Advertisement

Adherence to Oral Contraception in Women on Category X Medications

      Abstract

      Background

      Over 6% of women become pregnant when taking teratogenic medications, and contraceptive counseling appears to occur at suboptimal rates. Adherence to contraception is an important component in preventing unwanted pregnancy and has not been evaluated in this population. We undertook a pharmacy claims-based analysis to evaluate the degree to which women of childbearing age who receive Category X medications adhere to their oral contraception.

      Methods

      We evaluated the prescription medication claims for over 6 million women, age 18-44 years, with prescription benefits administered by a pharmacy benefits manager. Women with 2 or more claims for a Category X medication and 2 or more claims for oral contraception were evaluated in further detail. Adherence to oral contraception was measured by analyzing pharmacy claims. Multivariable logistic regression was performed to identify factors associated with adherence.

      Results

      There were 146,758 women of childbearing age who received Category X medications, of which 26,136 also took oral contraceptive medication. Women who received Category X medications were prescribed oral contraception (18%) at rates similar to others of childbearing age (17%). Women prescribed both Category X and oral contraception demonstrated adherence similar to the overall population. Age, class of Category X medication, number of medications, prescriber's specialty, and ethnicity correlated with lower adherence rates.

      Conclusions

      Despite added risk associated with unintended pregnancy, many women who receive Category X medications have refill patterns suggesting nonadherence to oral contraception. Compared with all women age 18-44 years, women receiving teratogenic medications do not have better adherence to oral contraception.

      Keywords

      Every year millions of women of childbearing age are prescribed medications labeled by the US Food and Drug Administration (FDA) as Category D or Category X.
      • Schwarz E.B.
      • Maselli J.
      • Norton M.
      • Gonzales R.
      Prescription of teratogenic medications in United States ambulatory practices.
      Category X medications are “contraindicated in women who are or may become pregnant” based on evidence that the risks of birth defects outweigh any potential benefits of the medication, and safer medications or other forms of therapy are generally available.
      US Food and Drug Administration
      Code of Federal Regulation, Title 21.
      Despite the FDA warning, approximately 6% of US pregnancies occur in women receiving medications with known teratogenic risk.
      • Andrade S.E.
      • Raebel M.A.
      • Morse A.N.
      • et al.
      Use of prescription medications with a potential for fetal harm among pregnant women.
      • A large percentage of women who receive Category X medications have refill patterns suggesting nonadherence to oral contraception.
      • Approximately 40% of unintended pregnancies in the US occur in women using contraception.
      • Increasing awareness of nonadherence and associated factors will be helpful to identify and target strategies that improve adherence and minimize potential risk associated with unintended pregnancy.
      Unfortunately, physicians and patients alike are often unclear about the risks associated with Category X medications,
      • Addis A.
      • Sharabi S.
      • Bonati M.
      Risk classification systems for drug use during pregnancy Are they a reliable source of information?.
      • Daniel K.L.
      • Goldman K.D.
      • Lachenmayr S.
      • Erickson J.D.
      • Moore C.
      Interpretations of a teratogen warning symbol.
      • Schwarz E.B.
      • Santucci A.
      • Borrero S.
      • Akers A.Y.
      • Nikolajski C.
      • Gold M.A.
      Perspectives of primary care clinicians on teratogenic risk counseling.
      • Santucci A.K.
      • Gold M.A.
      • Akers A.Y.
      • Borrero S.
      • Schwarz E.B.
      Women's perspectives on counseling about risks for medication-induced birth defects.
      and this may help explain the low incidence of documentation of contraceptive counseling
      • Schwarz E.B.
      • Postlethwaite D.A.
      • Hung Y.
      • Armstrong M.A.
      Documentation of contraception and pregnancy when prescribing potentially teratogenic medications for reproductive-age women.
      and the unexpected number of pregnancies in women receiving Category X medications.
      • Raebel M.A.
      • Ellis J.L.
      • Andrade S.E.
      Evaluation of gestational age and admission date assumptions used to determine prenatal drug exposure from administrative data.
      • Forrester M.B.
      • Stanley S.K.
      Exposures and treatments among women of childbearing age and pregnant women reported to Texas poison centers.
      In addition, the most common form of contraception used by women receiving Category X medications is oral contraception, a method that depends upon patient adherence for effectiveness.
      • Kost K.
      • Singh S.
      • Vaughan B.
      • Trussell J.
      • Bankole A.
      Estimates of contraceptive failure from the 2002 national survey of family growth.

