Medication decision making under uncertainty: Considering chronic disease and pregnancy potential

Abstract

People with chronic illness who are (or could become) pregnant and their clinicians may face difficult decisions about the use of medications that can affect fetal health and development. In these situations, people must make the choice whether to use a drug with limited information about teratogenicity to treat their illness, putting an existing or potential pregnancy at risk. This manuscript serves as a call to action for all clinicians whose patients face such treatment choices regarding teratogenic medications or medications with unknown fetal safety. It offers an approach to guiding patient-provider conversations on medication management for individuals who are or might become pregnant, grounded in a review of the existing literature on medication decision making for drugs with known or potential teratogenic properties and informed by the current tools available for offering decision support. As a case study for highlighting current challenges and potential solutions around safe prescribing, we focus on two systemic autoimmune and rheumatic diseases, systemic lupus erythematosus and rheumatoid arthritis, which disproportionately affect females compared to males (with incidence ratios of 9:1 and 3:1, respectively). We address how recent reproductive policy changes that restrict abortion access may affect both a provider and a patient’s decision calculus. Finally, we highlight several gaps that could be addressed by future research and the implementation of new data collection practices by federal bodies, while offering an approach to optimize patient-provider communication under conditions of ongoing uncertainty around the safety of drugs for those who may become pregnant while using them.

Introduction

Approximately 75% of adult females in the U.S. use one or more medications during pregnancy,¹ many of which can cross the placenta and enter the fetal circulation.¹ Medication use during pregnancy appears to be rising, in part due to an increasing prevalence of chronic and complex medical comorbidities among people of reproductive age.² To make a fully informed decision to initiate or continue a course of medication, clinicians and patients who are or may become pregnant must evaluate whether the potential benefits of the treatment will outweigh the potential risks, considering the available evidence of safety and the patient’s own needs, preferences, and values.

There are several challenges facing clinicians and patients when making medication decisions in the context of pregnancy. Approximately 12 million known teratogens—medications that can cause congenital anomalies– are prescribed to non-pregnant females of reproductive age.³ If these individuals find themselves pregnant, they are faced with the first decision: whether to continue or end the pregnancy, if this option is available to them and aligns with their values. Depending on their decision, they may then have to change their course of treatment. Similar medication choices are also faced by individuals, whose diagnosis occurs in the postpartum period, as the same drugs that are contraindicated for pregnancy are often contraindicated for lactation.

Due to limitations in the availability and quality of evidence, as well as the presentation of drug safety data, patients may find medication decision-making in the context of pregnancy to be confusing and difficult and their providers may not know how to best advise them. Some individuals will assume unnecessary risks and/or harms by ceasing critical treatments, while others will inadvertently expose themselves to medications with teratogenic properties due to inaccurate perceptions of the fetal risk. When the requisite risk-benefit evidence is limited, individuals may prioritize whatever element of a communication is most salient (e.g. the default warning on many medications of “If pregnant or breastfeeding, ask a health professional before use”).⁴ Some may infer inaccurate meaning from labels^5,6 assuming that the directive to ask a health professional indicates proof of harm for pregnancy or breastfeeding, while others may overweigh small probabilities of harm in contrast to similar or greater sized benefits.^7–10 This overweighting can happen for several reasons, including the availability heuristic, whereby the judged likelihood of an event is based on how easily an instance of that event comes to mind (e.g. recalling a devastating story of a pregnancy loss) and the “possibility effect,” whereby people are significantly more sensitive to changes in occurrence near a 0% or 100% chance (i.e. no chance of teratogenic exposure to a minuscule chance). The factors included in a medication decision may also extend beyond the physiological benefits of reduced disease burden and risks of medication side effects, including factors like financial cost, quality of life, and the personal values of both the individual and influential others, such as family members.

