2008. We excluded 3846 (3.7%) women who had evidence of prior hysterectomy or tubal sterilization at VA and 1928 (1.9%) women with infertility diagnoses. In addition, we excluded 4061 (3.9%) women who had evidence of pregnancy during FY2008, resulting in a final cohort of 94,115 women. Sample demographic and clinical characteristics are shown in Table 1. In our sample, 36.5% had mental illness only, 0.6% had SUD only, 5.3% had both diagnoses and 57.7% had neither diagnosis. Among women with SUD with or without mental illness, 38.4% had an alcohol use disorder diagnosis only, 28.5% had a drug abuse diagnosis only and 33.1% had both (data not shown). Among women with mental illness only, the most common ICD-9-CM codes included depression (71.1%), PTSD (32.0%), anxiety (29.8%) and bipolar disorder (12.0%). Among women with both mental illness and SUD, 75.9% had an ICD-9-CM code for depression, 48.4% for PTSD, 35.2% for anxiety and 32.5% for bipolar disorder. Overall, 19.3% of the study sample had documentation in FY2008 of having a contraceptive method: 22.1% of women with mental illness only, 14.6% of women with SUD only, 18.2% of women with both diagnoses and 17.7% of women with neither diagnosis (Table 2). More women with mental illness only or both mental illness and SUD used a highly effective method compared to women with neither diagnosis in unadjusted analyses. Table 3 presents adjusted associations of the four mental illness and SUD categories with having prescription contraception and use of highly effective contraception. The first model examines associations with the outcome of any prescription contraceptive use among our entire sample (N= 94,115). Women with mental illness only did not differ significantly from women with neither diagnosis [adjusted odds ratio (aOR) = 1.04, 95% confidence interval (CI) = 1.00–1.08, p=0.06], although a nonsignificant trend was observed. Women with SUD only had 27% lower odds of having prescription contraception (aOR = 0.73, 95% CI = 0.57–0.95, p=0.02), and women with both diagnoses had 21% lower odds of having prescription contraception (aOR = 0.79, 95% CI = 0.73–0.86, pb0.001), compared to women with neither diagnosis. The second model examines associations with use of a highly effective method among
100
L.S. Callegari et al. / Contraception 90 (2014) 97–103
Table 1 Demographic and clinical characteristics of the study sample by presence of mental illness and substance use disorder (SUD), N=94,115
Table 1 (continued) Mental illness & SUD Neither Mental SUD only Both illness only N= 542 N= 4,956 N=54,259 N=34,358 (0.58%) % (5.3%) % (57.7%) % (36.5%) %
Mental illness & SUD Mental SUD only Both Neither illness only N=542 N= 4,956 N= 54,259 N=34,358 (0.58%) % (5.3%) % (57.7%) % (36.5%) % Demographics Age 18–25 26–35 36–45 Race/Ethnicity Hispanic Non-Hispanic White Non-Hispanic Black Other Unknown/Missing Marital status Married Div/sep/Widowed Never married Unknown Adjusted annual income ≤ US$20K Has non-VA insurance Religion Protestant Catholic Other Unaffiliated Primary care clinic type Standard primary care clinic only Women's health clinic only Both Number of visits 1–2 3–5 6+ Seen in gynecology clinic Hospital-based clinic Geographic region a Midwest Northeast South West OEF/OIF Medical comorbidities b Types of mental health diagnoses Depression PTSD Anxiety Bipolar Schizophrenia Adjustment disorder Count of types of mental health diagnoses 0 1 2
12.9 37.6 49.5
11.1 25.5 63.5
10.1 29.4 60.5
17.6 38.4 44.0
6.4 47.2 22.5 2.9 20.9
4.6 42.8 31.0 3.3 18.3
5.2 54.4 21.9 2.8 15.6
6.3 34.3 25.1 2.9 31.4
33.5 28.6 36.9 1.0 37.0
16.1 34.1 49.4 0.4 23.8
20.4 36.2 42.9 0.5 32.6
33.3 22.5 36.9 7.3 36.5
33.2
17.9
26.3
29.2
47.9 17.8 10.2 24.1
53.3 17.2 9.6 19.9
47.6 18.8 10.6 23.0
