Slower metabolism, compromised adherence, and potential interactions present new challenges as patients get older. Here’s how to manage them.

Tom was 90. A retired businessman, he was still happily married, and his three children loved and supported him. His health was failing, however. He had been diagnosed with chronic atrial fibrillation and congestive heart failure, and he often complained about pain in the shoulders. Over time, his hands became shriveled and useless. We suspected a cervical neuropathy due to spinal stenosis or degenerative disk disease, but he had resisted diagnostic studies, always emphatic that he wasn’t going to have surgery anyway. Acetaminophen and heating pads were tried but to no avail; nonsteroidal anti-inflammatory drugs (NSAIDs) were not an option as he was taking warfarin (Coumadin).

A pain specialist suggested carbamazepine (Tegretol). It would work much better than narcotics, he explained, because this was neuropathic pain—and all I would have to do was check an occasional WBC count and drug level. How wrong he proved to be! Tom’s wife was on the phone within days, saying that he was feeling awful—that his left leg was jerking uncontrollably and that he was nauseated and fatigued. A quick house call confirmed her report: His left leg was twitching rhythmically every 30-40 seconds, and his color was an ashen gray. A stat carbamazepine level was normal, at 3.8 µg/mL, but his sodium was low—128 mEq/L.

Continue Reading

The carbamazepine was stopped, and within 24 hours Tom was feeling better. His muscle twitching stopped, and his sodium returned to normal levels.

Our error seems obvious now—the carbamazepine caused a syndrome of inappropriate antidiuretic hormone secretion and hyponatremia—but it is the kind of mistake most of us can make at least once in our clinical lifetime. That said, Tom’s case can teach us some important lessons.

Those of us privileged to write prescriptions hold an awesome power in our hands. We have thousands of medications at our disposal, and even if we write for only a small fraction of these on a regular basis, there is still plenty of room for error: using an antibiotic to which a patient has a known allergy; failing to recognize the potential for a serious drug interaction; ordering the wrong dose or the wrong dosing schedule.

Nowhere is this potential for error greater than with the elderly, who usually take multiple medications and who may—because of poor eyesight, dementia, or just forgetfulness—use their medications incorrectly even when they have been properly prescribed. Moreover, drug metabolism may be affected by the physiologic consequences of aging—decreased hepatic metabolism and renal clearance, along with increased volume of distribution.

Most older patients need to take prescription medicines; in fact, they may sometimes even be undertreated. But medicine is not always the answer to their complaints, and even when it is, observing some basic principles can help the busy clinician avoid scenarios similar to Tom’s.


Primum non nocere

Most of us learned this simple rule years ago, but it never hurts to go back and think about it: “First, do no harm.” That means asking ourselves whether every complaint really needs medicating. Would a simple conversation solve the problem instead? The use of sedative-hypnotics for insomnia provides a good example. Many elderly people live alone and spend their days napping; a talk about sleep hygiene—and the identification of possible depression—may be better than prescribing a benzodiazepine, which might send the patient to the emergency department (ED) with a hip fracture.

Nonpharmacologic interventions can’t be overemphasized. Can pain and stiffness be treated with stretching, exercise, heat, or ice instead of a nonsteroidal analgesic? Can headaches be magically cured with a new pair of glasses? Often elderly patients don’t want treatment; they just want to talk. Is the man who lives alone really in pain, or is he just lonely? If so, steering him to a senior center—or to a church or social worker—might be far more effective than prescribing another pill.

If a pill is absolutely necessary, consider all the medications within a given therapeutic category and choose the one with the best side-effect profile and the fewest drug interactions. Don’t settle for convenience or familiarity. In Tom’s case, gabapentin (Neurontin) would have been a far better choice than carbamazepine: It is no less effective in treating neuropathic pain, is mostly well tolerated, and interacts with fewer other drugs.

Some choices are fairly obvious. For example, it is foolish to use diphenhydramine (Benadryl), which is strongly anticholinergic and requires multiple daily doses, when loratadine (Claritin), taken once a day and much less sedating, will do the trick. Other substitutions may be less apparent: sertraline (Zoloft) is well tolerated by the elderly, whereas paroxetine (Paxil) and fluoxetine (Prozac) can be problematic because of multiple drug interactions and prolonged half-lives.