      Halpern V, Grimes DA, Lopez LM, Gallo MF. Strategies to improve adherence and acceptability of hormonal methods of contraception. Cochrane Database of Systematic Reviews 2006, Issue 1. Art. No.: CD004317. DOI: 10.1002/14651858.CD004317.pub2.

      Ideally, women on Category X medications would achieve near-perfect adherence to their oral contraception. However, prior research suggests that adherence to oral contraception in general is often suboptimal.
      • Nelson A.L.
      • Westhoff C.
      • Schnare S.M.
      Real-world patterns of prescription refills for branded hormonal contraceptives: a reflection of contraceptive discontinuation.
      Given that so few women receiving Category X medications appear to receive contraceptive counseling,
      • Schwarz E.B.
      • Postlethwaite D.A.
      • Hung Y.
      • Armstrong M.A.
      Documentation of contraception and pregnancy when prescribing potentially teratogenic medications for reproductive-age women.
      it is possible that these women have the same suboptimal adherence levels to oral contraception as seen in the general population. Efforts to improve adherence to oral contraception may be one important approach to minimizing unplanned pregnancies,
      • Frost J.J.
      • Darroch J.E.
      • Remez L.
      Improving contraceptive use in the United States.
      abortion, and the emotional turmoil that has been described in women who are on Category X medications at the time of conception.
      • Schwarz E.B.
      • Postlethwaite D.A.
      • Hung Y.
      • Armstrong M.A.
      Documentation of contraception and pregnancy when prescribing potentially teratogenic medications for reproductive-age women.
      Understanding adherence to oral contraception in women receiving Category X medications is an important step in understanding the role of poor adherence in unplanned pregnancies in these women and defining opportunities for improving the effectiveness of contraception. We are unaware of prior studies describing oral contraception adherence across this population of women. We undertook this study to evaluate adherence to oral contraception in women of childbearing age who received Category X medications. Our secondary aim was to identify patient-level risk factors for low adherence to oral contraception in this population.

      Methods

      Source of Data

      We utilized a database from Medco Health Solutions, Inc. (Franklin Lakes, NJ), a pharmacy benefits manager serving more than 13 million women utilizing long-term medications.

      Study Population

      Medco's de-identified database was used to create a dataset that included sociodemographic characteristics and prescription claims for all women with continuous eligibility from January 1, 2008 to June 30, 2009. The dataset included women who were of childbearing age (18-44 years) on January 1, 2008.

      Use of Category X Medications

      The study focused on a subset of drugs specifically identified by the FDA as Category X. Two sources, Micromedex
      DRUGDEX® System (electronic version). Thomson Reuters (Healthcare) Inc., Greenwood Village, Colorado, USA
      and Clinical Pharmacology
      Clinical Pharmacology [database online]. Tampa, FL: Gold Standard, Inc.; 2009
      were used to identify and validate the scope of medications included in the study (Appendix, available online). A woman was identified as a Category X medication user if she filled 1 or more claims for 1 or more Category X drugs between the dates of January 1, 2009 and June 30, 2009, as well as at least 1 additional Category X claim in the preceding 12 months.

      Use of Oral Contraceptives

      Use of oral contraception was determined by assessing pharmacy claims between January 1, 2008 and June 30, 2009. For the purposes of this study, a woman was considered to take oral contraception if she had at least 1 claim for any oral contraceptive medication during the review period (January 1, 2009 to June 30, 2009), plus at least 1 additional claim for oral contraceptive medication within the preceding 12 months.

      Oral Contraception Adherence

      To evaluate adherence to oral contraceptive medication, we used pharmacy claims to calculate a medication possession ratio (MPR). The MPR was calculated from the number of days the patient had a supply of her contraception divided by the total number of days between the first and last contraceptive medication refill date. For purposes of this study, adherence levels were defined as Low, Moderate, and Adherent, based on MPR rates of 0%-79%, 80%-94%, and 95%-100%, respectively. A 95% cutoff for adherence was used to reflect that missing more than 2 pills per month may lower effectiveness of oral contraception. A sensitivity analysis using MPR of 90%-100% as a cutoff for adherence also was performed.

      Patient Characteristics

      The following sociodemographic variables were analyzed: age, presence of children, estimated household income, ethnicity, marital status, education level, and census region of residence. We also evaluated the following clinical characteristics: total number of medications (acute and long-term medications), total number of Category X medications, class of Category X medications, and specialty for the prescriber of oral contraceptive medications. In addition, we examined the pharmacy fulfillment channels through which the patient received their oral contraception and Category X medications (mail order pharmacy, retail pharmacy, or both).

