AGE-ASSOCIATED CHANGES IN PHARMACOKINETICS
AGE-ASSOCIATED CHANGES IN PHARMACODYNAMICS
Persons aged 65 years or older are prescribed the highest proportion of medications in relation to their percentage of the U.S. population. Currently, approximately 13% of the U.S. population is aged 65 years or older; this age group purchases 33% of all prescription drugs. These figures are expected to increase to 25% and 50%, respectively, by the year 2040.
Drugs are the most common treatment for acute and chronic diseases. They are also used to prevent many of the diseases and disorders experienced by the older adult. Pharmacotherapy that is successful requires the correct drug at the correct dosage, for the correct disease or condition, for the correct patient. Unfortunately, achieving these goals is not simple or easy. Many other factors come into play, including the patient’s other disease states, other medications, adherence, beliefs, functional status, physiologic changes due to aging and disease, and ability to afford the medication. The basic principle of prescribing for older patients—briefly, start low, go slow—is repeated often. However, even the best clinicians who carefully adhere to this principle will encounter patients who have negative outcomes from one or more of their medications.
The principles of pharmacotherapy have not changed significantly during the past 20 years, but drug treatment has become more complex. More drugs are available every year, some with a new pharmacologic profile or mechanism of action. In addition, physicians must contend with expanded indications for available agents, both those that are approved by the U.S. Food and Drug Administration (FDA) and those that are off-label. Physicians also must be responsive to frequent changes in the managed-care formulary, the scientific advances in the understanding of drug-drug interactions (ie, the cytochrome P-450 system), the change of many drugs from prescription to nonprescription, and the boom in an unregulated third class of medications called nutriceuticals, that is, nutritional supplements, alternative medicines, and herbal preparations. Finally, very little information is available about the use of these unregulated medications in older patients—particularly, sick older patients on other medications.
Pharmacokinetic studies define the time course of a drug and its metabolites throughout the body with respect to four parameters: absorption, distribution, metabolism, and elimination. The effects of aging on each parameter have been studied, and the resulting generalizations have been incorporated into the principles of prescribing for the older patient.
Aging does not affect drug absorption via the gastrointestinal tract to any clinically significant degree. The rate of absorption may be slowed with age, but the extent of absorption remains unchanged. Consequently, the peak serum concentration of a drug in the older patient may be lower and the time to reach it delayed, but the overall amount absorbed (bioavailability) does not differ in younger and older patients. Exceptions include drugs that undergo an extensive first-pass effect (eg, nitrates); they tend to have higher serum concentrations or increased bioavailability, since less drug is extracted by the liver as a consequence of decreased liver size and blood flow.
Factors that have a greater impact on drug absorption include the way a medication is taken, what it is taken with, and a patient’s comorbid illnesses. For example, the absorption of many fluoroquinolones (eg, ciprofloxacin) is reduced when they are taken with divalent cations such as calcium, magnesium, and iron that are found in antacids, sucralfate, dairy products, or vitamins. Enteral feedings interfere with the absorption of some drugs (eg, phenytoin). An increase in gastric pH from proton-pump inhibitors, H2 antagonists, or antacids may increase the absorption of some drugs, such as nifedipine and amoxicillin, and decrease the absorption of other drugs, such as the imidazole antifungals, ampicillin, cyanocobalamin, and indinavir. Agents that promote or delay gastrointestinal motility, such as stimulant laxatives and metoclopramide, can, in theory, affect a drug’s absorption by increasing or decreasing the time spent in the segment of the intestinal tract necessary for dissolution or absorption. Another mechanism that can increase or decrease drug absorption is the inhibition or induction of enzymes in the gastrointestinal tract (this is discussed in the section on drug interactions).
Distribution refers to the locations in the body a drug penetrates and the time required for the drug to reach those locations. Distribution is expressed as the volume of distribution (Vd), with units of volume (eg, liters) or volume per weight (eg, L/kg).
