Approximately 7 in 10 family planning specialists reported a contraceptive risk event during their professional careers when pregnancy prevention was desired. While most participants were over a decade into their careers, 29 (13%) reported a risk event within the past year. These data show that even in the context of significant knowledge and high uptake of the most effective methods, risk of unintended pregnancy persists, underscoring the need for robust abortion access.

Our findings parallel metrics of similar contraceptive risks events in the general public. In an analysis of a national population of reproductive-aged women in 2015, 23% reported prior emergency contraception use, less than in our sample (35%) [12], which may be explained by improved access to, knowledge of, and comfort with reporting use of this contraceptive method among family planning specialists. This reported higher use among our study population may also be impacted by the measure of ever use since training (including older individuals, not just reproductiveaged) and ongoing increases in use since 2015 facilitated by lower costs and easier acquisition of emergency contraception. Participants’ report of under- or unprotected intercourse was similar to findings from a survey administered in 2014 to family planning specialists using the same definition: 76% lifetime risk and 7% past-year risk. Our ever-risk is likely lower because our query was limited to time since training commenced; our past-year risk of 13% may be accounted for by omission of withdrawal from the comparative study’s figure [4]. Regardless, among the family planning clinician population, contraceptive risks have been, and continue to be, part of the lived experience after the initiation of medical training. We measured perceived failure rather than pregnancy incidence. In a population of users highly trained to identify failure like incorrect or inconsistent use or device expulsion, capturing the potential for pregnancy may better address our research question than the overestimated performance deduced from clinically recognizable pregnancy used to calculate Pearl indices. Consequently, we refrain from situating our final metric of contraceptive failure in the context of the general typical use effectiveness measuring pregnancy incidence.

The most common methods used among both participants and the US population include oral contraception, external condoms, and intrauterine devices (IUDs), with a higher rate of IUD use among our participants compared to the general US population [13]. The hormonal IUD was the most common method still being used with the highest rate of discontinuation for planned conception. Individuals remain at risk of pregnancy, unsurprisingly, even with perfect use of contraception; multiple participants described experiences of IUD failures. However, it is unrealistic and unforgiving to expect that anyone—including reproductive health experts—will have perfect contraceptive use at each sexual encounter for a multitude of reasons, including the shortcomings of currently available contraceptive methods. Problems with access and adverse effects were infrequently reported reasons for discontinuation of a method in this cohort. Similarly, in the general population, side effects among oral contraception users have been reported as absent or mild, with minimal method discontinuation attributable to side effects [14]. However, side effects were a common reason for method avoidance in our study, particularly for the injection, nonhormonal IUD and implant. Prior study has found that among first time contraceptive users, nearly half were worried about side effects before starting contraception [9]; however, the degree to which these concerns have contributed to method avoidance among the general public is not clear. Given the unique expertise of family planning clinicians, extensive knowledge around potential side effects across methods likely contributed to informed decision-making and method avoidance. Participants’ report of side effects had overlap with other studies including bleeding and interference with sexual pleasure; although, based on write-in responses, weight and mood concerns were underrepresented in this population [9].

Participants also reported development of contraindications. In other studies, up to one third of individuals using combined oral contraceptives reported a relative or absolute contraindication to use, due to medical comorbidities [15, 16]. The high prevalence of these comorbidities may limit the number of contraceptive options safely available to many pregnancy-capable individuals. Notably, participants echoed the sentiments of many other contraceptive users in emphasizing the importance of control over the method – rather than reliance on a partner for use or a clinician for initiation or discontinuation [17,18,19]. As is the case in all populations, there are a diverse set of factors contributing to the (un)desirability of a contraceptive method, again highlighting that effectiveness is not the only metric influencing contraceptive decision-making. This is consistent with other work demonstrating that the contraceptive decision-making process is often a dynamic and nuanced process that changes over the course of decades [20]. Contraceptive decision-making changes with changing bodies, belief systems, environments and relationships [21].

In examining a population with a unique knowledge base and likely excellent access to contraception, including long-acting reversible methods, contraceptive risk events are common over the course of individuals’ professional lives, as is method discontinuation (for reasons other than conception) and method avoidance. These findings normalize contraceptive risk behaviors, emphasize that “typical use” describes use among all contraceptive users, and highlights the narrow range of contraceptive choice when accounting for method contraindications, performance features, and evolving user preferences. These findings work to dismantle the idea of an ideal contraceptive method or contraceptive user in an era characterized by intolerance of undesired pregnancy and loss of abortion access. Such considerations factor into clinical care, by, for example, reducing “otherization” in contraceptive counseling, building empathy for contraceptive dissatisfaction, and expanding the image of potential abortion beneficiaries to everyone. More tangibly, this translates to provision of universal guidance and access to emergency contraception, counseling on the reality of contraceptive switching and discontinuation for many users, and consideration of the inclusion of abortion counseling with contraceptive counseling [22].

These data have implications for the contemporary social and environmental factors affecting sexual and reproductive health by highlighting contraceptive shortcomings and events representing potential abortion need. Further exploration of contraceptive dissatisfaction may facilitate public understanding of the limitations of contraceptive technology and the demands put on pregnancy-capable people in navigating method use. These findings also emphasize the need for expansion of contraceptive options with critical research focused on development of novel agents and delivery systems, including male hormonal contraceptive methods [23].

The strengths of this study include its unique insight into contraceptive risk behavior and contraceptive choices among family planning specialists using quantitative input. These are salient data for generating a response to current questions around the role of contraception, particularly as it pertains to abortion need. Our study is limited by a design that did not allow for a comparison between contraceptive risk event and method at the time of event. However, the focus of this study was on the prevalence of risk in a population with access to and knowledge about all contraceptive options; the relevance of method data was intentionally focused on exploring imperfections of current technology. Our survey did not fully explore the adverse impacts of each individual method. Finally, our response rate, while consistent with or better than most online surveys, may be subject to non-response bias, including the possibility of preferential response among those with a specific interest in sharing their contraceptive risk histories [24]. While demographics of the Society of Family Planning membership are not publicly available data, the geographic diversity of this sample is similar to those in the member directory providing support of generalizability along one dimension.