      Statistical Analysis

      Chi-squared test was used to test the association between different categorical levels of variables and adherence levels. Logistic regression analysis was conducted to examine the relationship between adherence and demographic, clinical, and socioeconomic variables. For unadjusted logistic regressions, analyses were performed 1 set of covariates at a time; for adjusted regression analyses, all covariates were included in the logistic regression. All statistical analyses were conducted using SAS version 9.1 (SAS Institute Inc., Cary, NC).

      Results

      From January 2008 to June 2009, a total of 6,253,223 women in the study population were of childbearing age and had uninterrupted pharmacy benefit coverage. As shown in Table 1, approximately 38% (2,355,790) of these women took 1 or more long-term medications to treat a chronic disease, of which 6% (146,758) took a long-term medication classified as Category X. Among women who received Category X medications, approximately 18% (26,136) took oral contraceptive medication. This was only slightly higher than the prevalence of oral contraceptive use within the overall childbearing age population (16.7%), but significantly less than the prevalence in women who received long-term medications not classified as Category X (21.1%). However, these populations had different age distributions. Adherence to oral contraceptive medications was similar in all groups across ages (Table 1).
      Table 1Prevalence and Adherence to Oral Contraception in Women Aged 18-44 Years
      Medication Possession Ratio
      LowLow ModerateHigh
      All WomenOral Contraceptive Use0%-79%80%-94%95% and Up
      n (%)n (%)%n%n%n%
      Women 18-44 years6,253,2231,044,56916.7%204,43119.6%236,14122.6%603,99757.8%
       18-241,350,021 (21.6%)332,987 (31.9%)24.7%78,59823.6%81,69124.5%172,69851.9%
       25-342,010,116 (32.1%)414,538 (39.7%)20.6%78,33018.9%91,49422.1%244,71459.0%
       35-442,893,086 (46.3%)297,044 (28.4%)10.3%47,50316.0%62,95621.2%186,58562.8%
      Women 18-44, taking chronic medications2,355,790 (37.7%)492,19820.9%99,09020.1%111,42122.6%281,68757.2%
       Non-Category X2,209,032 (93.8%)466,06221.1%94,24220.2%105,75822.7%266,06257.1%
        18-24340,404 (15.4%)126,786 (27.2%)37.2%31,28924.7%31,15824.6%64,33950.7%
        25-34715,713 (32.4%)185,517 (39.8%)25.9%37,27820.1%41,35722.3%106,88257.6%
        35-441,152,915 (52.2%)153,759 (33.0%)13.3%25,67516.7%33,24321.6%94,84161.7%
       Category X146,758 (6.2%)26,13617.8%484818.5%566321.7%15,62559.8%
        18-249985 (6.8%)4384 (16.8%)43.9%106324.2%116226.5%215949.2%
        25-3425,386 (17.3%)7263 (27.8%)28.6%146720.2%156721.6%422958.2%
        35-44111,387 (75.9%)14,489 (55.4%)13.0%231816.0%293420.2%923763.8%
      Of the 26,136 women of childbearing age who received oral contraception with their Category X medication, 97.4% of Category X prescriptions were within 4 classes: sedative hypnotic (6.8%), antineoplastic (11.1%), retinoid (19.5%), and statin (57.7%). Women who received statins were more adherent to oral contraception, while those taking antineoplastic, retinoid, or sedative hypnotic medications were less adherent (Table 2; a more complete version of Table 2 appears online).
      Table 2Adherence to Oral Contraceptive Medications in All Women Aged 18-44 Years on Category X Medications Stratified By Patient and Medication Factors
      Oral Contraceptive Adherence Level (MPR)
      LowLow ModerateAdherent
      Women 18-44 Years0%-79%80%-94%95% and Up
      n (%)n%n%n%
      All Women26,136484818.55566321.6715,62559.78
      Class of Category X meds
      Totals exceed 100% because some women receive more than 1 Category X medication.
       Amphetamine, n (%)13 (0.05)430.77538.46430.77
       Androgen, n (%)175 (0.67)3117.714525.719956.57
       Antihypertensive, n (%)9 (0.03)333.33111.11555.56
       Antineoplastic, n (%)2894 (11.07)90931.4151117.661,47450.93
