Thinking (Re)Productively – June 2015

Performance measures for contraceptive care: what are we actually trying to measure?

Christine Dehlendorf, Helen Bellanca, Michael Policar

Over the past two decades, there has been increased attention to quality in health care, including the use of standardized measures to track performance within health care institutions, by plans and by individual providers [1]. These measures are a means to document quality of care so as to allow comparisons by consumers, payers and others. In addition, they can be used to incentivize quality improvement through such means as linking physician payments to their performance on quality measures. The Affordable Care Act (ACA) furthers this trend, with an emphasis on a pay-for-performance model based on quality and effectiveness measures [2].

The availability and use of concrete measures of quality in specific areas of health care can drive quality improvement and influence the degree to which these areas are prioritized by policy makers, providers of clinical services and funders of health care services. Unfortunately, family planning has been neglected in the development of performance measures, with no measures related to reproductive planning included in the 275 measures endorsed by the National Quality Forum or in the Adult Core Set of measures promulgated by Medicaid. To address this gap and to promote quality care, there has been increasing interest in developing standard quality measures for contraceptive care [3]. As an example, the 2013 Sexual and Reproductive Health (SRH) Workforce Summit meeting, convened by the Association of Reproductive Health Professionals, included in its Key Summit Recommendations to “Develop one Healthcare Effectiveness Data and Information Set (HEDIS®) measure on SRH” and to “Define SRH quality metrics for use in new models of care that provide incentives for quality of care/pay for performance” [4]. Several groups, including one convened by the Planned Parenthood Federation of America and a working group formed following the SRH Workforce Summit, have been actively working to explore potential metrics for contraceptive care.

At the same time, the reproductive health community is increasingly directing research and interventions towards the perceived underuse of highly effective forms of contraception, with a focus primarily on intrauterine contraception and implants, known collectively as long-acting reversible contraception (LARC). Due to their superior contraceptive efficacy, these methods are seen as desirable for individual women as a means to achieve their reproductive goals. On a public health level, increased use of these methods has the potential to lower unintended pregnancy rates and has been suggested as one means to decrease pregnancy complications by reducing unintended pregnancies [5]. In addition, some have connected the use of LARC methods with the goal of reducing poverty [6], including the statement, at a US House of Representatives Committee on Budget hearing, that “increasing the access of poor and low-income women to effective, long-lasting methods of birth control would reduce nonmarital births and the poverty rates as well” [7].

Consistent with this focus on LARC methods, some have suggested the use of measures designed to encourage increased use of these methods [3]. One possibility, for example, is to simply measure the percentage of women receiving family planning care who decide to use a highly effective method, with higher uptake being equated with higher quality of care. A proposed modification of this measure, termed the Contraceptive Protection Index, would use weighted averages of the typical use efficacy for each method multiplied by the percentage of women using the method [8]. The resulting averages would naturally weight LARC higher than shorter-acting hormonal methods, based on their higher efficacy, while also weighting shorter-acting hormonal methods more highly than barrier methods.

Is uptake of highly effective methods the best measure of quality in family planning care?

At first glance, the use of measures emphasizing the uptake of higher efficacy methods is logical based on the desire to improve prevention of unintended pregnancies on both an individual and societal level. This is particularly true given that many providers have misconceptions about LARC methods that could impede the provision of these methods [9]. Using a quality measure based on LARC use could therefore incentivize providers and practices to recognize the value of these methods and ensure women’s access to them.

However, there is cause for concern that this focused, outcome-based measure may not be appropriate in the setting of a decision as complex and contextualized as the choice of a contraceptive method. As detailed in two recent excellent commentaries on this topic, the choice of a contraceptive method is highly preference-sensitive due to the large number of available options and women’s varied preferences for method characteristics [[10], [11]]. There are up to 10 methods that are medically appropriate for the majority of women, and these methods have a variety of defining characteristics, including how often they are taken (e.g., daily, weekly, monthly, etc.), how they are taken (by mouth, by shot, on the skin, etc.) and what side effects are associated with them. Women have a broad range of preferences related to these characteristics [12]; for example, some women may prioritize efficacy above all other method characteristics, while for others the most important aspect of their method may be whether or not it affects their bleeding patterns or contains hormones.