45.1 16.1 8.9 29.8
52.1
49.3
46.2
60.2
11.9
16.2
12.4
14.3
36.0
34.5
41.4
25.6
39.1 38.5 22.3 20.8
47.0 36.0 17.0 19.7
33.3 36.4 30.3 21.5
63.5 28.2 8.2 16.5
92.0
94.5
93.5
88.8
17.8 10.8 49.8 21.6 71.9 25.1
19.0 10.0 47.2 23.8 58.1 43.4
21.4 13.6 41.5 23.4 63.0 41.7
16.4 9.4 52.8 21.5 74.7 17.2
71.1 32.0 29.8 12.0 3.0 13.4
0.0 0.0 0.0 0.0 0.0 0.0
75.9 48.4 35.2 32.5 8.0 13.3
0.0 0.0 0.0 0.0 0.0 0.0
0.0 54.2 32.6
100.0 0.0 0.0
0.0 29.9 38.4
100.0 0.0 0.0
3 4+ Drug use disorder Alcohol use disorder
11.0 2.2 0.0 0.0
0.0 0.0 54.1 68.5
22.0 9.7 62.4 71.8
0.0 0.0 0.0 0.0
Chi-squared tests were used to test group differences; all variables differed significantly with pb0.001 by mental illness and SUD status except adjustment disorder (p=0.86) and alcohol use disorder (p=0.10). a Geographic region is based on US Census regions. b Comorbidities include medical conditions that could impact the use of estrogen-containing methods: hypertension, history of venous thromboembolic disease, coronary artery disease, breast cancer, stroke, tobacco use among women aged 35 years old or older, diabetes with complications, liver disease and migraines with aura.
the subset of women veterans with a contraceptive method (N= 18,115). Compared to women with neither diagnosis, women with mental illness only had 17% higher odds of using a highly effective method (aOR = 1.17, 95% CI = 1.08–1.27, pb0.001) and women with both mental illness and SUD had 30% higher odds of using a highly effective method (aOR = 1.30, 95% CI = 1.10–1.54, p=0.002).
4. Discussion Little is known about the associations between mental illness, SUD and contraceptive use in adult women. In this cross-sectional analysis of over 90,000 women veterans, women with SUD either alone or in combination with mental illness were significantly less likely to have a prescription contraceptive method compared to women with neither mental illness nor SUD. Women with mental illness only, in contrast, were as likely as women with neither diagnosis to have a prescription contraceptive method. Among women veterans using contraception, those with mental illness were more likely to use a highly effective method (IUD or implant) compared to those with neither diagnosis. While no previous studies to our knowledge examine the relationship between diagnosed SUD and contraceptive use among adults, our finding that women with SUD were less likely to have a prescription method of contraception during the study period is generally consistent with prior literature linking use of alcohol and drugs to high-risk sexual behaviors such as inconsistent condom use in teens and young adults [12,23–25]. Several older studies that included associations between alcohol use and birth control in teens and young adults, however, reached conflicting conclusions [12,21]. Our findings provide preliminary evidence that women with diagnosed SUD may be at increased risk of unintended pregnancy due to nonuse of contraception or reliance on less effective, nonprescription methods.
L.S. Callegari et al. / Contraception 90 (2014) 97–103 Table 2 Contraceptive method type among women veterans with and without mental illness and SUD, N= 94,115 Mental illness & SUD Mental SUD only % Both % Neither % illness only % Any effective method 22.1 Highly effective 5.2 IUD 5.1 Implant 0.1 Moderately effective a 16.8 OCP 12.5 Patch 0.4 Ring 1.2 Injection 3.2
14.6 4.1 3.9 0.2 10.5 6.3 0.2 0.6 3.7
18.2 4.9 4.8 0.1 13.2 8.8 0.2 1.1 3.9
17.7 3.6 3.5 0.1 14.0 10.8 0.3 1.0 2.2
OCP=oral contraceptive pill. Chi-squared tests were used to test group differences; all variables differed significantly with pb0.001 by mental illness and SUD status. a Five hundred three (0.53%) women used more than one contraceptive method during the study period.
