Prescribing drugs is good. So is deprescribing.

There was nothing unique about that patient encounter except that it was my first appointment of 2017 and everyone at the office was buzzing about their New Year’s resolutions.

It went like this:

Me: “Are you taking all of your medications?”

Patient: “Yes, of course.”

Me: “Okay, let’s review them. Do you remember what they are?”

Patient: “I am not sure, but they are all here.” (Patient hands me a plastic bag brimming with orange pill bottles, boxes and over-the-counter containers.)

Me (in my head): Oh, brother!

That’s when I made my resolution: In 2017, I will try to tip the balance: I will not only try to write fewer prescriptions, I will also try to get more of my patients off their drugs altogether.

The “why” was obvious to me: Too many patients are taking too many drugs, for too long, in too-high doses, suffering harmful effects.

The “how”? That was less obvious. Starting in the second year of medical school (course name: Pharmacology) and continuing through residency training and beyond, doctors are taught how to prescribe drugs.

Here’s some of what we learn: which drugs are best for different conditions (for high blood pressure, diabetes, depression, pain and so on); which antibiotic is best for which type of infection; what are the most beneficial dosages and how frequently should certain drugs be taken; what is the best route for certain drugs (oral, rectal, IV); what the common side effects of most drugs are; which drugs are better for children, which for adults. There is much to know, and doctors have to stay up-to-date constantly, on old and new medications, recalls, generics, brand names, warning labels.

This is important, of course. Drugs have an important place. But rarely do we teach young doctors – that is one of the things I do now – how and when to deprescribe a drug. Doing so is not as simple as saying “stop.” Deprescribing is its own process, requiring extreme caution and a certain skill on the part of the physician. It is a skill, however, that is not being taught, and it is barely studied to develop best practices.

And that hurts patients, because few drugs are meant to be used forever, and all have potential to cause harm. For some drugs, those harms include addiction – much in the news these days – especially in the case of opioids, some anti-anxiety medications and certain sleeping aids. Cutting down on some of these drugs needs to be done very gradually and carefully.

Some people simply take medications for too long: Take certain heartburn medications (called proton pump inhibitors, or PPIs) for more than the recommended two weeks, and you risk pneumonia, intestinal infections, broken bones and vitamin B-12 deficiency.

Some people outgrow their medication: They change their lifestyle, and their diabetes, cholesterol or high blood pressure medications may not be needed anymore. But they keep taking them, because no one told them to stop.

But it’s my patient with the bag of medications who illustrates the situation most acutely: an older adult who is prescribed too many medications, by too many physicians, all at the same time, even if all are given for legitimate reasons.

“Polypharmacy” is the name we give to prescribing patients five or more medications at the same time.

Why is that problematic? First of all, drugs are chemicals that can interact with one another, potentially causing all kinds of complications that may not be apparent if you just take the one medication. Second, the aging process causes the kidneys and liver to be less efficient in processing medications. That often leads to more of the drugs sticking around in the body and magnifying their effects – and side effects. Polypharmacy has been shown to contribute to higher rates of hospitalizations and death and – of course – to higher costs.

The problem is widespread: According to some studies, about 20 percent of adult patients are routinely on five or more drugs, and in people older than 65, between 30 and 70 percent are treated with polypharmacy. In nursing homes and other residential facilities, that rate goes up to 90 percent.

Most patients – 89 percent, in one recent study of polypharmacy – have told researchers that they would be interested in stopping a medication if their physician agreed that was the right course of action.

So what can we physicians do?

First, we need to appreciate the scale of the problem and the potential harm of polypharmacy.

We need to recognize that there are professional and cultural norms that push us to prescribe (rather than find other solutions) and to overprescribe.

We doctors need to get out of our comfort zone. Yes, it is easier to keep somebody on a medication and just keep refilling it when the pharmacy calls, but is it better for the patient?

We need to get over our fear of causing harm by deprescribing. That’s not an irrational fear, of course. Indeed, many medications (for example, anti-depression medications, some high blood pressure drugs and steroids) need to be stopped gradually because stopping abruptly can be dangerous.

We also need to make sure we are treating the patient, not the disease. That means considering whether and what to prescribe while taking into consideration the patient’s age, other health conditions and overall life expectancy. As doctors, we need to ask ourselves, for example: For an 87-year-old woman with metastatic cancer, should I prescribe a medication to lower her cholesterol level? The answer is probably no: It is highly unlikely that the patient would benefit from this drug and very likely that she would suffer from harms caused by the drug and its interactions with her other medications.

Common classes of medications that are good candidates for deprescribing include:

Anti-anxiety medications known as benzodiazepines, which can contribute to cognitive impairment, delirium, falls (and related injuries), breathing problems and motor-vehicle accidents.

Atypical antipsychotics, which are often used to treat psychosis and, in the elderly, dementia.

Anti-cholesterol statins, which can cause muscle problems, cognitive impairment and a higher risk of diabetes. Statins also have a high risk for interaction with other medications and certain foods. Given that the benefits of statins are long term, they are not needed for elderly patients.

Tricyclic antidepressants, which are used for depression and dementia. These are not recommended in the elderly but are often used nonetheless, causing side effects or harms that can include low blood pressure (which contributes to falls and fractures), heart arrythmias and other disturbances, delirium, difficulty urinating, dry mouth and constipation.

Proton pump inhibitors, mentioned above.

Knowing which classes of drugs require special attention is important, but it is not enough. How to do it effectively, efficiently and with the lowest chances of harm is still anybody’s guess. We need researchers to help us by discovering and evaluating the best discontinuation protocols.

There is also a big-picture issue here: Deprescribing requires a lot of thought and planning. There are many more incentives for doctors to prescribe a medication than to stop one. Insurers and payers need to create incentives to allow primary-care physicians to spend adequate time with our patients to get them off drugs and carefully monitor their response when a medication is withdrawn in a supervised manner.

For my part, I am going to try to do that in 2017.