Physician Perspectives on Deprescribing Cardiovascular Medications for Older Adults

Parag Goyal, MD, MSc; Timothy S. Anderson, MD, MAS, MA; Gwen M. Bernacki, MD, MHSA; Zachary A. Marcum, PharmD, PhD; Ariela R. Orkaby, MD, MPH; Dae Kim, MD, MPH, ScD; Andrew Zullo, PharmD, PhD; Ashok Krishnaswami, MD, MAS; Arlene Weissman, PhD; Michael A. Steinman, MD; Michael W. Rich, MD


J Am Geriatr Soc. 2020;68(1):78-86. 

In This Article


This national physician survey showed that geriatricians, general internists, and cardiologists frequently consider deprescribing cardiovascular medications in the setting of adverse drug reactions, but variably consider deprescribing in the setting of other circumstances, like limited life expectancy. We also found that physicians report similar barriers to deprescribing irrespective of specialty. These findings extend prior studies on physician attitudes toward deprescribing by specifically examining cardiovascular medications and comparing perspectives across three specialties from a national sample. Our results have important implications for future efforts to promote deprescribing as a strategy to optimize the use of cardiovascular medications and to provide quality comprehensive patient-centric care to older adults.

While deprescribing is important after a new symptom or adverse drug reaction occurs, stopping medications only after they have caused a negative outcome is unlikely to significantly stem the ongoing risks of polypharmacy and the complex medication regimens endemic to older adults. A recent systematic review showed that an active process of targeted patient-specific deprescribing interventions can safely reduce total medication burden and potentially reduce mortality,[18] supporting the role of deprescribing in select circumstances. To make significant progress toward decreasing rates of adverse drug events, it is important to adopt a more proactive approach to medication management. For example, for individuals with limited life expectancy, the benefits of many cardiovascular medications may be diminished or even absent,[16,19] while the risk for adverse drug events in the setting of functional and/or cognitive impairment may be elevated.[20–22] Consequently, the harms of polypharmacy and risk of adverse drug events posed by continuing multiple preventative cardiovascular medications may outweigh the benefits for some older adults, such as those with dementia and those who struggle with performing their activities of daily living. Our finding that specialties varied in the frequency with which they considered deprescribing in these scenarios implies that there may be additional factors, such as those related to training, experience, and/or patient expectations, that affect whether and to what extent different specialists consider deprescribing. In addition, there may be variability with regard to the way life expectancy is evaluated and incorporated into decision making. For example, cardiologists were more likely to consider deprescribing for a 90-year-old woman compared to any of the other limited life expectancy scenarios, suggesting that chronological age may supersede physiological age when making decisions in some situations. In light of these observations, efforts to generate evidence supporting the potential benefits and safety of deprescribing cardiovascular medications remain important but may not be sufficient to improve prescribing practice for many older adults. There appears to be a need to sensitize clinicians to the growing body of evidence supporting the potential role of deprescribing, an effort that has begun in the cardiology community.[6] In addition, effective implementation strategies that can incorporate risk-benefit assessments, elicitation of health priorities, and deprescribing processes into routine clinical care in the primary and specialty care settings are much needed.[8,16]

We also found that medical specialties differed in the medications that they would consider deprescribing. Cardiologists less frequently considered deprescribing statins and/or aspirin in several scenarios compared to geriatricians or general internists. Historically, data on preventing cardiovascular events in older adults have been limited by the exclusion of older adults from clinical trials.[23] While statins for primary prevention are well studied in adults younger than 75 years, data in adults older than 75 years are limited and have shown little benefit.[4] There were similar gaps in knowledge regarding the use of aspirin for primary prevention in older adults until recently.[5] Reasons for reluctance toward deprescribing statins and aspirin in the absence of robust data supporting their benefits, especially when life expectancy is limited and exceeded by the time horizon to benefit, merit additional investigation. Whether the Aspirin in Reducing Events in the Elderly trial, which was published after conduct of this survey and showed that aspirin for primary prevention in older adults may be harmful,[5] will alter prescribing (and deprescribing) behavior remains to be seen. Tools like the Screening Tool of Older Persons Prescriptions in Frail Adults With Limited Life Expectancy[19] could assist clinicians with identifying cardiovascular medications that provide limited benefit. Additionally, guidelines for safe methods of deprescribing cardiovascular medications are also needed; guidelines for deprescribing other medication classes have been developed and could provide a useful starting point.[24–26]