While these factors alone make contraception more preference-sensitive than, for example, common measures of quality such as whether an individual receives appropriate therapy for asthma or is immunized against pneumococcus after the age of 65, perhaps the most important factor contributing to the preference-sensitive nature of contraceptive decision making is its relationship to intimate issues related to fertility, relationships and sexuality. Women’s feelings about method characteristics will be affected by their personal preferences related to such issues as the certainty of their desire to avoid pregnancy, whether or not they are comfortable with using a method that is inside their vagina or their uterus, and whether having irregular bleeding would negatively impact their sex life. Protecting women’s reproductive autonomy therefore requires the recognition that women’s preferences need to be paramount in the choice of a contraceptive method even if they are not consistent with the public health goal of decreasing unintended pregnancies.

In light of the unique nature of contraceptive decision making, quality measures that focus only on the short-term outcome of choice of a highly effective method are problematic, as they encourage the provision of counseling that emphasizes and/or promotes these methods at the expense of attention to patient preferences. Not only is this concerning from a reproductive autonomy perspective, but it also has the potential to be to the detriment of long-term outcomes, such as patient satisfaction and method continuation. Evidence supporting such an impact includes a qualitative study that found that women receiving contraceptive care value engaging in shared decision making in which their provider gives information that is in line with the woman’s stated values and preferences, allowing the woman to make the final decision herself [13]. In addition, there is evidence that women prefer more autonomy in contraceptive decision making than in other medical decisions [14]. The degree to which counseling is focused on patient preferences has also been associated with contraceptive use, as “client-centered” counseling has been linked with contraceptive continuation [15], while perceived pressure to use a long-acting method is associated with contraceptive discontinuation [16].

While it may be tempting to believe that instituting a LARC-based quality measure would not in fact result in the undesirable consequence of having providers shift towards more directive counseling, in fact, the potential for unintended consequences of performance measures has been documented [17]. These include such effects as clinicians bypassing informed consent in order to ensure that they meet thresholds for chlamydia screening, forcing the disenrollment of noncompliant patients from practices [18], referring high-risk patients to other providers [19] and overprescribing antibiotics [[20], [21], [22]]. While this is undoubtedly more likely in situations where performance on a quality measure is linked to payment, there are other mechanisms by which measurement can motivate behavior, including both external (e.g., public) and internal (e.g., practice or hospital level) reporting. There is no reason to expect that family planning providers would be more immune to the incentives associated with performance measures than clinicians in other areas of health care.

Implications for women of color and vulnerable populations

Incentivizing counseling that is focused on LARC methods could be particularly problematic among disadvantaged populations, given the history of forced sterilization and coercive contraceptive practices targeting women in these groups in the United States [23]. Further, reproductive coercion is not only a relic of the distant past. In the 1990s, for example, marketing for the Norplant contraceptive implant was targeted towards women of color; Medicaid in some states covered the cost of Norplant insertion for poor women, but not its removal [24]; and women convicted of crimes were given a “choice” between having Norplant inserted and serving prison time [25]. As recently as 2010, incarcerated women in California were reportedly subjected to illegal sterilization by prison doctors who did not obtain the required state approvals [26].

Considering the ongoing legacy of reproductive coercion in the United States, women who are members of the affected groups — including poor women, women of color, women with disabilities, young women and those in the correctional system — have reason to be suspicious about the degree to which clinicians and medical institutions will safeguard reproductive autonomy. A national survey of conspiracy beliefs about birth control found that one third of black women agreed with the statement, “Medical and public health institutions use poor and minority people as guinea pigs to try out new birth control methods” and that women who held strong contraceptive safety conspiracy beliefs were less likely to be using a provider-dependent method, such as intrauterine devices (IUDs) or implants [27]. A qualitative study of low-income women of color found that one third of the women felt that they had experienced coercion during contraceptive counseling [28]. In addition, there is evidence that clinicians promote long-acting methods differentially based on race, as a 2010 study using standardized patients found that family planning providers were more likely to recommend IUDs to low-income women of color than to low-income white women, holding other patient characteristics constant [29]. Incentivizing counseling that is biased towards long-acting methods, rather than focused on women’s needs and preferences, has the potential to amplify existing biases and disparities within the US health care system and worsen preexisting distrust among communities of color and other vulnerable populations. In addition, the use of a performance measure focused on LARC uptake may be problematic even if it did not influence counseling, as its mere existence could be perceived negatively by communities sensitized to these issues as evidence of a focus on controlling women’s reproduction, rather than on empowering women.