       Megestrol, n (%)66 (0.25)1827.271421.213451.52
       Migraine, n (%)192 (0.73)3317.195327.6010655.21
       Misc, n (%)287 (1.10)5519.165218.1218062.72
       Misoprostol, n (%)234 (0.90)8435.904017.0911047.01
       Pain, n (%)163 (0.62)3420.863622.099357.06
       Retinoid, n (%)5087 (19.46)118523.29131925.93258350.78
       Ribavirin, n (%)40 (0.15)1025.00615.002460.00
       Sedative hypnotic, n (%)1769 (6.77)38621.8243124.3695253.82
       Statin, n (%)15,092 (57.74)209613.89311620.65988065.47
       Qarfarin, n (%)808 (3.09)14117.4518122.4048660.15
      P value<.0001<.0001<.0001
      Number of different Category X meds
       1, n (%)25,461 (97.42)471018.50552021.6815,23159.82
       2, n (%)657 (2.51)13520.5513921.1638358.30
       3 or more, n (%)18 (0.07)316.67422.221161.11
      P value.4021.9481.7286
      MPR = medication possession ratio.
      low asterisk Totals exceed 100% because some women receive more than 1 Category X medication.
      A majority (68.9%) of these women received 5 or more medications (chronic and acute) during the study period. The greater the number of total medications a woman was prescribed, the less likely she was to be adherent to her oral contraception.
      Adherence rates among women prescribed contraception by an obstetrician/gynecologist were similar to that of the total population. When the prescriber was a primary care provider, women were more likely to be adherent. Women prescribed oral contraception by a dermatologist were more likely to be nonadherent.
      Single women and women with income <$20,000 had lower adherence. Older women and women with more education had better adherence. The majority of women in the study were white or “other,” however, women who identified as an ethnic minority were less likely to be adherent.
      Risk factors independently associated with nonadherence to oral contraception (MPR ≤95%) in the multivariable logistic regression (Table 3) include younger age, taking Category X medications other than statins, taking more than 1 Category X medication, contraception prescribed by a dermatologist, less education, single marital status, and identification as an ethnic minority. When we used a ≤90% MPR as a cutoff for nonadherence in the multivariable model, we found similar associations between these factors and nonadherence.
      Table 3Multivariable Analysis of Predictors of Adherence (MPR ≥95%) to Oral Contraception
      Odds Ratio (95% CI) for Adherence to OCs (MPR ≥95%)
      UnadjustedAdjusted
      Model also adjusted for total number of prescribed chronic medications, fulfillment channel, presence of children, household income, marital status, education and geographic region.
      Age (years)
       18-241.001.00
       25-340.92 (0.87-0.97)1.33 (1.22-1.44)
       35-441.45 (1.38-1.52)1.45 (1.33-1.58)
      Class of Category X meds
       Statins1.001.00
       Retinoids0.63 (0.60-0.67)0.70 (0.64-0.76)
       Warfarin1.02 (0.88-1.17)0.87 (0.75-1.01)
       Sedative hypnotic0.77 (0.70-0.85)0.68 (0.61-0.75)
       Antineoplastic0.67 (0.62-0.72)0.61 (0.56-0.66)
       Other0.83 (0.74-0.94)0.72 (0.63-0.82)
      Number of Category X meds
       11.001.00
       2 or more1.06 (0.91-1.24)0.73 (0.62-0.87)
      Prescriber specialty
       OB/GYN1.001.00
       Primary care1.30 (1.22-1.39)1.24 (1.16-1.34)
       Dermatology0.47 (0.40-0.56)0.70 (0.58-0.84)
       Rheumatology/Onc/Other0.91 (0.86-0.95)1.00 (0.94-1.06)
      Ethnicity
       White/other1.001.00
       Hispanic0.70 (0.63-0.77)0.69 (0.62-0.76)
       Black0.83 (0.74-0.0.93)0.76 (0.68-0.0.86)
       Asian0.78 (0.68-0.90)0.81 (0.69-0.93)
      CI = confidence interval; OC=oral contraceptives; MPR=medication possession ratio; OB/GYN=obstetrics/gynecology.
      low asterisk Model also adjusted for total number of prescribed chronic medications, fulfillment channel, presence of children, household income, marital status, education and geographic region.