We found that barriers to deprescribing were diverse but consistent across specialties and largely consistent with prior surveys of primary care physicians' perspectives on deprescribing PIMs.[9,10,12] Importantly, interfering with another clinician's treatment plan was the most frequently reported barrier to deprescribing cardiovascular medications, extending observations from other countries.[10] This concern is especially important within the US healthcare system, where fragmentation is common[27] and older adults routinely see multiple physicians.[17] Thus, shared communication between specialties is crucial for cardiovascular medications, which may be comanaged by geriatricians, general internists, and/or cardiologists. Yet, our study found that fewer than 60% of geriatricians and 50% of general internists and cardiologists communicated with other clinicians about deprescribing cardiovascular medications when concerns arose. Some clinicians may have been comfortable with deprescribing without discussion with another clinician, especially those who believed that cardiovascular medications were in their purview of care. However, given the prevalence of adverse drug events among older adults, these findings point to a potential gap in care. When considering deprescribing, cross-specialty communication may be valuable, as different specialties can offer different perspectives regarding the potential risks and benefits of continuing or discontinuing medications. Interestingly, cardiologists reported concerns about interfering with other clinicians as a barrier to deprescribing medications that would typically fall under their purview. The reason for this observation was not clear but could relate to a diffusion of responsibility. This highlights the complexities regarding the roles and responsibilities of specialists when it comes to deprescribing and identifies an important area of research that warrants further investigation. Taken together, our findings provide additional empirical evidence for the need to develop deprescribing protocols that incorporate formal processes of interdisciplinary communication.[18,28] Pharmacist involvement, as was studied in the recent developing pharmacist-led research to educate and sensitize community residents to the inappropriate prescriptions burden in the elderly (D-PRESCRIBE)[28] randomized controlled trial, could offer a particularly appealing strategy to bridge these communication gaps and should be examined as a potential strategy to combat this important barrier to deprescribing.

The second most common barrier to deprescribing reported by physicians was patient reluctance, which is consistent with findings from older studies evaluating physician-reported barriers.[10] Importantly, this commonly held perception might not actually reflect patient attitudes toward deprescribing. In a recent analysis of the National Health and Aging Trends Study, most older adults reported willingness to stop at least one of their medications if their physician said it was possible.[29] While it did not specifically assess attitudes toward cardiovascular medications or specify between primary and secondary prevention, that study coupled with our findings highlight a potentially important incongruity between patient attitudes and physician perceptions on deprescribing. This incongruity can erode patient-physician communication regarding deprescribing, and subsequently impair the shared decision-making process necessary for deprescribing.[13] Accordingly, our findings support the need to increase physician awareness about the role of deprescribing in providing patient-centered care and the need to develop tools that can facilitate patient-physician communication about deprescribing.

A major strength of this study is the examination of a national sample of physicians drawn from a large medical-specialty organization in the United States. Respondents were diverse in age, practice setting, and geographic region. However, our findings should be interpreted in light of several limitations. First, the overall response rate to the survey was low, especially among cardiologists. Low response rates for surveys are common, and not specific to deprescribing. Nonetheless, it is possible that the low response rate here could reflect limited knowledge or perhaps even a lack of interest in this topic. Respondent characteristics were similar to the full sample. However, those who did not respond to the survey may have different perspectives on deprescribing compared to respondents who may have been more familiar with and/or more interested in the concept of deprescribing. Accordingly, our findings likely overestimate physician willingness to consider deprescribing, and probably represent the best-case scenario for the different specialties. Second, our findings were subject to social desirability bias as they were based on physician self-report. Third, differences in reported deprescribing practices between specialties may have reflected differences in the patient populations cared for by each specialty. To address this, our survey included identical hypothetical cases, which allowed us to directly compare reported deprescribing practices across specialties. Fourth, our survey did not examine perspectives on who is responsible for deprescribing or respondent familiarity with the concept of clinical inertia as it relates to medication prescribing practice; these areas will be important to explore in future work.