Medication decision-making for pregnant or potentially pregnant people arguably became even more challenging in June 2022, when the U.S. Supreme Court overturned Roe v. Wade in favor of the Dobbs v. Jackson Women’s Health decision,¹¹ thereby abolishing the federal right to abortion. While rarely discussed in research studies or the media, the ability to terminate a pregnancy has served as a shield from the sequelae of fetal exposures to teratogenic medications prescribed for chronic disease. These reproductive rights to care enabled some pregnant individuals exposed to teratogens or medications with teratogenic potential to terminate a pregnancy with severe fetal congenital anomalies; clinicians to avoid the professional and legal consequences of prescribing^12–14 medications with known or unclear fetal risk to a patient who (inadvertently) becomes pregnant; and taxpayers and parents to avoid the financial costs of long-term care for children with congenital anomalies.¹⁵ The Dobbs decision provides a profound and urgent stimulus to augment research and safe medication prescribing for millions of U.S. people who are or may become pregnant.

Establishing safety profiles for medication use during conception, pregnancy or lactation: current processes and paradigms

Beginning in the 1970s, in the wake of the Thalidomide tragedy,¹⁶ most pregnant people were excluded from early-phase clinical research studies due to ethical concerns about fetal safety; however, these exclusions paradoxically placed pregnant individuals at greater risk of adverse health outcomes by undermining an evidence base for guiding peripartum treatment options. In 2018, revisions to The Common Rule changed definitions around vulnerable populations in research, and the FDA has since explicitly encouraged the inclusion of pregnant individuals in clinical trials, as failure to do so may ultimately compromise the health of the pregnant individuals and their fetuses.¹⁷ Pregnancy itself does not render an individual vulnerable. Yet, academic researchers and pharmaceutical companies remain disincentivized from including pregnant people in clinical trials. Outdated beliefs that individuals who are or could become pregnant remain a vulnerable, protected class in research studies persist, along with concerns about liability if a treatment is found to have fetal toxicity. Pharmacodynamic and pharmacokinetic (PD/PK) studies, which ideally enroll individuals at different stages of pregnancy to optimize dosing parameters that reduce fetal toxicity, are associated with significant costs and recruitment challenges. Moreover, for industry, the financial gains of establishing safety during pregnancy are uncertain, as pregnancy is a relatively brief state in the broader time course of chronic disease. Even in existing drug safety studies, inconsistent clinical and surrogate endpoints across studies preclude head-to-head safety comparisons between treatments. For example, medication risk/benefit ratios may not clearly identify the beneficiary of a treatment (e.g., patient or the fetus, or both) or define potential treatment benefits or risks consistently across studies (e.g., fetal malformations, 5-year child developmental index, maternal health indices). Ultimately, drug safety trials may not provide the data needed by a patient or prescriber to support informed and preference-sensitive treatment decisions in the context of reproduction.

Limited federal and private funding sources are available to support the conduct of rigorous drug studies that involve pregnant people or independent pregnancy exposure registries, with some notable exceptions discussed herein. At present, most information about drug safety in pregnancy is collected via post-marketing observational studies, including pregnancy exposure registries maintained by pharmaceutical companies. Pregnancy exposure registries are often statistically underpowered due to reliance on clinicians who are willing or able to share patient data and on patient volunteers. As some of these patients inadvertently became pregnant while using a medication of interest or took a medication with unknown safety throughout pregnancy due to limited availability of safe or tolerable alternatives, the patients included in these registries may not be representative of a particular population of interest.

The FDA has sought to provide greater clarity to the public about the known risks and benefits of medications, including data on pregnancy and lactation. In the past, a pregnancy risk categorization scheme assigned a simple letter grade (A, B, C, D, or X) to medications that indicated their compatibility with pregnancy and lactation based on available human and animal studies; however, this system was felt to be overly simplistic and consequently easily misinterpreted by clinicians and patients alike as a graded system. In 2015, FDA developed a Pregnancy and Lactation Labeling Rule (PLLR) that called for the development of narrative summaries for individual medications, which emphasize human studies and synthesize knowledge about the pregnancy risks.^18–21 The impact of the PLLR on safe prescribing practices remains unclear, with only approximately 20% of new drug labels containing any human data on pregnancy or lactation.²² In addition, limited studies suggest that many prescribers are not familiar with the PLLR.¹⁹ In one study, prescribers who did find the more detailed risk and disease information in the PLLR helpful still preferred the older, simpler categorization scheme given its concise, directive, and clear format. Similarly, studies suggest that patients would benefit from clear, interpretive, and, when possible, quantified estimates of a drug’s benefits and risks to guide their own decision making and medication discussions with providers.²³

