When less is more: De-prescribing medications

The what, why, and how of taking patients off their drugs.


Prescribing medications is by no means a foreign process to hospitalists. After choosing the most appropriate drug therapy for the desired therapeutic effect, they know to explain the risks and side effects to the patient and ensure that he or she is monitored for side effects as an outpatient.

“What's the 1 thing that we don't do?” asked Joseph Li, MD, FACP, during his presentation in March at Hospital Medicine 2016 in San Diego. “The 1 thing that I never do is talk about when to stop these medications....It's not surprising, when we think about this process we have for prescriptions, how patients end up with this long list of medications that seems to go on and on and on.”

Photo by Thinkstock
Photo by Thinkstock

Patients who come into the hospital with giant medication lists and bags stuffed with prescription bottles may not need to be on all those drugs, and that polypharmacy could be doing more harm than good, said Dr. Li, chief of hospital medicine at Beth Israel Deaconess Medical Center and an associate professor of medicine at Harvard Medical School in Boston.

De-prescribing is the systematic process of discontinuing drugs when the harms outweigh the benefits in terms of a patient's care goals, current level of functioning, life expectancy, values, and preferences. Dr. Li explained how hospitalists can de-prescribe safely and effectively, as well as which drugs are common candidates for discontinuation.

The polypharmacy problem

Polypharmacy, of course, is not a new phenomenon. “All of us in training have heard this term and are familiar with the term, but if you're anything like me in terms of my training, there really wasn't much in terms of discussion about this concept of de-prescribing,” said Dr. Li, who is an ACP Regent.

Data on the inappropriate use of medications show that up to 20% of drugs used in the elderly may be inappropriate, and up to 60% of the elderly are taking at least 1 potentially inappropriate drug, he said, citing a 2007 study published in Internal Medicine Journal and a 2011 study published in BMC Geriatrics. Data also show that the prevalence of polypharmacy increases with age. “That's consistent with what I see in terms of many of my older patients,” Dr. Li said, supporting his own experience with results of a 2008 article published in the Journal of the American Medical Association. “They seem to have longer drug lists than some of my younger patients, and the data suggest that 30% of the elderly (elderly being defined as 65 years or older) are prescribed at least 5 or more medications. To me, that seems pretty conservative, given that the patients I see are on many more than that.”

Drivers for polypharmacy include patient expectations and clinical guidelines, Dr. Li said. “I'm not here to tell you that clinical guidelines are a bad thing; they're a good thing. They help us boil down the data, help remind us what therapies are appropriate for certain conditions,” he said. “But I think that clinical guidelines do drive the issuing of additional medications to patients....And unfortunately, I do think that there are a number of medications currently prescribed [in] our country for adverse drug effects, which are due to other medications.”

How to de-prescribe

Dr. Li outlined a 5-step de-prescribing protocol that has been proposed by Ian Scott, MBBS, and his research group in Australia. First, clinicians must obtain a complete medication list—that is, what the patient is actually taking, not just what he or she has been prescribed—and know the exact indication for each drug, Dr. Li said. Next, they should consider the benefits and risks of stopping therapy before determining which drugs can be safely discontinued. The next step is to prioritize which drugs to reduce and which ones to stop. Last, clinicians must monitor the patient for side effects.

This must be a patient-centered intervention, Dr. Li emphasized. “Don't even begin to think about stopping one's medication without talking to the patient,” he said. “Some patients are going to be very open to the idea of stopping, whereas others are going to be very resistant to the idea. But either way, I think it's very important for you to have a conversation with the patient and recognize that there's going to be some inherent uncertainties.”

Indeed, clinicians don't know exactly what will happen after stopping a drug, but they should realize that this uncertainty is not any different than at the time of prescription, Dr. Li said. “This is going to be a shared decision between you and the patient and also their outpatient providers, and somebody's going to have to follow up on monitoring the patient for side effects,” he said.

The most important part of the de-prescribing process is communication, said Dr. Li. “There's no reason for that patient to trust you as much as they trust their [primary care physician]. They've never seen you before, and you're going to have to engage that patient and convince them that you and your [primary care physician] are in this together and having a conversation about the risks and benefits of the medications, and you're going to need to absolutely document this in the medical record and make sure that the plans for monitoring the patient are handed off to the outpatient provider postdischarge,” he said.

What to de-prescribe

Although most of the research on de-prescribing is conducted in outpatient settings, there are some data pertinent to inpatient de-prescribing, Dr. Li said.

In a pilot study of 50 hospitalized patients 65 years and older who were taking at least 8 medications routinely on admission, clinicians were able to reduce or discontinue at least 2 in more than 84% of patients and 4 or more in more than half of patients, Dr. Li said, citing a 2016 study published in Internal Medicine Journal. The most commonly discontinued medications were proton-pump inhibitors (PPIs), H2 blockers, statins, antihypertensive agents, and inhaled bronchodilators.

After 78 days, symptom relapse caused only 5 medications to be restarted in 3 out of 39 patients who maintained follow-up. “So it's really compelling that you can actually...reduce the risk of medications from anywhere from 2 to 4 medications in a population of patients very similar to ones that you and I take care of and not have these patients go back on these medications,” said Dr. Li.

A systematic review of 31 drug withdrawal trials published in 2008 by Drugs & Aging showed that anywhere from 20% to 100% of elderly patients could have their antihypertensives, psychotropics, and benzodiazepines withdrawn safely, he added. And a 2013 Cochrane review of 9 randomized controlled trials showed that in more than 80% of patients with dementia, it was safe to withdraw antipsychotics. Unsurprisingly, safely withdrawing psychotropic medications and benzodiazepines reduced falls and improved cognitive and psychomotor function, Dr. Li said.

Statins are prime candidates for de-prescription, he said. “I know in my practice, it's fairly commonplace for me to see some really old patients on statins,” Dr. Li said. “And they absolutely probably were appropriately started on a statin a number of years ago as a risk prevention matter, but when you're 94 years old and you're thinking you're on a statin for primary prevention, I think the circumstances change a little bit in terms of risk-benefit ratio.” In the U.S. alone, if all the inappropriate statins were eliminated, there would be a potential cost benefit of more than $600 million, said Dr. Li, citing a 2015 study published in JAMA Internal Medicine.

And consider, for instance, how often patients come out of the ICU on a PPI just because a prescription was written for gastrointestinal prophylaxis, he said. (For more on addressing inappropriate PPI use in the hospital, read “Reducing unnecessary PPI use may help save lives” in the March 2016 ACP Hospitalist.)

Although the de-prescribing process may seem daunting, a simple way to start is to target 1 class of medications, Dr. Li said. Choosing to focus first on PPIs, for example, can be a viable first step to rolling out a de-prescribing protocol, he said. “And that may be the low-lying fruit because I do think that convincing people to stop their atypical antipsychotics or their tricyclic antidepressants may be more challenging and more difficult than just stopping PPIs,” said Dr. Li.

Prescriptions are often inadvertently continued or renewed, unbeknownst to a patient's primary care physician, Dr. Li noted. “That's why I do think a hospitalist can be a good person to bring this up because it's the 1 opportunity for a fresh set of eyes to say, ‘Hey, this doesn't make sense. Why are we doing this?’ and we can potentially intervene in this cascade,” he said.