Reaching Out to Patients to Identify Adverse Drug Reactions and Nonadherence

Abstract

Many millions of dollars have been spent on preventing adverse drug reactions at the point of prescribing. Automated systems help identify drug-drug interactions and excessive drug doses. Computerized alerts warn prescribers about potentially inappropriate drugs in older adults. Yet, only one-quarter of adverse drug reactions can be prevented by catching errors or problems at the time of prescribing.1 The remainder of adverse drug reactions are not the result of prescriber error, but simply represent the known side effects of drugs. Some patients who take calcium channel blockers will develop peripheral edema. Some patients who take selective serotonin reuptake inhibitors (SSRIs) will experience marked sexual dysfunction. For some drugs, risk factors have been identified that place a patient at higher risk of developing an adverse event. However, in most cases, we cannot predict who will develop an adverse drug reaction, and who will not. We prescribe and hope for the best. Unfortunately, physicians don’t do a good job of identifying and appropriately managing adverse reactions when they do occur. Many patients don’t tell their doctors when they are experiencing an adverse event, and we often don’t ask.2–3 Moreover, physicians often misattribute the symptoms of an adverse drug reaction as the manifestation of an underlying disease, leading to diagnostic workups and a prescribing cascade of new medications rather than treating the problem at its source by stopping the offending drug.4 On a broader level, only a small fraction of adverse drug reactions are reported to the FDA Adverse Event Reporting System (http://www.fda.gov/Safety/MedWatch/), hindering efforts for post-marketing surveillance of drug safety. These problems with recognizing and managing adverse drug reactions occur not because physicians are incompetent, but because we lack the systems that would allow us to systematically identify and address medication-related problems. The research described by Forster et al. in this issue of Archives shows a promising approach to bridge this quality gap. Building on past studies that have shown the benefits of reaching out to patients to identify adverse drug reactions, the authors developed a hybrid system. Three days after a drug was newly prescribed, the system generated a phone call to the patient. Using interactive voice response technology, the system asked the patient four simple questions about problems they may be having with their drugs and whether they wanted to talk to a pharmacist. The process was repeated two weeks later. One-third of contacted patients needed a follow-up call from the pharmacist. Overall, the system identified slightly under half of the 22% of patients who experience an adverse drug reaction. In addition, it identified one-third of the 6% of patients who were non-adherent to their medications. This is exciting and highly promising. It is also not ready for widespread implementation. While the system detected a number of medication-related problems, it missed more than half of adverse drug reactions and two-thirds of episodes of non-adherence in patients - and would likely have done worse outside the controlled environment of a research setting. For most patients, the simple act of reaching out is necessary but not sufficient. People don’t develop adverse drug reactions – they develop symptoms, which may be mistakenly attributed to causes other than drugs (including “getting old”), and which they may be hesitant to disclose. (Other adverse reactions may be completely asymptomatic but nonetheless serious, such as progressive hyperkalemia or anemia). Outreach calls may also be asynchronous with when the patient develops a medication-related problem. These challenges bedevil the widespread practice of calling patients several days after hospital discharge to inquire on their wellbeing and identify problems with their medications. While a wonderful idea, relatively little is known about how well these follow-up procedures actually identify problems, and although there is some evidence that these interventions are effective, the benefits are not as great as one might hope.5 What might be most helpful is a multifocal approach, in which the surveillance strategies being developed by Forster and like-minded colleagues are coupled with efforts to educate and encourage patients to be active partners in monitoring adverse reactions and non-adherence to their medications.6 This latter approach is best exemplified by health-coach based approaches pioneered by Coleman and others, in which impressive improvements in health resulted not from bringing services to patients, but by helping patients be engaged participants in their own care. 7 These interventions are complex, and their potential benefits do not diminish the substantial contribution of surveillance-based approaches. Nonetheless, the solution to the problems of adverse drug reactions and non-adherence cannot solely rest on bringing the health care system closer to the patient. We need to empower our patients to come closer to us.