Highlighting the complexities around safe medication prescribing: case examples of systemic lupus erythematosus and rheumatoid arthritis

The care of people with systemic autoimmune and rheumatic diseases (SARDs) exposes complexities that have led to widespread prescription of teratogens and medications with unknown pregnancy safety for people who are or could become pregnant. SLE and RA are chronic, disabling, and potentially organ-threatening diseases that can worsen in the context of pregnancy—particularly SLE, which is associated with a risk of maternal mortality that is 5 to 20 times greater than that of the general population.^{24, 25} Medications are often required in the preconception period and throughout pregnancy to control SLE and RA disease activity. In this context, medication decisions are not neutral – undertreated disease can threaten both maternal and fetal health, while teratogenic medications that bolster maternal health directly threaten fetal health.

Medications with known teratogenicity or unknown pregnancy compatibility are widely prescribed to people with SARDs. Approximately 44.5% of reproductive-age females with RA and 30% of reproductive-age females with SLE are prescribed at least one teratogenic medication, usually methotrexate or mycophenolate mofetil/mycophenolic acid.²⁶ Some teratogens are uniquely effective in the treatment of systemic autoimmune and inflammatory diseases. The gold standard therapies for lupus nephritis (i.e., cyclophosphamide, mycophenolate) - an organ-threatening condition that affects 40% of patients with SLE – are teratogens.^27,28 Some teratogens are less expensive or easier to access than pregnancy-compatible treatment options. For example, the teratogen methotrexate, which is the gold standard for RA care and is used to treat many other SARDs, costs $102 per year, as compared to $30,000-$50,000 in annual costs for pregnancy-compatible tumor-necrosis factor inhibitors (TNFi) and biosimilar agents.²⁹ In fact, insurance prior authorizations for pregnancy-safe TNFi are often refused for non-pregnant patients with RA, who have not already used or failed methotrexate, forcing some to use a teratogenic medication for disease control irrespective of their plans for pregnancy. Explosive growth in new immunomodulatory drugs has heralded effective novel pathways in treating disease. Again, however, few of these medications have proven safety or efficacy during pregnancy. Most, in fact, contain warnings to discontinue before or upon confirmation of pregnancy, likely because of scant data about their safety in pregnancy.

Structural and social influences on health also affect drug prescribing patterns, medication decision-making, and medication safety data that reflect the information needs of the public. Data from the Rheumatology Informatics System for Effectiveness (RISE) registry, which collects data from 30% of rheumatology practices nationwide, indicate that reproductive-age Black, Asian, and Hispanic patients with SLE and RA are more likely to be prescribed teratogenic medications than are white patients (e.g., methotrexate, mycophenolate) .²⁶ Other work has found that Black Americans with RA are significantly less likely to obtain biologic disease-modifying anti-rheumatic drugs, often regardless of insurance status, including those with demonstrated safety in pregnancy.³⁰ Adult females of color are also significantly less likely than white females to participate in clinical trials to establish drug safety, further limiting the evidence base. A qualitative study used critical race theory to understand how Black individuals with SLE perceived participation in clinical trials, found that factors that negated participation included mistrust due to racism, concerns about how individuals were assigned to placebo groups, and culturally insensitive marketing of trials.^28,31–33 With data lacking on medication treatment and side effects among diverse populations, some patients with RA and SLE may feel that the available safety data are not relevant to their particular health needs.