      Discussion

      Despite the availability of safe and effective contraception in the US, approximately 40% of unintended pregnancies occur in women using contraception, and 90% of these are due to inconsistent or incorrect use.
      • Homco J.D.
      • Peipert J.F.
      • Secura G.M.
      • Lewis V.A.
      • Allsworth J.E.
      Reasons for ineffective pre-pregnancy contraception use in patients seeking abortion services.
      • Frost J.J.
      • Darroch J.E.
      Factors associated with contraceptive choice and inconsistent method use, United States, 2004.
      Inconsistency is especially problematic in women who take teratogenic medications. In this study we examined prescription claims for over 6 million women and described prevalence of oral contraception use and adherence in women aged 18-44 years who also took Category X medications. We found that the percentage of women who received oral contraception was no greater in women taking Category X medications than in the general population. In addition, approximately 40% of women who received Category X medications and oral contraception had refill patterns that suggested suboptimal adherence to their oral contraception. While the potential teratogenicity of Category X medications would be expected to compel women to adhere to contraception, we found no better adherence to oral contraception than in the general population. Finally, we identified certain groups of women at greater risk of nonadherence, including those who identified as an ethnic minority and reported less education. Our findings suggest that nonadherence to oral contraception may cause many women who receive Category X medications to be at risk for unintended pregnancy complicated by concerns about teratogenesis.
      There are several possible explanations for why women who received Category X medications were not more adherent to oral contraception than women in the general population. First, patients may not understand information they receive about the teratogenic risk posed by Category X medications or the importance of >95% adherence to oral contraception. Researchers at the Centers for Disease Control and Prevention tested the effectiveness of a teratogen warning symbol to communicate risk and found that only 1 in 5 women correctly interpreted the symbol to mean that they should not get pregnant while taking the medication.
      • Daniel K.L.
      • Goldman K.D.
      • Lachenmayr S.
      • Erickson J.D.
      • Moore C.
      Interpretations of a teratogen warning symbol.
      Accurate interpretation of labels can be particularly challenging for patients with low literacy levels, and patients with low literacy are more likely to ignore warning labels.
      • Davis T.C.
      • Wolf M.S.
      • Bass P.F.
      • et al.
      Literacy and misunderstanding prescription drug labels.
      Given that low health literacy levels have been found in approximately 1 in 3 English-speaking patients and over half of Spanish-speaking patients, low literacy may be an important risk factor for low adherence in this population.
      • Williams M.V.
      • Parker R.M.
      • Baker D.W.
      • et al.
      Inadequate functional health literacy among patients at two public hospitals.
      • Gazmararian J.A.
      • Baker D.W.
      • Williams M.V.
      • et al.
      Health literacy among Medicare enrollees in a managed care organization.
      We found that Hispanics were less likely to be adherent. Conversely, women with higher levels of education were more likely to be adherent. In prior research, ethnic minorities and patients with less education were more likely to misread labels, however, after adjusting for literacy level and other confounders, ethnicity and education had no independent effect.
      • Davis T.C.
      • Wolf M.S.
      • Bass P.F.
      • et al.
      Literacy and misunderstanding prescription drug labels.
      Additional research is warranted to extend our understanding of the effect of low literacy and limited English proficiency on comprehension of teratogenic risk and adherence to contraception in women receiving Category X medications.
      Second, it is possible that women who received Category X medications were not effectively counseled about teratogenic risks. Although women want to be informed about teratogenic risks, physicians do not routinely counsel patients about such risks or the need for contraception. Suboptimal rates of contraceptive counseling have been described with women who are prescribed potentially teratogenic medications, ranging from 17% to 69%.
      • Schwarz E.B.
      • Maselli J.
      • Norton M.
      • Gonzales R.
      Prescription of teratogenic medications in United States ambulatory practices.
      • Schwarz E.B.
      • Postlethwaite D.A.
      • Hung Y.
      • Armstrong M.A.
      Documentation of contraception and pregnancy when prescribing potentially teratogenic medications for reproductive-age women.
      • Lohr P.A.
      • Schwarz E.B.
      • Gladstein J.E.
      • Nelson A.L.
      Provision of contraceptive counseling by internal medicine residents.
      • Weisman C.S.
      • Maccannon D.S.
      • Henderson J.T.
      • Shortridge E.
      • Orso C.L.
      Contraceptive counseling in managed care: preventing unintended pregnancy in adults.
      • James L.
      • Barnes T.R.E.
      • Lelliott P.
      Informing patients of the teratogenic potential of mood stabilizing drugs: a case note review of the practice of psychiatrists.
      The largest of these studies, by Schwarz et al,
      • Schwarz E.B.
      • Postlethwaite D.A.
      • Hung Y.
      • Armstrong M.A.
      Documentation of contraception and pregnancy when prescribing potentially teratogenic medications for reproductive-age women.
      evaluated the documentation of contraceptive counseling among 488,175 women of reproductive age. Fifty-one percent of women who filled a Category X medication had no documentation of contraceptive counseling or contraception dispensed within the 2 years before filling the Category X prescription.
      • Schwarz E.B.
      • Postlethwaite D.A.
      • Hung Y.
      • Armstrong M.A.
      Documentation of contraception and pregnancy when prescribing potentially teratogenic medications for reproductive-age women.
      There are several possible reasons for this low rate of counseling. Many primary care physicians have difficulty accessing relevant clinical resources or prescription data that might help identify those patients who most need contraceptive and teratogenic risk counseling.