Previous studies have linked psychological symptoms of depression and stress to inconsistent contraceptive use or nonuse [10,11], discontinuation of contraception [9] and use of less effective methods [7,8] and hypothesized that these symptoms may predispose women to riskier contraceptive behavior. Two recent analyses, however, did not detect these associations [19,20], suggesting that there is variation by context and patient population. The first study among postpartum women in Brazil found no association between perinatal psychiatric disorders and subsequent use of less effective contraceptive methods postpartum [19]; the second found that low-income women with more psychological distress prior to an abortion were more likely to choose more effective rather than less effective postabortion methods [20]. In our sample, women veterans with mental illness diagnoses were as likely to have a prescription contraceptive method and more likely to use a highly effective method than women veterans without these diagnoses. Our study differs from previous analyses in that we defined mental illness using diagnosis codes in the year prior to or during the study period, rather than using current symptoms. In addition, our study occurred in the unique setting of the VA health care system, where the emphasis on identifying and treating mental illness and close integration of mental health services into primary care may mitigate any effects of mental illness on contraceptive use. Although women with mental illness were slightly more likely to have non-VA insurance compared to women without mental illness, women with mental health needs may still preferentially use the VA system for contraception rather than seeking care outside the VA [25], which could bias towards higher rates of documented contraception. Among women veterans who had a documented method of contraception from VA, those with mental illness, with or without coexisting SUD, were more likely to use a highly effective, compliance-free method compared to those
101
without these diagnoses. These findings are encouraging in light of a recent study by Borrero et al. which found that adherence to compliance-based methods among women veterans using VA is low [33]. One possible explanation for our findings is that providers in VA may encourage use of compliance-free methods such as IUDs (rates of implant use were low) differentially in women with mental illness or SUD, as these conditions have been linked to lower compliance with health maintenance [34]. We also cannot exclude the possibility that residual confounding influenced our findings. While we adjusted for marital status, we were unable to adjust for other reproductive factors associated with use of highly effective methods such as parity, age at first sex and number of partners [35]. Because our data are cross-sectional, we could not determine whether mental illness or SUD occurred before or after the initiation of contraceptive methods. We are therefore only able to evaluate associations between mental illness and SUD with contraceptive use, rather than infer causal relationships. Additional longitudinal studies in VA and other settings are needed to further assess these relationships and their variation by context and patient population. Prospective designs are also needed to investigate potential mechanisms for observed associations. Additional limitations of our study include that we were not able to determine with certainty from the available clinical data whether women were at risk of unintended pregnancy. For example, while we were able to exclude those women with documented sterilization, hysterectomy or infertility in the VA, we may have included women in our sample with surgery or treatment outside VA. In addition, we could not exclude women in same-sex relationships or women not sexually active during the study period, as this information is not available in VA clinical data. Rates of mental illness and SUD are higher among women veterans in
Table 3 Adjusted comparisons of prescription contraceptive use by mental illness and SUD among women veterans Any prescription contraceptive method a N= 94,1150
Highly effective method a,b N= 18,115
aOR (95% CI) p value aOR (95% CI) p value Mental illness & SUD diagnoses None Reference Mental illness only 1.04 (1.00–1.08) SUD only 0.73 (0.57–0.95) Mental illness +SUD 0.79 (0.73–0.86)
Reference 1.17 (1.08–1.27) 0.02 1.26 (0.75–2.13) b0.001 1.30 (1.10–1.54) 0.06
b0.001 0.39 0.002
aOR = adjusted odds ratio. a Adjusted for age, race/ethnicity, marital status, income, religion, nonVA insurance, clinic type and number of visits, gynecologist visit, geography, military conflict and medical comorbidities. b Among the subset of women veterans with a prescription contraceptive method.