Age-associated changes in body composition can alter drug distribution. In older patients, drugs that are water soluble (hydrophilic) have a lower volume of distribution, as older people have less body water and lean body mass. Drugs affected include ethanol and lithium. Digoxin, which distributes and binds to skeletal muscle, has been reported to have a reduced volume of distribution in older persons because of the reduced muscle mass of older adults. Drugs that are fat soluble (lipophilic) have an increased volume of distribution in older persons because they have greater fat stores than do younger persons. Thus, it takes longer for an older patient taking a lipophilic drug to reach a steady-state concentration and longer for the drug to be eliminated from the body. Examples of fat-soluble drugs include diazepam, flurazepam, thiopental, and trazodone.
The extent to which a drug is bound to plasma proteins also influences its volume of distribution. Albumin, the primary plasma protein to which drugs bind, is often decreased in older patients; thus, a higher proportion of drug is unbound (free) and pharmacologically active. Drugs that bind to albumin and whose unbound fraction has been shown to be increased in elderly persons include ceftriaxone, diazepam, lorazepam, phenytoin, valproic acid, and warfarin. Normally, additional unbound drug is eliminated; however, age-related decreases in the organ systems of elimination may result in the accumulation of unbound drug in the body. Phenytoin provides an example of the way an increase in unbound drug can lead to an unnecessary and potentially harmful dosage increase. A patient with a low serum albumin (≤ 3 g/dL) whose phenytoin dose is increased because his or her total phenytoin concentration is subtherapeutic may develop symptoms and signs of phenytoin toxicity after a dose increase because the concentration of free phenytoin is elevated.
The liver is the most common site of drug metabolism, but metabolic conversion also can take place in the intestinal wall, lungs, skin, kidneys, and other organs. Aging affects the liver by decreasing liver blood flow as well as decreasing liver size and mass. Consequently, in the older patient the metabolic clearance of drugs by the liver may be reduced. Drug clearance is also reduced with aging for drugs that are subject to the phase I pathways or reactions, which include hydroxylation, oxidation, dealkylation, and reduction. Most drugs metabolized through phase I pathways may be converted to metabolites of lesser, equal, or greater pharmacologic effect than the parent compound (eg, diazepam). Drugs metabolized through the phase II pathways are converted to inactive compounds through glucuronidation, conjugation, or acetylation (eg, lorazepam). Medications subject to phase II metabolism are generally preferred for older patients, as their metabolites are not active and will not accumulate.
Age and gender differences also have been reported. For example, oxazepam is metabolized faster in elderly men than in elderly women. The reason is unknown. Nefazodone concentrations have been reported to be 50% greater in older women, but no differences were found between older men and younger persons.
In drug metabolism, factors other than aging can exaggerate or override the effects of aging. For example, hepatic congestion due to heart failure decreases the metabolism of warfarin, resulting in an increased pharmacologic response. Smoking stimulates monooxygenase enzymes and increases the clearance of theophylline even in older patients.
Elimination refers to a drug’s final route(s) of exit from the body. For most drugs, this involves elimination by the kidney as either the parent compound or as a metabolite or metabolites. Terms used to express elimination are a drug’s half-life and its clearance.
A drug’s half-life is the time it takes for its plasma or serum concentration to decline by 50%, for example, from 20 µg/mL to 10 µg/mL. Half-life is usually expressed in hours. Steady state is reached when the amount of drug entering the systemic circulation is equal to the amount being eliminated. For a drug administered on a regular basis, 95% of steady state in the body is achieved after five half-lives of the drug.
Clearance is usually expressed as volume per unit of time (eg, L/hour or mL/minute) and represents the volume of plasma or serum from which the drug is removed (ie, cleared) per unit of time. Clearance may also be expressed as volume per weight per unit of time (L/kg/hour). Half-life and clearance can also refer to metabolic elimination.
The effects of aging on kidney function have been studied to a greater extent than have the effects of aging on liver function. Glomerular filtration declines as a consequence of a decrease in kidney size and renal blood flow and a decrease in functioning nephrons. On average, kidney function begins to decline when people reach their mid-30s, with an average decline of 6 to 12 mL per minute per 1.73 m2 per decade. Follow-up studies (conducted in men only) over 10 to 15 years found three normally distributed groups: those whose creatinine clearance declined to the extent that it was clinically significant; a second group whose creatinine clearance declined to the extent that it was statistically but not clinically significant; and a third group whose creatinine clearance did not change. Renal tubular secretion also declines with age.