      • Schwarz E.B.
      • Santucci A.
      • Borrero S.
      • Akers A.Y.
      • Nikolajski C.
      • Gold M.A.
      Perspectives of primary care clinicians on teratogenic risk counseling.
      Others may have limited training or are not comfortable with contraceptive counseling.
      • Steinauer J.
      • LaRochelle F.
      • Rowh M.
      • Backus L.
      • Sandahl Y.
      • Foster A.
      First impressions: what are preclinical medical students in the US and Canada learning about sexual and reproductive health?.
      Despite evidence that contraceptive counseling is often omitted, many physicians believe they should be counseling women when they prescribe potentially teratogenic medications.
      • Schwarz E.B.
      • Santucci A.
      • Borrero S.
      • Akers A.Y.
      • Nikolajski C.
      • Gold M.A.
      Perspectives of primary care clinicians on teratogenic risk counseling.
      In a qualitative study that described preferences for communication about the risks of medication-induced birth defects, women expressed their desire for health care providers to initiate discussions about potentially teratogenic medications at the time they are prescribed regardless of expressed intention to become pregnant.
      • Santucci A.K.
      • Gold M.A.
      • Akers A.Y.
      • Borrero S.
      • Schwarz E.B.
      Women's perspectives on counseling about risks for medication-induced birth defects.
      Improvement in adherence to oral contraception in this population likely will require efforts by physicians, pharmacists, other health care professionals, and health care delivery systems. Pharmacy claims can be used to alert members of the health care team about women who are late to fill their oral contraception. Pharmacists can assist in counseling patients about potential teratogenic risks of Category X medications and can identify and address barriers to adherence to oral contraception. Such counseling should not replace discussions between patients and physicians at the point of prescribing. Patients have been shown to want these discussions,
      • Santucci A.K.
      • Gold M.A.
      • Akers A.Y.
      • Borrero S.
      • Schwarz E.B.
      Women's perspectives on counseling about risks for medication-induced birth defects.
      and strategies should be developed to support physicians in identifying this issue at the point of care. Electronic prescribing with built-in reminders may be one promising technology for reminding physicians of the need to address teratogenic counseling at the time of prescribing.
      • Aspden P.
      • Wolcott J.
      • Bootman J.L.
      • Cronenwett L.R.
      Preventing Medication Errors: Quality Chasm Series.
      Counseling by any member of the health care team should utilize multiple instructional modalities to reach patients of all literacy levels.
      While this study performed in a large national database has many strengths, our findings must be interpreted in light of the following limitations. First, we measured adherence to oral contraception using pharmacy claims data without medical record data or patient interviews. Therefore, we could not assess literacy levels to confirm whether women were taking oral contraception specifically for contraception. Some women may have been using oral contraception for other medical conditions, such as polycystic ovarian syndrome or dysmenorrhea, in which case, 95% adherence is less critical. We could not assess possible reasons for nonadherence. Some women with low adherence may have used a second method of contraception, making adherence to oral contraception less important (eg, abstinence, condoms, diaphragm, nonoral hormonal contraceptives). However, the MPR is calculated between the first and last claim during the study period, so a low MPR indicates a gap between refills rather than a cessation of use, and therefore, it is unlikely that an intervening contraceptive method explains the significant number of low MPRs seen in this study. Women may have also received oral contraception intermittently from sources outside the pharmacy claims system (eg, physician samples, self-pay, sharing medications). While these limitations may bias our study toward overestimating nonadherence to oral contraception, given the high percentage of nonadherence seen in this study, the true number of women missing more than 5% of their oral contraception doses is likely to be significant. In addition, our database does fully capture all claims submitted under the pharmacy benefit regardless of channel, increasing the chances that the large majority of oral contraception use is included in the calculation of MPR.
      Second, we chose to define adherence to oral contraception as ≥95%. Recognizing that recommendations for how many pills a woman can miss before she experiences a decrease in effective contraception range from 1-3 pills per month, we also performed our analysis using a ≥90% cutoff. Analysis using this more liberal definition of adherence to oral contraception showed nearly identical findings.
      Finally, we did not look at pregnancy rates in this population to confirm that nonadherence in these women is associated with a greater likelihood of pregnancy. Other studies have evaluated pregnancy rates and have found only slightly lower incidence of pregnancy in women on Category X and D versus those taking a safer class of medications (Category A or B medications).
      • Schwarz E.B.
      • Postlethwaite D.A.
      • Hung Y.
      • Armstrong M.A.
      Documentation of contraception and pregnancy when prescribing potentially teratogenic medications for reproductive-age women.
      There are 2 important conclusions from this study. First, many women on Category X medications who are using oral contraception may not meet the adherence rate required for effective contraception. A pharmacy claims database is one approach to identify these women. Second, women who are within an ethnic minority and women who have less education appear most vulnerable. Further research will be required to confirm whether low literacy is responsible for this association. Given physicians' potentially limited time and access to drug information, pharmacists may play an important role by providing feedback about contraceptive adherence and assisting with counseling using instructional formats that are effective for all literacy levels. Future studies should evaluate the impact of interventions, including the use of pharmacists, to improve adherence to oral contraception in women of childbearing age on Category X medications.