102
L.S. Callegari et al. / Contraception 90 (2014) 97–103
sexual minority relationships [36], and it is unclear how our results would be altered if these women were excluded from our analyses. We were also unable to exclude women who were trying to get pregnant; however, studies using national data have demonstrated that at any given time less than 5% of reproductive-aged women are seeking to become pregnant [37]. Lastly, our results may not be generalizable to women veterans who do not use VA or to civilian women. In summary, this study of women Veterans found that mental illness and SUD diagnoses are associated with having a prescription method of contraception in the VA healthcare system. Women with SUD appear to be least likely to have a contraceptive method, placing them at elevated risk of unplanned pregnancy. This finding is particularly concerning given the substantially increased risk of both fetal and maternal morbidity among women with SUD. Prospective research is needed to confirm these associations. If confirmed, preventive interventions to increase uptake of effective contraceptive methods will be critical among women veterans and other populations at high risk of SUD.
[15]
[16]
[17]
[18]
[19]
[20]
[21]
[22]
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): S43–8. [2] Trussell J. Contraceptive failure in the United States. Contraception 2011;83:397–404. [3] Jones RK, Darroch JE, Henshaw SK. Contraceptive use among U.S. women having abortions in 2000-2001. Perspect Sex Reprod Health 2002;34:294–303. [4] Frost JJ, Singh S, Finer LB. Factors associated with contraceptive use and nonuse, United States, 2004. Perspect Sex Reprod Health 2007;39:90–9. [5] Wu J, Meldrum S, Dozier A, Stanwood N, Fiscella K. Contraceptive nonuse among US women at risk for unplanned pregnancy. Contraception 2008;78:284–9. [6] Frost JJ, Singh S, Finer LB. U.S. women's one-year contraceptive use patterns, 2004. Perspect Sex Reprod Health 2007;39:48–55. [7] Hall KS, Moreau C, Trussell J, Barber J. Role of young women's depression and stress symptoms in their weekly use and nonuse of contraceptive methods. J Adolesc Health 2013;53:241–8. [8] Garbers S, Correa N, Tobier N, Blust S, Chiasson MA. Association between symptoms of depression and contraceptive method choices among low-income women at urban reproductive health centers. Matern Child Health J 2010;14:102–9. [9] Hall KS, White KO, Rickert VI, Reame N, Westhoff C. Influence of depressed mood and psychological stress symptoms on perceived oral contraceptive side effects and discontinuation in young minority women. Contraception 2012;86:518–25. [10] Berenson AB, Breitkopf CR, Wu ZH. Reproductive correlates of depressive symptoms among low-income minority women. Obstet Gynecol 2003;102:1310–7. [11] Stidham Hall K, Moreau C, Trussell J, Barber J. Young women's consistency of contraceptive use - does depression or stress matter? Contraception 2013;88:641–9. [12] Cooper ML. Alcohol use and risky sexual behavior among college students and youth: evaluating the evidence. J Stud Alcohol Suppl 2002:101–17. [13] Dunkel Schetter C, Tanner L. Anxiety, depression and stress in pregnancy: implications for mothers, children, research, and practice. Curr Opin Psychiatry 2012;25:141–8. [14] Grote NK, Bridge JA, Gavin AR, Melville JL, Iyengar S, Katon WJ. A meta-analysis of depression during pregnancy and the risk of preterm
[23] [24]
[25]
[26]
[27]
[28]
[29]
[30]
[31]
[32]
[33]
[34]