The Dobbs decision may also have a profound impact on prescribing patterns of efficacious and cost-effective teratogens in rheumatology care. Only approximately one-third of reproductive-age females with SARDs use contraception.³⁴ Despite a lack of high-quality data to estimate the incidence of fetal teratogen exposures in rheumatology, studies confirm that these exposures do occur. For example, in a small national survey of reproductive-age females with RA, SLE, and other inflammatory arthridities (e.g., psoriatic arthritis) who had experienced at least one pregnancy, 29% who had been prescribed methotrexate had become pregnant—usually inadvertently– while using that drug.³⁴ In a separate national survey of U.S. rheumatologists, approximately half reported that prior to the Dobbs decision, they had cared for at least one patient who had become pregnant while using a teratogen.³⁵ Approximately one-third of those rheumatologists subsequently recommended termination of pregnancy, because of fetal teratogen exposures and/or disease severity that threatened maternal health (71% and 69%, respectively). Among rheumatologists who practiced in abortion-restricted states, 83.1% felt comfortable referring a patient for a termination under the protections of Roe, but only 34.8% felt comfortable doing so after the Dobbs decision. Importantly, the Dobbs decision may also affect the care of non-pregnant, reproductive-age persons with SARDs. Of rheumatologists practicing in abortion-restricted states, 13% changed or planned to change their prescribing of methotrexate and 8.7% changed or planned to change their prescribing of mycophenolate post-Dobbs, as compared to only 5.3% and 1.3%, respectively, of rheumatologists practicing in abortion-protected states. These findings, captured in the first few months after the Dobbs decision, must be updated given the rapidly evolving state policy environments (e.g., judicial injunctions, constitutional amendments), but suggest that prescribing patterns and risk calculations around teratogens for pregnant and non-pregnant individuals may change in rheumatology care and likely other clinical contexts in response to Dobbs-era policies. Such changes may particularly impact reproductive-aged females with SARDs that identify as members of socially marginalized groups or who are financially disadvantaged, as these individuals have greater need for abortion access and utilization.³⁶

Existing resources that support informed medication decision-making in rheumatology and other clinical contexts for individuals who are or could become pregnant

Patient- and clinician-directed educational resources around medication safety in pregnancy are available for patients, rheumatologists, and other clinicians. For example, the American College of Rheumatology, European League Against Rheumatism, and British Society for Rheumatology have published freely accessible, clinician-directed systematic literature reviews, pamphlets, online resources, and handouts that synthesize pregnancy and lactation safety data for anti-rheumatic and immunomodulatory drugs. The American College of Rheumatology supported a multi-year Reproductive Health Initiative, which led to the development of patient-facing online educational modules around various aspects of pregnancy planning, including medication safety review.³⁷

Several national programs also provide guidance around safe prescribing of anti-rheumatic drugs, as well as other medications used among people who are or could become pregnant. The Mother-to-Baby program (https://mothertobaby.org) ,^{38, 39} developed by the nonprofit Office for Teratology Information Specialists (OTIS), provides free online educational materials on drug safety in pregnancy—many of which are anti-rheumatic drugs – including a patient/clinician informational hotline, a podcast and lactation medication app (associated with the National Institute of Health’s LactMed database,⁴⁰ which provides information on drug and chemical exposures for breastfeeding parents), and registries for pregnant patients exposed to medications with known or potential teratogenicity. Similar to OTIS, InfantRisk Center (https://infantrisk.com/) is a nonprofit center that provides information about the infant safety risks of medication exposures through lactation.^41,42 InfantRisk provides medication safety fact sheets, a call center for patients and providers to query about medication and breastfeeding safety, and a breastmilk registry that evaluates medication transfer into breastmilk. Both OTIS pregnancy registries and InfantRisk Center have shown effectiveness at quantifying the safety of certain medications during pregnancy and lactation.

To increase the baseline safety of initial prescribing, including for those considering pregnancy, FDA authorized widespread Risk and Evaluation Mitigation Strategies (REMS) programs in 2007. REMS programs provide a multi-faceted approach to ensure that certain drugs are prescribed only when the likely benefit outweighs the risks. When FDA deems a REMS program necessary for a specific drug, the pharmaceutical company must provide a patient-directed medication guide available to patients at the point of drug dispensing, a plan for communication outreach to clinicians and professional societies, and sometimes more restrictive measures, such as limited dispensing plans and certifications for pharmacies and healthcare settings. For drugs with teratogenic properties, some REMS requirements have included that patients must pledge to use birth control or abstain from sex, and clinicians must document negative pregnancy tests before prescribing or refilling a teratogen.^1,8 The REMS program for mycophenolate (www.mycophenolaterems.com) appears to have increased the rate of counseling around teratogenicity in some rheumatology clinics⁴³ and lowered overall fetal exposure to the drug among those who were pregnant at the point of prescribing. However, it has not clearly reduced the number of pregnancies that occur once mycophenolate is prescribed, or the rate of inadvertent fetal teratogen exposures.⁴⁴ In fact, 1 in 16 pregnancies (17,800 pregnancies in 2017 alone) occur among those individuals who used one of the 12 teratogens associated with REMS programs.¹

