Concerns, not ignorance, inhibit stroke med adherence

Physicians should work with patients to develop solutions to medication concerns.


It's clear that more than a spoonful of sugar is required to get a stroke survivor to take the host of medications he or she is typically prescribed. But the exact causes of and cures for patient nonadherence have so far eluded hospitalists and other physicians.

Hoping to gain some further insight, researchers at Columbia University and Mount Sinai Medical Center in New York City recently surveyed 600 survivors of stroke or transient ischemic attack. The patients reported their medication adherence and answered a series of questions about their perceptions and knowledge about stroke care, medications and the health care system generally.

Courtesy of Ian M Kronish
Courtesy of Ian M. Kronish.

ACP Hospitalist spoke with lead author Ian M. Kronish, MD, MPH, assistant professor of medicine at Columbia, about the results, which were published in the May Journal of General Internal Medicine.

Q: What motivated this study?

A: I'm a general internist, and a lot of what my job entails is keeping track of medications of my patients. I've been interested in reasons for low rates of adherence in primary care patients in general. I had an opportunity to work with a colleague on a project in stroke survivors. The focus of this project was to enhance the stroke survivors' ability to self-manage their risk factors after having had the stroke. We thought it was a great opportunity to try to understand the barriers to medication adherence in this high-risk population.

Q: In addition to being high-risk, what other factors defined your study population?

A: We included a lot of patients who were of low socioeconomic status. We recruited participants from a mix of community-based and more traditional registries. That was one great thing—that we were able to enroll a substantial number of disadvantaged minority patients who are often excluded [from research].

Q: How did the findings compare to your expectations?

A: On the one hand, they met my hypothesis of what I thought would happen. In the literature, there's a high prevalence of nonadherence to medications. A lot of times they'll quote that about 50% of patients are not adherent to medications. We found 40% were nonadherent.

One of the things that was surprising was [the finding on] knowledge about stroke and medication adherence. Participants were asked a question about what they thought were the three most important things they could do to prevent a recurrent stroke, to tap into their understanding of the big risk factors. A large percentage of them didn't know more than one of those risk factors, but that didn't predict nonadherence. Doctors spend a lot of time thinking we need to better educate our patients about the risk factors for stroke and other medical conditions, but the connection between knowledge about a disease and whether patients are going to comply with their medications doesn't seem to be that strongly related.

Q: What factors better predicted nonadherence?

A: It's more their deeper health beliefs gained from a variety of sources and from their own experiences [identified in the study by concerns about having to take medication, long-term and disruptive effects of medication, and becoming dependent on medication]. I had done some work on a qualitative paper where we asked focus groups of Hispanic and African-American patients about adherence to hypertension medications, and we saw that the patients' own beliefs about their hypertension medications seemed to play a large role in their decisions to take medications.

In this study, we also found that patients who perceived they were discriminated against by the health system had lower adherence. Maybe it's a result of the vulnerable study population that we found this link. The association wasn't as strong as with concerns about medication, but it was there. I don't how much doctors think about discrimination as being a risk factor for nonadherence to medical therapy.

Q: Can physicians do anything to correct these concerns about discrimination?

A: It wasn't about perceived discrimination from your personal doctor, it was perceived discrimination from the health system, so it's not really clear. More work would need to be done to better identify which parts of the health system patients feel discrimination from. There may be things physicians can do: They may think about the quality of the support staff and what communication is going on, how easy it is to access the practice. We can try to make our practices as comfortable an environment as possible for patients of different racial and ethnic backgrounds. These things may sound like common sense, but if you realize that these things impact patients' ability to self-manage their illnesses, they become even more important to implement.

Q: What about the deeper beliefs about medication? Are they modifiable by physicians?

A: I don't think there are great studies yet directly addressing this. Because beliefs come from so many different sources, not just doctors, it is a challenge. Other studies have shown that collaborative communication with doctors is at least associated with better medication adherence. The only measure that [our study had] like that was a question about their trust in doctors, and that didn't appear to be related. It's possible if clinicians are trained in collaborative communication, this has the potential to modify beliefs, or at least outweigh certain beliefs. We need more creative approaches.

Q: What kinds of approaches are you investigating?

A: Patients in this study were randomized to a trial in which the intervention group participated in a [chronic disease self-management] workshop that was led by peers in their community who had suffered strokes or had similar health problems. The idea is that peers may be better at delivering information and teaching the self-management skills that doctors want their patients to know. The workshops also included a problem-solving component where, at the end of each session, participants would think about something concrete they could do to help control their risk factors. In the case of stroke, a lot of that is finding ways to take their pills more regularly. But it was up to the patient to decide and then they would come back to the group the next week and report on what they had done—for example, buying a pillbox or asking someone in their family to help remind them to take their pills. The idea is that this will build patients' confidence that they can successfully take action to improve their health. We're still waiting to look at those results, but teaching self-management skills in this manner may be one approach to improve medication adherence.

Q: What lessons should physicians take from your research?

A: While doctors realize nonadherence is a problem, I don't think they realize how common a problem it really is. Doctors could take time to explore their patients' understanding of what caused their stroke and the reasons for the different medicines. A lot of times they'll be surprised by their patients' beliefs.

For example, I hear a lot from my patients about statin medications. There are different stories percolating in the media and through social networks about potential toxicity of these drugs and [as a result] a lot of patients are taking them every other day or once a week. Doctors think that the risks of statins are very, very low, but a lot of patients are getting mixed messages.

Exploring some of the concerns that patients have or asking if they have any concerns about the medication could go a long way. Have an honest discussion with the patient about those concerns and try to accommodate them the best that you can. Maybe it's coming to a halfway point and saying, “Why don't we try a lower dose, but take it every day? Or why don't we see how this affects you? We'll monitor your blood tests and make sure it's not causing any harm.” Acknowledging patients' concerns about medications may go a long way toward increasing trust and hopefully increasing adherence as well.