Written by: Michael A. Steinman, M.D.; C. Seth Landefeld, M.D.
Media contact: Alicia Rohan
This editorial was originally published in the Journal of the American Medical Association on Nov. 13, 2018, by Michael A. Steinman, M.D., Division of Geriatrics, University of California San Francisco; and San Francisco VA Health Care System; and C. Seth Landefeld, M.D., Department of Medicine, University of Alabama at Birmingham.
Headshot of C. Seth Landefeld, M.D.Inertia is a powerful force. Stopping or starting is difficult in health care as well as in other sciences. Ineffective or potentially harmful treatments are often not stopped, even years after they have been started, and effective treatments are too often not started at all.
Once started, medications can be difficult to stop. It takes time for office-based clinicians to reassess use of medications prescribed to patients with chronic diseases, particularly therapies that are not clearly related to the symptoms or conditions that are the focus of a given patient encounter. Even if a clinician recognizes a medication as potentially inappropriate and a candidate for discontinuation, both the clinician and the patient may be concerned that “the devil they know is better than the devil they don’t know,” and that stopping a medication may worsen symptoms or biomarkers or may be perceived as giving up. Clinicians also may be unsure about how to best taper different medications or how to recognize and manage adverse drug withdrawal events. Thus, use of unnecessary and potentially harmful medications is common among older adults.
In the latest issue of JAMA, three articles highlight the power of inertia in stopping or starting medicines. The clinical trial by Martin and colleagues found that a structured, pharmacist-led educational intervention stopped the use of an inappropriate medication in 43 percent of eligible community-dwelling older adults versus 12 percent who received usual care. The investigators randomly assigned pharmacies in Quebec to an intervention that involved an approach for recommending deprescribing (n = 34) or to usual care (n = 35). The trial included 489 older adults who were prescribed one of the following potentially inappropriate medications: sedative-hypnotic drugs, nonsteroidal anti-inflammatory drugs, glyburide or first-generation antihistamines. Two hundred forty-one patients received usual care, and 248 patients received the D-PRESCRIBE intervention. In the intervention group, community pharmacists were encouraged to deliver (via mail or in person) educational brochures to patients about one of their potentially inappropriate medications, which included why the medication may be inappropriate, potential alternatives and strategies for stopping the drug. Pharmacists were also encouraged to provide evidence-based pharmaceutical opinions about the drugs to the patients’ physicians using a standard template; such recommendations are reimbursable in Quebec. After six months, 106 of 248 patients (43 percent) in the intervention group stopped using the inappropriate medication compared with 29 of 241 (12 percent) in the usual care group.
This is an important study with a powerful effect. Interventions that require substantial effort by pharmacists or physicians are often highly effective for deprescribing medications but are challenging to implement in actual clinical care (i.e., “real-world”) settings because of the time, effort and costs required. Interventions similar to D-PRESCRIBE are likely more feasible than interactions that involve more effort by pharmacists or physicians than dispensing educational material, and the outcomes compare favorably with other lower-intensity interventions that targeted similar types of medications. Also, the evidence-based pharmaceutical opinions used in the D-PRESCRIBE trial that recommended deprescribing to the primary physician likely augmented the patient education component of the intervention, yielding more powerful effects than a similar study that examined the effects of only patient education on the discontinuation of inappropriate medications, conducted by the same research group. Although the study methods provide confidence in the validity of the results, limitations should be recognized. Only one in four eligible individuals agreed to participate in the D-PRESCRIBE study, which possibly selected for patients more likely to respond and introduced biases in favor of the intervention. Also, it is not clear whether this intervention would be as effective in other health care systems where comprehensive primary care is not as widespread. Despite these limitations, the results are potent and teach two key lessons.
First, interventions that engage patients, physicians and pharmacists in the deprescribing process may be more effective than interventions that do not. Given the inertia that accompanies long-term medication use, it is often insufficient to motivate only one of these groups to make a change. In a 2014 study by Tannenbaum and colleagues on deprescribing, which used pharmacist-led education for patients without the physician component, patients who did not stop their medications cited clinician discouragement as a major contributor. Similarly, engaging only physicians neglects the critical role of patient buy-in to make change happen. Second, pharmacists can initiate and catalyze deprescribing by providing the expertise, focus and targeted recommendations that can guide physicians while serving as trusted arbiters and sources of advice for patients, including for issues that some patients may be reluctant to tell their physician. This pharmacist role should not be underestimated. Studies have repeatedly shown that involving pharmacists in quality improvement activities can yield substantial benefits, yet historically, these professionals have not been used to their fullest potential. Support for the pharmacist’s role, including the standard in Quebec that pharmacists can bill for providing pharmaceutical opinions to a physician, may promote beneficial deprescribing at little cost.
Also in this issue of JAMA, two research letters highlight another example of inertia by reporting inadequate dispensing of naloxone by pharmacies to prevent fatal opioid overdoses. In these secret shopper studies, investigators evaluated the availability of naloxone from pharmacies in Texas and California, two of the 50 states that, in recent years, have passed legislation to empower pharmacists to dispense this lifesaving medication without a direct physician’s prescription. Evoy et al report that, among four pharmacy chains in Texas, where standing orders allow pharmacists to dispense naloxone, 83.7 percent of 2,127 pharmacies contacted would dispense naloxone without an outside prescription and 69.4 percent stocked and would dispense naloxone under standing order. Puzantian and Gasper report that, in a random sample of all pharmacies in California, where pharmacists are allowed to furnish the drug without a physician’s prescription, 23.5 percent of 1,147 pharmacies that were contacted and provided data indicated that naloxone was available and could be dispensed without a prescription. The differences reported are not explained, but may reflect differences in the pharmacies included (chains versus all pharmacies) and in state laws. Other actionable reasons for these differences may be the limitations of availability in both states, including the training and comfort levels of pharmacists, awareness of the legislation, and insurance coverage. Improved training of pharmacists may be needed to make naloxone universally available for the prevention of opioid-related deaths. Such efforts will require resources, investment and organizational support.
Older adults often take too many medications. More naloxone in the hands of friends and family can save lives. Change, or overcoming inertia in health care, is difficult but essential to deprescribing harmful medications and making effective medications available to individuals who need them. Involving groups that are less often the target of interventions to overcome inertia, such as patients and pharmacists, will likely be necessary for medicine to achieve such changes. These changes are more likely to be achieved if patients, pharmacists, nurses and organizations all push and pull in the same direction.