While each of these resources have achieved national reach in providing essential educational support for patients and clinicians, it is unclear if they reduce the risk of fetal teratogenic medication exposures. Moreover, these programs were not developed to explicitly address the root causes of behaviors and practices that lead to inadvertent fetal teratogenic medication exposures, which may include gaps in patients’ contraceptive access or clinicians’ assumptions about a patient’s pregnancy intentions.²

Considering socially marginalized groups

Supporting shared decision making around medication use

To make an informed choice about medication use, individuals must be able to weigh the key decision factors that play a role for them, including risks that extend beyond side effects (such as financial hardship) and benefits that extend beyond health maintenance (such as achieving reproductive goals). Shared decision-making (SDM) is a process in which patients and clinicians review the best evidence available on the risks, benefits, and alternatives to a course of treatment or medical tests so that the patient has agency to make an informed, value-concordant decision about their own healthcare.^45,46 In the clinical care of those who are or may become pregnant, using a SDM process may facilitate safer prescribing behaviors by clinicians and serve as an opportunity to incorporate multiple decision factors, including a clarification of values around reproductive goals.⁴⁷ SDM is ideal for preference-sensitive decision-making— when more than one option exists and must be deliberated in the context of a patient’s preferences, priorities, and risk tolerance. While predominantly implemented when making initial treatment decisions, SDM can be intentionally used to periodically re-visit long-term prescription use with no anticipated end date, such as methotrexate for those with SLE.^46,48

Decision aids are educational tools (written, audiovisual, or web-based) that can enhance the shared decision-making process by providing patients with critical information on the risks, benefits, and alternatives to a course of treatment or medical test.^49–51 In 2003, the first international standards were established to guide and improve the content, development, implementation, and evaluation of these tools. While, decision aids may be used outside or within the context of the clinical encounter, they always have the goal of supporting a patient in deliberate, shared decision making and have been found to improve clinician satisfaction⁵² in addition to supporting patient choice.

Several decision aids have been developed to support patients with SLE or RA. Table 1 offers a description of those decision aids that focus on or are inclusive of individuals of pregnancy who are or may become pregnant and have published their development and/or evaluation process. Some tools from the literature are not included in this brief review if their primary population was primarily outside the range of typical reproductive age (e.g.⁵³) or if their evaluation was primarily qualitative (e.g.⁵⁴).

Table 1.

Decision aids to support medication for individuals with RA or SLE that are inclusive of individuals who are or may become pregnant.