Serum creatinine is not an accurate reflection of creatinine clearance in elderly patients. Because of the age-related decline in lean muscle mass, the older person’s production of creatinine is reduced. The decrease in glomerular filtration rate counters the decreased production of creatinine, and serum creatinine stays within the normal range, not revealing the change in creatinine clearance.
The conservative approach in treating the older patient is to calculate the appropriate dose for renally eliminated medications as if the patient’s kidney function actually has declined with aging. Measuring a patient’s 24-hour creatinine clearance would be the most accurate way to determine the appropriate dose, but this is time consuming and requires an accurate 24-hour urine collection. An 8-hour collection time has been shown to be accurate but has not been widely accepted.
To initially estimate a patient’s creatinine clearance (CrCl), the clinician can use the Cockroft and Gault equation (see below).
(140 − age) × weight
CrCl = ——————————————
72 × serum creatinine
Weight in kg; serum creatinine in mg/100 mL; 85% less in women.
The equation is widely applied, but it has limitations. First, not all patients experience a significant age-related decline in renal function, and for them, the equation would underestimate creatinine clearance. Second, for patients whose muscle mass is reduced beyond that of normal aging, the creatinine clearance would be overestimated. This would apply to persons whose serum creatinine is less than normal, that is, < 0.7 mg/dL. It has been suggested that 1 mg/dL be substituted for a low serum creatinine. However, normalizing the serum creatinine has not been shown to be a precise estimate, and it generally underestimates the actual creatinine clearance.
In cases in which the patient’s kidney function may be impaired but estimates of function are uncertain, the clinician should consider the following:
The pharmacodynamic action of a drug—that is, its time course and intensity of pharmacologic effect—may change with the increasing age of the patient. An excellent example of such pharmacodynamic changes in older persons has been demonstrated with the benzodiazepines. Older persons have been found to have more sedation and lower performance than younger persons on a psychomotor test following a single dose of triazolam. These differences are attributed to pharmacokinetic changes, that is, to significantly higher plasma triazolam concentrations that are due to reduced clearance in the elderly persons. However, a different pattern has been found for nitrazepam: The pharmacokinetics of nitrazepam (an intermediate-acting benzodiazepine similar to lorazepam) were found to be no different in young and older persons after a single 10-mg dose; yet, 12 hours and 36 hours after a 10-mg dose, older persons were found to make significantly more mistakes on a psychomotor test than when they had taken placebo. Younger persons did not demonstrate significant impairment at any time. In addition, even with short-term use, young and elderly patients may experience impaired balance and posture following a single dose of a benzodiazepine.
It is uncertain whether the age-associated pharmacokinetic changes of morphine account for the increased level and prolonged duration of pain relief experienced by older patients. In older adults, morphine has been shown to have a smaller volume of distribution, higher plasma concentrations, and longer clearance than in younger adults. Older patients achieve pain relief at least equivalent to younger patients at half the intramuscular dose, and they have an increased duration of pain relief. Thus, the dose or frequency, or both, of morphine given intramuscularly or by intravenous infusion should be lower, at least initially, in older patients.
Pharmacodynamic and pharmacokinetic changes, alone or together, generally result in an increased sensitivity to medications by the older adult. In some patients, particularly those who are frail, the use of lower doses, longer intervals between doses, and longer periods between changes in dose are ways to successfully manage drug therapy and decrease the chances of medication intolerance or toxicity. Disease- and drug-specific monitoring are also necessary to ensure a successful outcome.
Optimizing drug therapy for older adults means achieving the balance between overprescribing and underprescribing of beneficial therapies. Overprescribing of drug therapies not only refers to the use of multiple medications but also implies a lack of appropriateness in medication selection, dose, or use. One survey found that 40% of nursing-home residents had an order for at least one potentially inappropriate medication. Analyses of national medication use surveys in the ambulatory setting have consistently shown that > 20% of older patients received at least one potentially inappropriate medication, with at least one potentially inappropriate medication prescribed at ~ 8% of office visits. Furthermore, nearly 4% of office visits and 10% of medical hospital admissions resulted in a prescription for one or more medications classified as “never” or “rarely appropriate” for older patients. The potential consequences of overprescribing include adverse drug events (ADEs), drug-drug interactions, duplication of drug therapy, decreased quality of life, and unnecessary costs.