And don’t be misled into thinking that older and cheaper drugs are necessarily better—no matter what that HMO formulary says. Take the obstructive symptoms of benign prostatic hyperplasia (BPH), for example. Doxazosin (Cardura) may work as well as the newer and more expensive tamsulosin (Flomax), but the difference in side effects, such as the reduced incidence of orthostatic hypotension with the latter, justifies the increased cost.

Start with the brown bag

One of the most valuable things clinicians can do for their elderly patients is to review their medications. This is especially true in an era of increasingly fragmented care and for those patients who spend their days shuffling from one subspecialist to the next.

I insist that my patients bring all their medications to the office at least once a year, and this “brown-bag review” never fails to uncover problems. One patient has gotten nabumetone (Relafen) from his orthopedist and indomethacin (Indocin) from his neurologist; another patient is taking a beta blocker for BP and a calcium channel blocker for headaches; a third is using alprazolam (Xanax) for anxiety and temazepam (Restoril) for sleep. Addressing these problems may be time-consuming—and it’s never why the patient came into the office in the first place—but it almost always results in a safer and happier patient.

Polypharmacy is common in the elderly. One study, for example, found that one in four women over the age of 65 took at least five medications; one in eight took 10.1 Of course, the more medications, the greater the chance of adverse drug reactions. But often many of these drugs can be discontinued, and sometimes the creative clinician will be able to find a new drug that replaces two or three old ones.

One patient, for example, came to my office taking warfarin, nabumetone, and temazepam—for atrial fibrillation, for intractable low back pain that affected him only at night, and for sleep, respectively. His family was concerned about increasing sedation and confusion, which had started around the time his temazepam was first prescribed. However, they were certain that he needed something for sleep—without the temazepam, he was up smoking cigarettes at 3 AM.

The first two drugs are potentially risky taken together, and the last is hardly ever appropriate in the elderly, at least for long-term use. Within 15 minutes we had settled on a plan: He would stop smoking by 5 PM, and we would give him a small bedtime dose of long-acting oral morphine instead of the NSAID and the benzodiazepine. The plan worked perfectly, and apart from adding prune juice at breakfast for predictable constipation, he stayed happily on this regimen until his death one year later.

Stop unnecessary meds

Many clinicians are reluctant to stop medications started by other prescribers, but this reluctance should be weighed against the very real opportunity to improve care when a potentially harmful drug is removed. Think creatively, and tailor the medication list to the patient. For example, a 95-year-old with dementia is not served by a daily dose of lovastatin (Mevacor) —no matter what her lipid profile—and the NSAID that seemed so necessary for her knee pain 10 years ago may no longer be needed once she is bedridden. Similarly, an antidepressant started after the death of a spouse might be stopped once the surviving partner has readjusted to life alone or moved into a new setting.

Keep medication lists up to date

It doesn’t take more than a few low-tech interventions to make a big difference in a patient’s life. I keep an updated problem list and med sheet in the front of every chart. I also cross out old meds rather than erasing them and note why they were stopped so I am not tempted to try them again. A pill organizer can also be life-changing, as can annotated labels that explain what a medication is for: “Lovastatin for cholesterol” and “Metoprolol for BP.”

Use a drug-interaction program to prevent errors

Those low-tech interventions notwithstanding, there is one thing that never leaves my desk or pocket: my Palm Pilot. No matter how many meds you stop, there will always be others that are clearly indicated, and while the advice to “start low, go slow” is always wise, dosing is only part of the picture. Given that many elderly patients are walking pharmacies, addressing the possibility of drug interactions is of paramount importance. That’s where the Palm Pilot comes in.

As compared with standard printed manuals, which require the clinician to search through pages of fine print, such programs as Epocrates Essentials ( allow instant and easy access not only to standard prescribing information but also to an invaluable medication interaction program. Simply bring up the name (brand or generic), click on the drug, add any number of other medications, and then hit “run.” And this program doesn’t stop at telling you that a certain drug combination is incompatible; you can click on the hyperlink to find out why. Now that this technology is widely available, there is no good reason not to use it.