Decision Aid & year of initial publication	Decision Focus	Population	Description	Theory or framework	Format	Outcomes
Motherhood Decision Aid; 201255–57	Having a child or more children	Individuals of pregnancy capacity with RA	Content on the effects of RA and RA medication on conception, pregnancy, and parenting. Contains worksheets in different formats that individual can use to clarify their values and make a choice (e.g. pros/cons scales, a knowledge checklist)	Ottawa Decision Support Framework58; International Patient Decision Aid Standard (IPDAS) guidelines59	Booklet (45 pages) or web- based	In a randomized study (n=144), those who received the decision aid had higher RA, pregnancy, and parenting knowledge scores and lower decisional conflict than those who received no intervention (control). No significant differences were found in anxiety, depression, self-efficacy, or the degree to which the diagnosis of RA complicated their decision and certainty around having children.
Unnamed Rheumatoid Arthritis Decision Support; 201260, 61	Initiating biologic agents	Individuals with RA who have not been responsive to prior treatment with a non-biologic disease modifying anti-rheumatic drug	Interactive educational modules (with voiceovers) that offers content about 4 different biologic agents. Equivalent amount of information provided about benefits and risks.	Fuzzy trace theory62; IPDAS guidelines	Web-based	In a randomized study (n=125), knowledge of biologics, self-reported clarity of knowledge and clarity of values were all greater among those viewing the decision aid than among those receiving usual care (control). For those who escalated care, there was greater value-concordant choice at 2 weeks in the intervention group compared to the control. No significant differences were found in satisfaction with risk communication.
ANSWER (Animated, Self-serve, Web-based Research Tool); 201463,65,66	Initiating methotrexate	Individuals with newly diagnosed RA.	6 vignettes in animated graphic novel form. Asks patients to rate the importance of potential risks and benefits of using methotrexate for the individual in the vignette. Included a rating of the importance of becoming pregnant (or starting a family).63,66,58	IPDAS guidelines	Web-based	In a pre-/post-intervention study (n=30), ANSWER showed improvements in medication knowledge and decisional conflict 58, 66 after using the tool. No significant differences were found in patients’ perceptions of their ability to manage their
Shared Decision Making in Lupus (SMILE); 201567	Initiating immunosuppressive drug(s).	Individuals with lupus nephritis. Tailored for use primarily by Hispanic/Latine and non-Hispanic Black individuals of pregnancy capacity.	Information tailored to the treatment context of induction (experiencing a current flare with no prior immunosuppressive medication use) or maintenance, (anticipating a future flare when 1 or 2 prior immunosuppressive medications have failed). The decision aid walks the patient through two possible medication options. It is available in both Spanish and English.	IPDAS guidelines	Computer/web-based	In a randomized study (n=301), patients who used SMILE, when compared to a patient group receiving an information pamphlet on lupus from the ACR (control), showed a significant reduction in self-reported decisional conflict. A significantly higher number of patients in the SMILE group reported that the information was excellent for understanding their medication options. No significant differences were found in measures of patient-provider interaction following use of the tool.
ANSWER 2 (Animated, Self-serve, Web-based Research Tool); 201863,–66	Initiating biologic and small molecule agents	Individuals with RA receiving a recommendation to initiate or switch to a new bDMARD or tsDMARD	General risk/benefit information provided on using conventional DMARDS vs. adding or switching to a new agent (bDMARD or tsDMAR) via embedded videos. Tailored risk-benefit information on specific agents then presented; specific agents prioritized by an attribute rating task completed by the patient.	IPDAS guidelines	Web-based	In a pre-/post-intervention study (n=50), ANSWER-2 showed improvements in medication knowledge and decisional conflict after using the tool.64 No significant differences were found in a measure of patients’ self-efficacy.
Unnamed patient decision aid; 2020	Initiating conventional, biologic, or targeted synthetic (DMARDs)	Spanish-speaking individuals who have not been responsive to prior treatment with a nonbiologic disease modifying anti-rheumatic drug	Information about RA followed by a matrix comparing choice options to add or change to another csDMARD; add or change to bDMARD; add or change to tsDMARD. Patients completed a values clarification exercise, which included one question on desire to become a mother in the near future.	IPDAS guidelines	Both paper- based and (interactive, multi-media) web-based	In a pre-/post beta test study (n=54), participants showed decrease in decisional conflict after using the tool.68
Decision Aid & year of initial publication	Decision Focus	Population	Description	Theory or framework	Format	Outcomes
MyVoice:Rheum; 2024	Family planning decisions	Women and people with pregnancy capacity who have systemic autoimmune and rheumatic diseases	Information about medication safety during pregnancy and lactation; genetics and heritability of SARDs (including SLE and RA); contraception options and safety/efficacy; fertility; pregnancy planning and management; and parenting. MyVoice:Rheum also includes family planning decision clarification exercises, and a summary sheet that a patient can present to a rheumatologist to initiate a family planning discussion.	IPDAS guidelines. Self-Determination Theory	Web-based	MyVoice:Rheum was found to be feasible and acceptable in a pilot study. Sixty-one percent of MyVoice:Rheum users as compared to 44% of individuals who reviewed a pamphlet on reproductive health in rheumatology had family planning discussions with their rheumatologists.69