The factors associated with inappropriate prescribing or overprescribing are listed in Table 10.1. Simply limiting the number of medications for a given patient, however, is not always possible or desirable. For example, a patient with heart failure may be appropriately treated with three or four drugs: a diuretic, an angiotensin-converting enzyme (ACE) inhibitor, a β-blocker, and perhaps digoxin. If this patient has hyperlipidemia and diabetes mellitus, another two or three medications could be required. Hence, such a patient would be taking five to seven indicated medications for major medical conditions alone.
The underprescribing of medications to older adults is also of concern. Underprescribing may result from an effort to avoid overprescribing, a complex medication regimen, or adverse effects. It may also result from the thinking that older adults will not benefit from medications intended as primary or secondary prevention, or from aggressive management of chronic conditions, such as hypertension and diabetes mellitus.
Medications often cited as underprescribed in older adults include ACE inhibitors and β-blockers for heart failure and at discharge following an acute myocardial infarction (MI), aspirin within 24 hours and at discharge following an acute MI, warfarin for atrial fibrillation, HMG–CoA reductase inhibitors for primary prevention of cardiovascular events, gastroprotective agents for patients at high risk for NSAID-induced gastrointestinal bleeding, and narcotic analgesics for pain control.
Underprescribing was recognized by the Centers for Medicare and Medicaid Services (CMS), which implemented quality indicators in 1998 to measure the use of recommended medications prior to discharge for acute MI, stroke, and pneumonia in hospitalized patients with fee-for-service Medicare. Comparison of data from 1998–99 to 2000–2001 found that the percentage of patients receiving appropriate care had improved for the nation as a whole in the management of these conditions, but the prescribing of an ACE inhibitor in heart failure had declined. In spite of this overall improvement, between 16% and 29% of patients did not receive appropriate pharmacotherapy prior to discharge after their MI.
Investigators from the Assessing Care of Vulnerable Elders (ACOVE) project developed explicit medication quality indicators divided into four categories: prescribing indicated medications; avoiding inappropriate medications; education, continuity, and documentation; and medication monitoring. These indicators were applied to “vulnerable” older adults enrolled in two managed care organizations and the results were reported as the percentage of eligible patients who met the indicator or “pass rate.” The prescribing of indicated medications category had an overall pass rate of 50% (range 11% to 94% for the 17 indicators); avoiding the prescription of inappropriate medications had an overall pass rate of 97% (range 79% to 100% avoidance across the 9 indicators). The overall pass rate for education, continuity, and documentation indicators was 81% (range 10% to 99% for the 8 indicators). The 9 indicators for medication monitoring had an overall pass rate of 64% (range 22% to 80%). These results suggest that the prescribing of inappropriate medications is less of a problem than the underprescribing of indicated medications and the failure to monitor medications, document information about medications, maintain continuity, and educate patients.
An ADE is defined as an injury resulting from the use of a drug. Preventable adverse drug events are among the most serious consequences of inappropriate drug prescribing among older adults. An adverse drug reaction (ADR) is a type of ADE; it refers to harm that is directly caused by a drug at usual doses. For a listing of risk factors for ADEs in older patients, see Table 10.2.
ADEs are estimated to be responsible for 5% to 28% of acute geriatric medical admissions; the estimated annual incidence rate is 26 per 1000 beds for hospitalized patients. It has been estimated that in the nursing home, for every dollar spent on medications, $1.33 in health care resources is consumed in the treatment of drug-related morbidity and mortality. A cohort study of all long-term-care residents in 18 nursing homes in Massachusetts demonstrated that ADEs are common and often preventable in nursing homes. During the 28,839 resident-months of observations, 546 ADEs were identified. Overall, 51% of these adverse events were judged to have been preventable. Most of the errors occurred at the ordering and monitoring stages. A cohort study of residents of two long-term-care facilities found the overall rate of ADEs to be 9.8 per 100 resident months. Atypical antipsychotics, anticoagulants, and diuretics were the drug classes found to be most frequently associated with ADEs.