      Acknowledgment

      The authors thank the following Medco Health Solutions, Inc., employees for their statistical support: Inderpal Bhandari, PhD, Rocco Lulic, MS, Mahj Rafiq, BBA, and Zhuliang Tao, MPH. Additionally, we wish to thank the following consultants: Cynthia Fenton, MD and Steven Haffner, MD for their help with interpretation of statistical results and revisions of the manuscript.

      Appendix

      AppendixCategory X Medications Included in Analysis
      Class DescriptionGeneric Name
      AmphetamineBenzphetamine hcl
      AndrogenDanazol
      Fluoxymesterone
      Methyltestosterone
      Nandrolone decanoate
      Nandrolone phenpropionate
      Oxandrolone
      Testosterone
      Testosterone cypionate
      Testosterone enanthate
      Testosterone propionate
      AntihypertensiveAmbrisentan
      Bosentan
      AntineoplasticAnastrozole
      Bexarotene
      Bicalutamide
      Fluorouracil
      Goserelin acetate
      Leuprolide acetate
      Methotrexate sodium
      Methotrexate sodium/pf
      Raloxifene hcl
      Thalidomide
      Tositumomab
      Tositumomab iodine-131
      Triptorelin
      MegestrolMegestrol acetate
      MigraineDihydroergotamine mesylate
      Ergoloid mesylates
      Ergonovine ergotrate
      Ergotamine tartrate
      Ergotamine tartrate/caffeine
      MiscAcetohydroxamic acid
      Amyl nitrite
      Dronedarone
      Dutasteride
      Finasteride
      Leflunomide
      Lenalidomide
      Mifepristone
      Miglustat
      Sodium iodide
      MisoprostolMisoprostol
      PainDiclofenac sodium/misoprostol
      RetinoidAcitretin/emollient comb
      Isotretinoin
      Tazarotene
      Tretinoin/mequinol
      RibavirinRibavirin
      Sedative hypnoticEstazolam
      Flurazepam hcl
      Quazepam
      Temazepam
      Temazepam/diet8
      Triazolam
      StatinAmlodipine/atorvastatin
      Atorvastatin calcium
      Ezetimibe/simivastatin
      Fluvastatin sodium
      Lovastatin
      Niacin/lovastatin
      Niacin/simvastatin
      Pravastatin sodium
      Rosuvastatin calcium
      Simvastatin
      WarfarinWarfarin sodium