The tools that have been created, to date, highlight the role that SDM can play in knowledge acquisition and clarification of patient values. Almost all tools showed decreases in decisional conflict – an individual’s perceived uncertainty when making a choice—and improvements in knowledge specific to the decision context. However, it is notable that most of these decision aids did not measure or did not improve concordance between knowledge, attitude (or values) and choice; suggesting an additional layer of complexity in decision making that is not being adequately addressed in the risks, benefits, and alternatives being discussed in the existing set of tools. Several decision aids, while inclusive of individuals of who may become pregnant, were intended to support decision making for a broader patient population and focused primarily on side effects and disease progression. Only one offered information tailored specifically to a minoritized population.⁶⁷ Two decision aids explicitly addressed reproductive goals as the focal decision or as a key consideration in medication decision making,⁶⁹ yet did not prioritize medication considerations like cost or medication delivery-type. None explicitly addressed pregnancy termination or how an individual’s decision calculus may change with abortion access (previously driven largely by values, currently driven by values and policy). While the treatments presented across these decision aids ranged in teratogenic properties and degree of evidence of human teratogenicity, this was also not included as a specific decision factor. To our knowledge, there are no decision aids that comprehensively address the complexity of the medication decision making process for individuals who are or can become pregnant. Thus, there is a continued need for decision supports that can communicate risk, benefits, and uncertainty around treatment options, facilitating informed choice around treatment initiation or continuation while considering both reproductive goals and realistic options if an unwitting teratogenic exposure occurs.

Recommendations for patient-provider conversations on medication management for those individuals who are or may become pregnant

Drawing from various frameworks of person-centeredness, health equity, and reproductive justice and autonomy,^70–72 we recommend the following considerations for facilitating incorporation of reproductive goals into medication management SDM, acknowledging the very real time constraints on patient-provider interactions. 1) If any medication option has known or unknown teratogenic potential, it is an ethical imperative to first identify the potential for pregnancy, through an understanding of sexual activity and contraceptive practices and next elicit reproductive goals, which can be done by asking patients open-ended questions about their priorities for pregnancy now and in the future, with recognition that some people will be uncertain about their reproductive goals and preferences.⁷³ This conversation may also need to occur longitudinally, incorporating contraceptive counseling and discussions of a patient’s specific priorities and needs with respect to attributes of contraceptive options. 2) Develop professional familiarity on the evidence base of medications so as to clearly communicate the requisite information that patients need to make an informed decision about their treatment options, including sharing what is known and what is unknown about risk and benefit for pregnancy and lactation. 3) Acknowledge access constraints, which may affect an individual’s medication decision making, including discussion of the cost of medications and potential for insurance authorization, without assumption of a patient’s financial circumstances, and explicitly acknowledging abortion restrictions or access, depending on state of practice. 4) Keep the conversation open by recognizing that people’s situations and preferences may change and allowing for revisiting a decision, including reproductive decision making, throughout the course of care.

Conclusion

This review outlined the uncertainty associated with medication decision making for individuals who are or could become pregnant when the medication options have known or unknown teratogenic properties. The primary reason for uncertainty lies in lack of evidence from clinical trials and gaps in how medication decision making is supported in clinical care. We conclude with a call to action for stakeholders in medication choices for those who are or can become pregnant. Specifically, we call for an improvement in the quality of available data through federal and industry investment in research and clinical trials that include individuals who are pregnant and/or lactating. We call for enhanced drug labels, using the best practices in risk communication. We call for a centering of the voices of marginalized individuals as trials are developed, data are collected, and decision supports are created. Lastly, we believe it is imperative to develop technology or healthcare workflow processes and resources that enable patients to receive the necessary information and values clarification needed to support medication decision-making, preferentially prior to a clinic appointment to facilitate efficiency and avoid gaps in treatment as a medication decision is being made.

Abbreviations:

FDA

Food and Drug Administration

PLLR

Pregnancy and Lactation Labeling Rule

SARDS

Systematic autoimmune and rheumatic diseases

SLE

Systemic lupus erythematosus

RA

Rheumatoid arthritis

OTIS

Office for Teratology Information Specialists

REMS

Risk and Evaluation Mitigation Strategies

SDM

Shared decision-making

Data availability

No data was used for the research described in the article.