Think “drugs” in the differential diagnosis

A cursory glance at any drug’s package insert will attest that medications can give rise to almost any symptom. The astute clinician will therefore include drugs in the differential diagnosis of every complaint and take a careful history to identify which drug, if any, could be the offender. Failure to consider drugs as the cause of a presenting complaint can lead to a second drug’s being prescribed to counteract the side effects of the first, a sequence known as a cascade.

Studies have shown, for example, that elderly patients being treated with antipsychotic medication frequently end up being treated with antiparkinsonian drugs; those treated for dementia with cholinesterase inhibitors are likely to be given anticholinergics, such as oxybutynin, when they develop predictable cholinergic side effects such as urinary incontinence.2

Stay away from “bad” drugs

Careless use of commonly prescribed “good” drugs may cause more harm in the elderly than appropriate use of rarely prescribed “bad” drugs. For example, the addition of a seemingly innocuous drug such as acetaminophen to an otherwise appropriate warfarin regimen can push the international normalized ratio (INR) through the roof and lead to a disastrous bleed.

However, some meds are so likely to cause problems in the elderly that they should never be used (at least not without careful consideration). These include indomethacin, which has a high risk of causing central nervous system side effects, and meperidine (Demerol), which should always be replaced by morphine. Drugs that are poorly efficacious and interact with multiple other medications, such as digoxin (Lanoxin), also fit into this category, as do muscle relaxants (metaxalone [Skelaxin], cyclobenzaprine [Flexeril], carisoprodol [Soma]) and strong anticholinergics, such as the belladonna alkaloids, GI antispasmodics, and diphenhydramine.

In 1991, M.H. Beers and a team of other doctors at the University of California, Los Angeles, identified groups of medications that were considered to be particularly dangerous in the elderly.3 These categories were further defined in a later study to identify drugs that should almost never be used in the elderly patient because of their toxicity or inefficacy and those that may need to be used but are often used incorrectly (Table 1). This study found that up to a quarter of elderly patients were receiving inappropriate medication.4

The importance of the Beers criteria notwithstanding, merely identifying which medications are most harmful in the older patient is not enough to reduce their use. In fact, a study from 1997 showed that nearly a third of ED visits for adverse drug reactions involved three medications not on the Beers list: warfarin, digoxin, and insulin.5 While digoxin’s efficacy is dubious in most cases of heart failure, the other two medications are clearly an important part of the therapeutic armamentarium; however, their usefulness should not overshadow their apparent additional toxicity.

It has been said that a pill without side effects is a pill that doesn’t work. This axiom aside, we should choose the medicines we give our older patients carefully and rethink both diagnoses and therapeutic plans before we end up making a treatment worse than the presenting disease.


Mr. Zimmerman is a full-time practitioner with Esopus Medical, PC, an independent family practice in Rifton, N.Y., and a contributing editor to The Clinical Advisor.



1. Kaufman D, Kelly JP, Rosenberg L, et al. Recent patterns of medication use in the ambulatory adult population of the United States: The Slone survey. JAMA. 2002;287:337-344.

2. Rochon PA. Drug prescribing for older adults. In: Rose BD, ed. UpToDate. Wellesley, Mass.: UpToDate; 2008.

3. Beers MH, Ouslander JG, Rollingher I, et al. Explicit criteria for determining inappropriate medication use in nursing home residents: UCLA Division of Geriatric Medicine. Arch Intern Med. 1991;151:1825-1832.

4. Zhan C, Sangl J, Bierman A, et al. Potentially inappropriate medication use in the community-dwelling elderly: findings from the 1996 Medical Expenditure Panel Survey. JAMA. 2001;286:2823-2829.

5. Budnitz DS, Shehab N, Kegler SR, Richards CL. Medication use leading to emergency department visits for adverse drug events in older adults. Ann Intern Med. 2007;147:755-765.