A health care quality measure goes beyond simple counts of how many patients are being served. A quality measure tells us about the quality of the service they are receiving and thus provides some information on the likely health impact. For example, one current quality measure is the percentage of children age two who receive certain vaccinations. Quality measures exist for many areas of medicine including 700 National Quality Forum endorsed measures. However, there is not yet a measure for contraception.1 A number of others have highlighted the need for one or more quality measures for contraception. For example, the recommendations from the Association of Reproductive Health Professionals’ Sexual and Reproductive Health (SRH) Workforce Summit include “Develop one Healthcare Effectiveness Data and Information Set (HEDIS) measure on SRH” and “Define SRH quality metrics for use in new models of care.”2 A quality measure that affects how much health centers get paid by insurers can encourage providers to focus on contraception and deliver valuable information to managers on how health centers and agencies are performing. Primary care providers in particular have many demands on their time. They understandably allocate their time in part based on financial incentives. Establishing a quality measure for contraception that affects payment is a key tool in the effort to have primary care sites provide higher quality contraceptive services.
Measuring social impact
A quality measure is also important for management decisions about how to maximize social impact. These decisions face health care organizations and donors both in government and at private foundations. The Hewlett Foundation has a particular interest in measuring impact as we describe our grant making approach as “outcome-focused grant making”.3 We attempt to specifically articulate the outcome we are trying to achieve and to measure whether our investments are making progress towards that outcome. Health centers exist in large part to achieve social impacts such as improving health. Health centers also need to consider the financial bottom line. However, it is essential that managers consider the impact of financial decisions on the achievement of social impact. If we want to make good management decisions, we need clear measures of social impact. A quality measure for contraception could enable health center managers to track the impact of financial and other operational decisions on the health center’s social impact. However, few health centers use social impact measures because they are difficult to calculate. The problem is that the ultimate outcomes we are trying to change such as unintended pregnancy are difficult to measure.
The contraceptive protection index
The Contraceptive Protection Index offers a promising solution.4 Although it does not measure the incidence of the condition we are trying to prevent (unintended pregnancy), it does measure the delivery of the preventive treatment, effective contraception. Measuring whether the effective treatment is being provided to those in need is one way to measure quality. The contraceptive protection index is a weighted average. It is calculated by multiplying the typical use efficacy rates for each contraceptive method (how well does the method protect against pregnancy in the real world) by the percentage of women using each method. This index is an improvement over simply measuring the percentage of women who use intrauterine devices (IUDs) and implants because it also takes into account the efficacy of the methods used by those women not using IUDs and implants. Pill and condom users are generally the majority of users at most health centers, so a measure of IUD and implant use alone obscures what impact most health centers are having on the majority of their clients.
One can calculate the St. Louis Contraceptive CHOICE Project’s contraceptive protection index by starting with the percentage of women in that project who chose each method of contraception, for example 46% chose the levonorgestrel IUD.5 The percentage of women using each method is then multiplied by the typical use efficacy rate for that method, for example 46% * 99.8% for the levonorgestrel IUD.6 These efficacy rates for each method, now having been weighted by the percentage of women using each method, are next added together to create a weighted average, the contraceptive protection index, of 97.8%. This calculation is based on the methods used by those women who received contraception because in the CHOICE project all women received a method of contraception. An improvement would be to calculate the index for all women served by a given health center who are not planning to get pregnant in the next year. This would likely include some women who left the health center with no method despite their desire to avoid pregnancy. It is essential that these women are captured in the index so that it reflects all women who are not planning to get pregnant served by a particular health center or in a given population. A 2006 report from the Guttmacher Institute, Estimating the Impact of Expanding Medicaid Eligibility for Family Planning Services, includes such a calculation.7 This report indicates a possible increase in the contraceptive protection index from 72.6 to 91.9 among women who would become eligible for no-cost contraception if Medicaid were expanded.
Quality not quantity
While serving more women has some benefits, it may not always mean more impact. Imagine one health center that provides oral contraceptives to 1000 women in a month and another health center that sees only 900 but provides them with IUDs. More women served may seem like more impact. But if part of the goal of the health center is to help women avoid unintended pregnancy, then the health center that saw fewer women but provided them with a far more effective contraceptive method is the one that achieved more impact.
Putting the index into practice
Consider calculating the contraceptive protection index over time for your own health center and see if quality is improving and how you compare to your peers. Multiply the proportion of women served by your health center using each contraceptive method by that method’s typical use efficacy rate and sum the results. The analysis can be done for all women currently using a method or all women who do not wish to become pregnant in the next year if those data are available. If you are involved in a research intervention aimed at improving contraceptive use, consider including the index as an outcome in your results so that your intervention’s impact can be more easily compared with others.
If one of the goals of health centers that provide family planning services is to reduce unintended pregnancy, we must measure not only the quantity of clients served but also the quality and effectiveness of the services delivered. If we want health centers that do not focus on family planning to provide quality family planning services, we need a quality measure for contraception that is tied to payment. The contraceptive protection index would be an excellent place to start.
- National Quality Forum . Field guide to NQF resources.
- Nothnagle M, Cappiello J, Taylor D. Sexual and reproductive health workforce summit report. Contraception. 2013;88:204–209
- The William and Flora Hewlett Foundation . Our approach to philanthropy.
- Sonfield A, Frost JJ, Gold RB. Estimating the impact of expanding Medicaid eligibility for family planning services: 2011 update, New York: Guttmacher Institute, 2011 p. 38.
- Peipert JF, Madden T, Allsworth JE, Secura GM. Preventing Unintended Pregnancies by Providing No-Cost Contraception, Obstetrics & Gynecology, Vol. 120, No. 6, December 2012, Table 2.
- Trussell J. Contraceptive efficacy. In: Hatcher RA, Trussell J, Nelson AL, Cates W, Kowal D, Policar M editor. Contraceptive Techology: Twentieth Revised Edition. New York NY: Ardent Media; 2011;
- Sonfield A, Frost JJ, Gold RB. Estimating the impact of expanding Medicaid eligibility for family planning services: 2011 update, New York: Guttmacher Institute, 2011, Table A9 p. 55.