In the ambulatory setting, the ADE rate has been reported to be 50.1 per 1000 person-years, and the preventable ADE rate to be 13.8 per 1000 person-years. Cardiovascular drugs, diuretics, NSAIDs, hypoglycemics, and anticoagulants are the drug classes found to be most often associated with preventable ADEs. Again, errors occurred most often at the time of prescribing or were related to inadequate monitoring. The majority of ADEs (≥ 95%) experienced by older patients are considered to be predictable.
Lists of potentially inappropriate medications and medication-disease pairs have been published. One list that was developed by the consensus of experts in geriatric medicine and pharmacology (1997 Beers criteria) has been adapted by the CMS for identifying potentially inappropriate or unnecessary medication use by nursing-home residents. An updated version was published in late 2003 (see reference by Fick, 2003, in the list at the end of the chapter); however, CMS has not adopted this list as of the time of this writing. (For the current version, also see the most recent edition of Geriatrics At Your Fingertips, published annually by the American Geriatrics Society.)
A common pathway for ADEs and polypharmacy has been described as the “prescribing cascade.” One form of this cascade occurs when a medication results in an ADE that is mistaken as a separate diagnosis and treated with more medications, which puts the patient at risk for additional ADEs and more medications. Examples that have been studied include metoclopramide-induced parkinsonism and the subsequent prescribing of antiparkinson medications, and calcium channel blockers that result in peripheral edema and the subsequent use of diuretics.
A drug-drug interaction (DDI) is defined as the pharmacologic or clinical response to the administration of a drug combination that differs from that anticipated from the known effects of each of the two agents when given alone. Drug-drug interactions are important because they may lead to ADEs. The likelihood of DDIs increases as the number of medications a patient takes increases. Among prescription drugs, cardiovascular and psychotropic drugs are most commonly involved in DDIs. A positive correlation exists between the number of potential DDIs and the number of adverse effects experienced by hospitalized older patients. The most common adverse effects are neuropsychologic (primarily delirium), arterial hypotension, and acute kidney failure. Drug combinations that are reported to result in increased risk for hospitalization for older patients are shown in Table 10.3. Drugs commonly involved in DDIs in long-term care are listed in Table 10.4. Risk factors associated with DDIs include the use of multiple medications, receiving care from several prescribing clinicians, and using more than one pharmacy.
Drug interactions can take many forms. For example, absorption can be enhanced or diminished (as described above), drugs with similar or opposite pharmacologic effects can result in exaggerated or impaired effects, and drug metabolism may be inhibited or induced. Research focusing on the cytochrome P-450 system has proposed or studied in vivo or in vitro numerous DDIs involving the different P-450 isozymes. The effect of aging on the cytochrome P-450 system and the clinical implications for prescribing have not been completely determined. Cross-sectional data have shown that cytochrome P-450 content declines incrementally, once in the fourth decade and again after age 70. In vitro microsomal activity of cytochrome (CYP) 3A4 is not altered by aging, but in vivo age- and gender-related reductions in drug clearance have been found for CYP3A4 substrates erythromycin, prednisolone, verapamil, alprazolam, nifedipine, and diazepam. CYP3A4 accounts for 30% of the P-450 content in the liver and is also prominent in the intestinal tract. This isozyme is involved in the metabolism of more than 50% of drugs on the market and can be induced by drugs such as rifampin, phenytoin, and carbamazepine, and inhibited by many drugs, including the macrolide antibiotics, nefazodone, itraconazole, ketoconazole, as well as grapefruit juice. The isozyme CYP2D6 is involved in the metabolism of 25% to 30% of marketed drugs and has been associated with only minimal age-related changes. CYP2D6 is involved in the metabolism of many psychotropic drugs and can be inhibited by many agents. In addition, approximately 10% of white people are deficient in this isozyme and have reduced ability to clear and increased sensitivity to CYP2D6 substrates. Clinically, these patients and those taking CYP2D6 inhibitors (eg, quinidine, paroxetine, fluoxetine) cannot convert codeine and tramadol to their active metabolites and have a reduced analgesic response to these agents.
For DDIs involving herbal preparations, see Complementary and Alternative Medicine.