Patient-centered measures of quality of contraceptive care

The potential problems associated with performance measures for contraception focused solely on uptake of LARC methods illustrate the need for measures that better capture the quality of contraceptive care in the context of women’s needs and preferences. A combination of measures will likely be necessary to address the multidimensional nature of quality, including interpersonal quality, availability of information and access to services. In combination, these measures should take into account the importance of both the patient experience of counseling and whether each woman is able to choose the appropriate contraceptive method for her.

Measures of patient experience with care have long been of recognized value in the measurement of quality, including having been collected since 1995 by the Agency for Healthcare Research and Quality through the Consumer Assessment of Healthcare Providers and Systems and being a mandated part of quality assessment implemented as part of the Affordable Care Act [30]. This emphasis on patients’ perceptions of the care they receive is motivated by both an inherent belief in the value of ensuring patient satisfaction, as well as findings that measures of patient experience are associated with improved clinical outcomes [[30], [31]]. Studies support the association between patient experience and improved clinical outcomes in the specific context of family planning care [[32], [33], [34], [35]]. While obtaining robust measurements of patient experience requires a much greater investment of resources for data collection than claims-based measures, especially given challenges in obtaining adequate response rates to these types of surveys, they are irreplaceable in providing information about whether care is meeting patients’ needs. This is of particular value in the assessment of contraceptive care, given the personal nature and complex context of contraceptive decision making.

Process measures designed to determine whether women are able to choose an appropriate contraceptive method could address some of the same issues targeted by those measures focused on LARC uptake — namely, provider resistance to provision of these methods and the resulting limitation on women’s ability to choose these methods — while avoiding their pitfalls. For example, the National Health Service in the United Kingdom has implemented a measure that incentivizes the provision of information about LARC methods [36], while the American College of Obstetricians and Gynecologists has recommended a performance measure based on whether or not women are offered LARC methods [37]. Both of these approaches focus on giving women options and information and are consistent with maintaining an emphasis on women’s preferences during counseling and ensuring that they have adequate information to make an informed choice. An alternative claims-based measure, currently being validated by the Centers for Disease Control and the Office of Population Affairs, is designed to incentivize the provision of LARC methods, while minimizing the potential for over-promotion, by explicitly not setting a benchmark for the percentage of women using these methods and by aggregating data at the level of the health plan. The goal of this approach, which is considered an access measure, would be to identify practices that are well below the mean in order to provide the opportunity to address barriers to LARC use [38]. A similar approach could consist of a measure that used a minimum “floor” standard, such as 1% of all eligible patients, in order to differentiate providers who offer these methods at all from those who do not. In implementing this type of measure, care would need to be taken to ensure that their intent was clear in order to avoid the interpretation that promotion of LARC methods was being incentivized.

Some have advocated for an intermediate-outcome measure similar to LARC-based measures, but which focuses on use of any method of contraception considered to be either moderately or highly effective, which therefore includes hormonal methods in addition to LARC methods [3]. This measure allows for greater consideration of patient preferences, as compared to a measure that is focused on LARC alone or that gives greater priority to higher-efficacy methods, and is therefore preferable. However, it is possible that this measure could incentivize providers to deemphasize counseling about condoms, which could have implications for women for whom this would be an appropriate method choice due to their personal preferences and/or their risk of sexually transmitted infections. As a result, this might best be implemented in conjunction with an experience-based measure that simultaneously tracks the degree to which patients’ needs are met in the counseling encounter. An additional process-based measure that has been discussed in the literature is whether women of reproductive age are screened for the need for contraception at any given health care encounter. This type of measure allows for a more population-based assessment of whether women’s needs are being met, as opposed to focusing on those who are already accessing family planning services [39], and would be facilitated by consistent documentation of women’s risk for unintended pregnancy in electronic health records as currently being advocated by the Office of Population Affairs [40].

