Despite enormous effort and billions of dollars of expense, the vast majority of the published biomedical literature is not useful, meaning that it does not impact decision-making favorably for either healthcare consumers or providers. That is the reason that creative and impactful research is so valuable. In this issue of the Journal, Hubacher et al report their innovative research that assesses whether long-acting user-independent contraceptive methods are truly more effective at preventing unintended pregnancy than short-acting, user-dependent methods. This and another article in this issue inform and challenge a current central tenet of family planning research: that increasing access to long-acting reversible contraception (LARC), intrauterine devices (IUDs), and implants is the most efficient and cost effective way to improve the US public health crisis of unintended pregnancy.

The LARC-focused family planning practice is based on the results of large studies conducted in St. Louis, MO, and statewide initiatives in Colorado and Iowa. These studies demonstrated dramatic reductions in unintended pregnancy, teen pregnancy, and abortions with cost-free access to IUDs and implants. The Contraceptive CHOICE Project may be the contraceptive research project of our time. It meticulously has described, in >60 articles thus far, the outcomes of 9256 women who received counseling that focused on the benefits of LARC methods and who were offered no-cost access to short- and long-acting reversible methods of contraception for 2-3 years. The primary findings were that 75% of the women chose LARC methods and that contraceptive pill, patch, or ring users were >20 times more likely to have an unintended pregnancy than LARC users. Moreover, teenage study participants had pregnancy and abortion rates less than one half that of St. Louis residents and the nation. These impressive results have created a basic assumption in family planning research and care delivery that prioritizes LARC methods.

Based on this information, family planning clinicians and researchers shared the benefits of increasing access to LARC methods with women who came to the clinics and hospitals where they worked, instructed trainees on the benefits of their use, and researched methods to remove barriers to LARC methods. Based on the accumulating evidence that demonstrates the benefits of LARC, I worked hard to increase the supply of devices in my community, where family planning Medicaid expansion is not a reality, so that my uninsured patients and those with financial challenges could obtain LARC methods.

Then advocates sounded resistance to the focus on LARC methods as a cure-all with particular warnings about promoting LARCs over patient-centered care. They cited concerns of autonomy and possible reproductive coercion by providers who over zealously encourage IUDs and implants, which cannot be removed easily without a medical provider. This insight highlights the need to discuss all available contraceptive methods during adequate counseling.

The Hubacher et al and Karpilow and Thomas articles boldly address the “LARC is superior” hypothesis in different and interesting ways. Hubacher et al sought to find out whether a group of women at a family planning clinic who intended to start a new short-acting method (oral contraceptive pills or depo-medroxyprogesterone acetate) would consider a long-acting method (an IUD or implant) and whether they were assigned a LARC method or chose one, how their pregnancy, continuation, and satisfaction rates would compare with those who were assigned to short-acting methods (pills or injection). Essentially, they wanted to find out whether it is just that the methods are more effective or whether there is something about women who choose the more effective methods that contribute to their lower pregnancy risk over time.

The research team came up with an ingenious method to get at this question. They recruited women who were seeking oral contraceptives or depo medroxyprogesterone acetate and randomly assigned those who were willing to be assigned to no-cost short-acting or long-acting methods. Women who were assigned to the short-acting methods group could then pick between oral contraceptives or depo medroxyprogesterone acetate. Those women who were assigned to the long-acting methods group could choose the subdermal implant, levonorgestrel IUD, or a copper IUD. The research team informed participants that they would not be charged if they wanted a device removed at any time during the year of follow up but that anyone who changed methods was responsible for the cost of the new method. Women who were willing to participate but not willing to be assigned randomly could enter a patient preference arm in which they were responsible for the cost of the contraception through private insurance, Medicaid, Title X coverage, or out-of-pocket contributions. When I first heard about this study, I doubted that it would ever be completed. My bias was that women are clear in their desire for contraceptive methods and would not agree to random assignment. Others expressed concern regarding the ethics of randomizing poor women to different methods of contraception when they may not have the ability to obtain more effective methods unless they agreed to study participation. I believed that there was a difference between the women who choose LARC methods and doubted that women would agree to randomization of such different treatments. It might be this difference, rather than LARC itself, that is responsible for the higher effectiveness. This was exactly the hypothesis that the study team sought to assess, and I am glad they did.