      Supplementary data

      References

        • Schwarz E.B.
        • Maselli J.
        • Norton M.
        • Gonzales R.
        Prescription of teratogenic medications in United States ambulatory practices.
        Am J Med. 2005; 118: 1240-1249
        • US Food and Drug Administration
        Code of Federal Regulation, Title 21.
        Fed Regist. 2003; 21 (21 C.F.R. Section 201.57): 21-32
        • Andrade S.E.
        • Raebel M.A.
        • Morse A.N.
        • et al.
        Use of prescription medications with a potential for fetal harm among pregnant women.
        Pharmacoepidemiol Drug Saf. 2006; 15: 546-554
        • Addis A.
        • Sharabi S.
        • Bonati M.
        Risk classification systems for drug use during pregnancy.
        Drug Saf. 2000; 3: 245-253
        • Daniel K.L.
        • Goldman K.D.
        • Lachenmayr S.
        • Erickson J.D.
        • Moore C.
        Interpretations of a teratogen warning symbol.
        Teratology. 2001; 64: 148-153
        • Schwarz E.B.
        • Santucci A.
        • Borrero S.
        • Akers A.Y.
        • Nikolajski C.
        • Gold M.A.
        Perspectives of primary care clinicians on teratogenic risk counseling.
        Birth Defects Res A Clin Mol Teratol. 2009; 85: 858-863
        • Santucci A.K.
        • Gold M.A.
        • Akers A.Y.
        • Borrero S.
        • Schwarz E.B.
        Women's perspectives on counseling about risks for medication-induced birth defects.
        Birth Defects Res A Clin Mol Teratol. 2010; 88: 64-69
        • Schwarz E.B.
        • Postlethwaite D.A.
        • Hung Y.
        • Armstrong M.A.
        Documentation of contraception and pregnancy when prescribing potentially teratogenic medications for reproductive-age women.
        Ann Intern Med. 2007; 147: 370-376
        • Raebel M.A.
        • Ellis J.L.
        • Andrade S.E.
        Evaluation of gestational age and admission date assumptions used to determine prenatal drug exposure from administrative data.
        Pharmacoepidemiol Drug Saf. 2005; 14: 829-836
        • Forrester M.B.
        • Stanley S.K.
        Exposures and treatments among women of childbearing age and pregnant women reported to Texas poison centers.
        Vet Hum Toxicol. 2004; 46: 210-212
        • Kost K.
        • Singh S.
        • Vaughan B.
        • Trussell J.
        • Bankole A.
        Estimates of contraceptive failure from the 2002 national survey of family growth.
        Contraception. 2008; 77: 10-21
      1. Halpern V, Grimes DA, Lopez LM, Gallo MF. Strategies to improve adherence and acceptability of hormonal methods of contraception. Cochrane Database of Systematic Reviews 2006, Issue 1. Art. No.: CD004317. DOI: 10.1002/14651858.CD004317.pub2.

        • Nelson A.L.
        • Westhoff C.
        • Schnare S.M.
        Real-world patterns of prescription refills for branded hormonal contraceptives: a reflection of contraceptive discontinuation.
        Obstet Gynecol. 2008; 112: 782-787
        • Frost J.J.
        • Darroch J.E.
        • Remez L.
        Improving contraceptive use in the United States.
        Issues Brief (Alan Guttmacher Inst). 2008; 1: 1-8
        • DRUGDEX® System (electronic version). Thomson Reuters (Healthcare) Inc., Greenwood Village, Colorado, USA
        (Accessed September 19, 2009)
        • Clinical Pharmacology [database online]. Tampa, FL: Gold Standard, Inc.; 2009
        (Accessed September 19, 2009)
        • Homco J.D.
        • Peipert J.F.
        • Secura G.M.
        • Lewis V.A.
        • Allsworth J.E.
        Reasons for ineffective pre-pregnancy contraception use in patients seeking abortion services.
        Contraception. 2009; 80: 569-574
        • Frost J.J.
        • Darroch J.E.
        Factors associated with contraceptive choice and inconsistent method use, United States, 2004.
        Perspect Sex Reprod Health. 2008; 40: 94-104
        • Davis T.C.
        • Wolf M.S.
        • Bass P.F.
        • et al.
        Literacy and misunderstanding prescription drug labels.
        Ann Intern Med. 2006; 145: 887-894
        • Williams M.V.
        • Parker R.M.
        • Baker D.W.
        • et al.
        Inadequate functional health literacy among patients at two public hospitals.
        JAMA. 1995; 274: 1677-1682
        • Gazmararian J.A.
        • Baker D.W.
        • Williams M.V.
        • et al.
        Health literacy among Medicare enrollees in a managed care organization.
        JAMA. 1999; 281: 545-551
        • Lohr P.A.
        • Schwarz E.B.
        • Gladstein J.E.
        • Nelson A.L.
        Provision of contraceptive counseling by internal medicine residents.
        J Womens Health (Larchmt). 2009; 18: 127-131
        • Weisman C.S.
        • Maccannon D.S.
        • Henderson J.T.
        • Shortridge E.
        • Orso C.L.
        Contraceptive counseling in managed care: preventing unintended pregnancy in adults.
        Women's Health Issues. 2002; 12: 79-95
        • James L.
        • Barnes T.R.E.
        • Lelliott P.
        Informing patients of the teratogenic potential of mood stabilizing drugs: a case note review of the practice of psychiatrists.
        J Psychopharmacol. 2007; 21: 815-819
        • Steinauer J.
        • LaRochelle F.
        • Rowh M.
        • Backus L.
        • Sandahl Y.
        • Foster A.
        First impressions: what are preclinical medical students in the US and Canada learning about sexual and reproductive health?.
        Contraception. 2009; 80: 74-80
        • Aspden P.
        • Wolcott J.
        • Bootman J.L.
        • Cronenwett L.R.
        Preventing Medication Errors: Quality Chasm Series.
        National Academies Press, Washington, DC2007