Conclusion

Measurement of quality in contraceptive care may ensure that family planning services are prioritized in our evolving health care system and that attention is paid to continuous quality improvement in order to ensure that women receive the best possible care. In order to accomplish these goals, it is essential that performance measures motivate care that best reflects quality in the context of women’s needs and preferences. While we recognize the importance of working to ensure that all women have access to LARC methods, measures used to accomplish this goal should not have the effect of inadvertently undermining quality of care by incentivizing directive or potentially coercive practices. Measures that, either individually or in combination, reward the quality of contraceptive care from both a patient and systems perspective, while protecting women’s autonomy, should be prioritized by those developing performance measures. In addition, organizations such as the National Quality Forum, which consider the potential impact of measures from a variety of perspectives in their formal endorsement processes, can help to ensure that the preference-sensitive nature of contraceptive decision making is reflected in measures designed to incentivize quality family planning care.

Acknowledgments

The authors thank Nora Anderson, MPA, for her assistance with conceptualization and preparation of this commentary.

References

1. Chassin, M.R., Loeb, J.M., Schmaltz, S.P., and Wachte, R.M. Accountability measures — using measurement to promote quality improvement. N Engl J Med. 2010; 363: 683–688
2. Health policy brief: pay-for-performance. Health Aff. 2012;
3. Sonfield, A. Pay-for-performance: making it work for safety-net family planning centers and the clients they serve. . 2014; 17
4. Association of Reproductive Health Professionals. in: ARHP (Ed.) Sexual and Reproductive Health Workforce Project: overview and recommendations from the SRH Workforce Summit. ; 2014
5. Services, C.f.M.M. CMCS maternal and infant health initiative: improving maternal and infant health outcomes in Medicaid and CHIP. ([cited 2015 January 15]; Available from: http://www.medicaid.gov/medicaid-chip-program-information/by-topics/quality-of-care/maternal-and-infant-health-care-quality.html; 2012.
6. Amato, P.R. and Maynard, R.A. Decreasing nonmarital births and strengthening marriage to reduce poverty.Future Child. 2007; 17: 117–141
7. Haskins, R. Poverty and opportunity: begin with facts. in: Committee on Budget. Brookings Institute, ; 2014
8. Belden, P. Counting contraception. Contraception. 2014; 89: 145–147
9. Harper, C.C., Blum, M, de Bocanegra, HT, Darney, PD, Speidel, JJ, Policar, M, and Drey, EA. Challenges in translating evidence to practice: the provision of intrauterine contraception. Obstet Gynecol. 2008; 111: 1359–1369
10. Higgins, J.A. Celebration meets caution: LARC’s boons, potential busts, and the benefits of a reproductive justice approach. Contraception. 2014; 89: 237–241
11. Gomez, A.M., Fuentes, L., and Allina, A. Women or LARC first? Reproductive autonomy and the promotion of long-acting reversible contraceptive methods. Perspect Sex Reprod Health. 2014; 46: 171–175
12. Lessard, L.N., Karasek, D., Ma, S., Darney, P., Deardorff, J., Lahiff, M. et al. Contraceptive features preferred by women at high risk of unintended pregnancy. Perspect Sex Reprod Health. 2012; 44: 194–200
13. Dehlendorf, C., Levy, K., Kelley, A., Grumbach, K., and Steinauer, J. Women’s preferences for contraceptive counseling and decision making. Contraception. 2013; 88: 250–256
14. Dehlendorf, C., Diedrich, J, Drey, E, Postone, A, and Steinauer, J. Preferences for decision-making about contraception and general health care among reproductive age women at an abortion clinic. Patient Educ Couns.2010; 81: 343–348
15. Abdel-Tawab, N. and Roter, D. The relevance of client-centered communication to family planning settings in developing countries: lessons from the Egyptian experience. Soc Sci Med. 2002; 54: 1357–1368
16. Kalmuss, D., Davidson, AR, Cushman, LF, Heartwell, S, and Rulin, M. Determinants of early implant discontinuation among low-income women. Fam Plann Perspect. 1996; 28: 256–260