Hubacher et al enrolled 916 women; 43% of the women agreed to random assignment. Of the 524 women (57%) who enrolled in the patient preference arm, all but 2 choose short-acting methods, which supported the supposition that women at these clinics may have been clear about their desire for pills and the injection and may have been unable to pay for IUDs and implants if they were not willing to undergo random assignment. The 12-month method continuation rates were highest for women who were assigned randomized to long-acting methods (77.8%), compared with 63.3% for women who were assigned randomly to short-acting methods, and with 53.0% for women who selected short-acting methods in the patient preference arm. The group’s respective unintended pregnancy rates over the year were 0.7%, 7.7%, and 6.4%. The satisfaction assessments also revealed interesting findings. First, women with the expectation of getting pills or depo medroxyprogesterone acetate were highly satisfied when they got an IUD or implant. It is less surprising that women who were assigned randomly to an IUD or implant were significantly less likely to have a pregnancy. A particularly interesting and telling finding was that those who chose random assignment were asked after assignment whether they had hoped for short- or long-acting contraception. The answer was that the assignment did not matter if the product was free.

The Karpilow et al study used a sophisticated model that was informed by data from the National Survey of Family Growth and other national databases to assess pregnancy rates in 3 simulated populations that are sociodemographically similar to CHOICE participants: (1) women who used the pre-enrollment method of CHOICE participants, including those who used neither contraceptive methods nor condoms, (2) women who received the CHOICE intervention where 75% of the women used LARC methods and 25% of the women used the pill, patch, or ring, and (3) women who transitioned from no method or condoms to the pill, patch, or ring. The model output shows that three-quarters of the pregnancy reduction that was achieved by the CHOICE intervention could be achieved by just transitioning women who used no method or condoms to the pill, patch, or ring. The Karpilow and Thomas models that clearly switching from nothing to something is beneficial. However, it does not acknowledge the frustration and lack of satisfaction that women have with contraception, especially hormonal methods, where commonly less than one-half of the women are still using the method after 1 year.

The question for clinicians becomes, how do we use the information from these two studies? We certainly cannot interpret the Hubacher et al article to say that women who did not want LARC initially got it and in the end were satisfied, so you should get it also. Nor can we turn to the Karpilow and Thomas article to say that we do not even need access to LARC methods because we can gain most of the benefits just by offering more women pills, patches, and rings.

By combining the wisdom of my colleagues who are critical of LARC-focused care, who use patient autonomy as a foremost guiding principal, and these study results, we can improve contraceptive counseling and care. Most importantly, we should focus our efforts where they matter most: on our patients. Patient-based counseling helps women choose a contraceptive method that fits their personal priorities the best. An excellent starting point ideally is beginning the contraceptive conversation with a simple question such as, “What is important in a method for you?” Adding the study results discussed here, we can inform women who enter the clinic with a preference for pills that, if they choose an IUD or implant, they are likely not only to continue the method, be satisfied, and avoid unintended pregnancy but also, importantly, that choosing some kind of contraception is far better than none or condoms. An additional modest, but meaningful, gain in lowering the pregnancy rate is made by including LARC methods. The efficacy of LARC is not just because women who choose LARC are different in some key way from women that choose short acting methods but is, in fact, due to the LARC itself. Finally, providing free access to all methods of contraception permits true autonomy in decision-making and will provide the greatest gains in lowering the pregnancy rate. It turns out that our job as medical providers is to do just that, provide patient-centered medical care; when we provide our patients with what they want and the care that they believe will best meet their needs, they will be satisfied and have the best outcomes.