birth, low birth weight, and intrauterine growth restriction. Arch Gen Psychiatry 2010;67:1012–24. Seng JS, Low LK, Sperlich M, Ronis DL, Liberzon I. Post-traumatic stress disorder, child abuse history, birthweight and gestational age: a prospective cohort study. BJOG 2011;118:1329–39. Rogal SS, Poschman K, Belanger K, et al. Effects of posttraumatic stress disorder on pregnancy outcomes. J Affect Disord 2007;102: 137–43. Floyd RL, O'Connor MJ, Sokol RJ, Bertrand J, Cordero JF. Recognition and prevention of fetal alcohol syndrome. Obstet Gynecol 2005;106:1059–64. Witt WP, Wisk LE, Cheng ER, et al. Poor prepregnancy and antepartum mental health predicts postpartum mental health problems among US women: a nationally representative population-based study. Womens Health Issues 2011;21:304–13. Faisal-Cury A, Menezes PR, Huang H. The relationship between perinatal psychiatric disorders and contraception use among postpartum women. Contraception 2013;88:498–502. Steinberg JR, Tschann JM, Henderson JT, Drey EA, Steinauer JE, Harper CC. Psychological distress and post-abortion contraceptive method effectiveness level chosen at an urban clinic. Contraception 2013;88:717–24. Santelli J, Carter M, Orr M, Dittus P. Trends in sexual risk behaviors, by nonsexual risk behavior involvement, U.S. high school students, 1991-2007. J Adolesc Health 2009;44:372–9. Scott-Sheldon LA, Carey MP, Vanable PA, Senn TE, CouryDoniger P, Urban MA. Alcohol consumption, drug use, and condom use among STD clinic patients. J Stud Alcohol Drugs 2009;70:762–70. Leigh BC. Alcohol and condom use: a meta-analysis of event-level studies. Sex Transm Dis 2002;29:476–82. Peipert JF, Zhao Q, Meints L, Peipert BJ, Redding CA, Allsworth JE. Adherence to dual-method contraceptive use. Contraception 2011;84:252–8. Halpern-Felsher BL, Millstein SG, Ellen JM. Relationship of alcohol use and risky sexual behavior: a review and analysis of findings. J Adolesc Health 1996;19:331–6. Haskell SG, Gordon KS, Mattocks K, et al. Gender differences in rates of depression, PTSD, pain, obesity, and military sexual trauma among Connecticut War Veterans of Iraq and Afghanistan. J Womens Health (Larchmt) 2010;19:267–71. Lehavot K, Hoerster KD, Nelson KM, Jakupcak M, Simpson TL. Health indicators for military, veteran, and civilian women. Am J Prev Med 2012;42:473–80. Eisen SV, Schultz MR, Vogt D, et al. Mental and physical health status and alcohol and drug use following return from deployment to Iraq or Afghanistan. Am J Public Health 2012;102(Suppl 1):S66–73. Borrero S, Mor MK, Zhao X, McNeil M, Ibrahim S, Hayes P. Contraceptive care in the VA health care system. Contraception 2012;85(6):580–8. Frayne SM, Phibbs CS, Friedman SA, Berg EA, Ananth L, Iqbal S, et al. Sourcebook: women veterans in the Veterans Health Administration. Volume 1. Sociodemographic characteristics and use of VHA care. Washington, DC: Veterans Health Administration; 2010. Steiner MJ, Trussell J, Mehta N, Condon S, Subramaniam S, Bourne D. Communicating contraceptive effectiveness: a randomized controlled trial to inform a World Health Organization family planning handbook. Am J Obstet Gynecol 2006;195:85–91. Mattocks KM, Skanderson M, Goulet JL, et al. Pregnancy and mental health among women veterans returning from Iraq and Afghanistan. J Womens Health (Larchmt) 2010;19:2159–66. Borrero S, Zhao X, Mor MK, Schwarz EB, Good CB, Gellad WF. Adherence to hormonal contraception among women veterans: differences by race/ethnicity and contraceptive supply. Am J Obstet Gynecol 2013;209:103e1–11. DiMatteo MR, Lepper HS, Croghan TW. Depression is a risk factor for noncompliance with medical treatment: meta-analysis of the effects of
L.S. Callegari et al. / Contraception 90 (2014) 97–103 anxiety and depression on patient adherence. Arch Intern Med 2000;160:2101–7. [35] Kavanaugh ML, Jerman J, Hubacher D, Kost K, Finer LB. Characteristics of women in the United States who use long-acting reversible contraceptive methods. Obstet Gynecol 2011;117:1349–57.
103
[36] Mattocks KM, Sadler A, Yano EM, et al. Sexual victimization, health status, and VA healthcare utilization among lesbian and bisexual OEF/ OIF veterans. J Gen Intern Med 2013;28(Suppl 2):S604–8. [37] Mosher WD, Jones J. Use of contraception in the United States: 19822008. Vital Health Stat 23 2010:1–4.