17. Casalino, L.P. The unintended consequences of measuring quality on the quality of medical care. N Engl J Med. 1999; 341: 1147–1150
18. McDonald, R. and Roland, M. Pay for performance in primary care in England and California: comparison of unintended consequences. Ann Fam Med. 2009; 7: 121–127
19. Omoigui, N.A., Miller, DP, Brown, KJ, Annan, K, Cosgrove, D 3rd, Lytle, B et al. Outmigration for coronary bypass surgery in an era of public dissemination of clinical outcomes. Circulation. 1996; 93: 27–33
20. Welker, J.A., Huston, M., and McCue, J.D. Antibiotic timing and errors in diagnosing pneumonia. Arch Intern Med. 2008; 168: 351–356
21. Kanwar, M., Brar, N., Khatib, R., and Fakih, M.G. Misdiagnosis of community-acquired pneumonia and inappropriate utilization of antibiotics: side effects of the 4-h antibiotic administration rule. Chest. 2007; 131:1865–1869
22. Baker, D.W. and Qaseem, A. Evidence-based performance measures: preventing unintended consequences of quality measurement. Ann Intern Med. 2011; 155: 638–640
23. Stern, A.M. Sterilized in the name of public health: race, immigration, and reproductive control in modern California. Am J Public Health. 2005; 95: 1128–1138
24. Arnow, R.S. The implantation of rights: an argument for unconditionally funded Norplant removal. Berk Women’s Law J. 1996; 11: 19–48
25. ACLU. Norplant: a new contraceptive with potential for abuse. (January 27, 2015; Available from: https://www.aclu.org/reproductive-freedom/norplant-new-contraceptive-potential-abuse; 1994.
26. Johnson, C.G. Female inmates sterilized in California prisons without approval. in: C.f.I. Reporting (Ed.) Center for Investigative Reporting, ; 2013
27. Thorburn, S. and Bogart, L.M. Conspiracy beliefs about birth control: barriers to pregnancy prevention among African Americans of reproductive age. Health Educ Behav. 2005; 32: 474–487
28. Yee, L.M. and Simon, M.A. Perceptions of coercion, discrimination and other negative experiences in postpartum contraceptive counseling for low-income minority women. J Health Care Poor Underserved. 2011; 22:1387–1400
29. Dehlendorf, C., Ruskin, R., Grumbach, K., Vittinghoff, E., Bibbins-Domingo, K., Schillinger, D. et al.Recommendations for intrauterine contraception: a randomized trial of the effects of patients’ race/ethnicity and socioeconomic status. Am J Obstet Gynecol. 2010; 203: 319.e1–319.e8
30. Anhang Price, R., Elliott, M.N., Zaslavsky, A.M., Hays, R.D., Lehrman, W.G., Rybowski, L. et al. Examining the role of patient experience surveys in measuring health care quality. Med Care Res Rev. 2014; 71: 522–524
31. Doyle, C., Lennox, L., and Bell, D. A systematic review of evidence on the links between patient experience and clinical safety and effectiveness. BMJ Open. 2013; 3
32. Rosenberg, M.J., Waugh, M.S., and Burnhill, M.S. Compliance, counseling and satisfaction with oral contraceptives: a prospective evaluation. Fam Plann Perspect. 1998; 30: 89–92 ([104])
33. Forrest, J.D. and Frost, J.J. The family planning attitudes and experiences of low-income women. Fam Plann Perspect. 1996; 28: 246–255 ([277])
34. RamaRao, S., Lacuesta, M., Costello, M., Pangolibay, B., and Jones, H. The link between quality of care and contraceptive use. Int Fam Plan Perspect. 2003; 29: 76–83
35. Sanogo, D., RamaRao, S., Jones, H., N’Diaye, P., M’Bow, B., and Diop, C.B. Improving quality of care and use of contraceptives in Senegal. Afr J Reprod Health. 2003; 7: 57–73
36. Arrowsmith, M.E., Majeed, A., Lee, J.T., and Saxena, S. Impact of pay for performance on prescribing of long-acting reversible contraception in primary care: an interrupted time series study. PLoS One. 2014; 9: e92205
37. Long-acting reversible contraception: implants and intrauterine devices: ACOG practice bulletin no. 121.Obstet Gynecol. 2011; 118: 184–196
38. Personal communication. Lorretta Gavin. ; 2014
39. Bellanca, H.K. and Hunter, M.S. ONE KEY QUESTION(R): preventive reproductive health is part of high quality primary care. Contraception. 2013; 88: 3–6
40. Committee, Q.T. IHE quality, research and public health technical framework supplement: family planning.(January 30, 2015; Available from: http://www.ihe.net/uploadedFiles/Documents/QRPH/IHE_QRPH_Suppl_FP.pdf;2014.