Drug-disease combinations common in older patients can affect drug response and lead to adverse drug events. Obesity and ascites alter the volumes of distribution of lipophilic and hydrophilic drugs, respectively. Patients with dementia may have increased sensitivity or paradoxical reactions to drugs with central nervous system or anticholinergic activity. Patients with renal insufficiency or impaired hepatic function due to cirrhosis or hepatic congestion have impaired detoxification and excretion of drugs. (See reference by Fick, 2003, at the end of the chapter.)
Principles of prescribing for older patients are shown in Table 10.5. This basic approach applies primarily to medications that will be used to treat chronic conditions for which an immediate, complete therapeutic response is not necessary. A dose adjustment may still be needed for medications used to treat conditions requiring an immediate response (eg, when prescribing antibiotics for a patient with impaired kidney function).
Overprescribing can be prevented by reviewing a patient’s medications on a regular basis and each time a new medication is started or a dose is changed. The importance of maintaining accurate records of all medications taken by the patient cannot be overemphasized. Knowing what other prescribers have given the patient and what patients are self-prescribing, and documenting all of this in the patient’s records is crucial. Many patients do not consider vitamins, herbal preparations, or over-the-counter medications (even aspirin) to be medications, so clinicians must be specific when they inquire about a patient’s use of other medications. Medications that have been newly approved by the FDA should be used cautiously in treating older patients. Such drugs are likely to be more expensive, and information about their use in older patients is often limited.
It is best if the patient brings all medications to the review, including over-the-counter medications, vitamins, and any herbal preparations or other types of supplements (a “brown-bag” evaluation). Examining the containers and labels and asking what each medication is for and how and when it is taken can provide insight into the patient’s understanding and adherence to his or her medication regimens. Discontinue any medications when there is no longer an indication for their continued use. Any new complaints or worsening of an existing condition should prompt the consideration of whether it could be drug-induced. When considering treatment for a new medical condition, always consider nonpharmacologic approaches first. If a drug therapy is still indicated, select within the class in order to minimize the risk of an ADE.
When initiating therapy, the basic principle should be to start low go and go slow. Though the FDA requires that labeling for new drugs regarding dosing in older patients not be extrapolated from another patient population (eg, patients with kidney impairment), it does not require that a drug be studied explicitly in older persons. Older persons are used in the studies of many new drugs for their phase I and II dose tolerability and their pharmacokinetic and pharmacodynamic actions, but the older persons chosen for these studies are usually healthy and free of concomitant illnesses. Much of what is known about medications and how to use them in sick older patients, particularly those who are frail, is learned only after a drug has been available for several years.
Nonadherence and underadherence to medications is a huge and often unrecognized problem in drug treatment. It is estimated that nonadherence among elderly people may be as high as 50%. Patients may be reluctant to admit that they are not taking medications or not following directions. If nonadherence is suspected, the clinician needs to consider the patient’s financial, cognitive, and functional status, as well as his or her beliefs about and understanding of medications and diseases.
Prescription drug costs have increased substantially, and supplemental prescription drug benefit plans are expensive. Some plans still may leave a patient with a copayment that he or she cannot afford or with only a fixed dollar amount for the year. Clinicians should avoid prescribing expensive new medications that have not been shown to be superior to less expensive generic alternatives.
Cognitive impairment may also cause nonadherence, as patients forget to take medications or confuse them. Simplifying the regimen and involving a caregiver to oversee medication management can be helpful approaches. Medication trays or drug calendars also may help with organization, and they are very useful for patients who have difficulty remembering when they last took a medication.
The older patient’s ability to read labels, open containers, or pour medications or even a glass of water may be impaired, so functional assessment can be useful. Some patients may need additional education or reinforcement about the purpose of a medication, especially those used to treat conditions that are usually asymptomatic, such as diabetes mellitus and hypertension. Older patients also may need reassurance regarding the safety and possible adverse effects of certain medications, particularly newly prescribed medications or those associated with serious adverse events, such as warfarin.
■ Fick DM, Cooper JW, Wade WE, et al. Updating the Beers criteria for potentially inappropriate medication use in older adults: results of a U.S. consensus panel of experts. Arch Intern Med. 2003;163(22):2716–2724.
Recognizing that several of the drugs on the Beers criteria were no longer available or their use had declined substantially, the authors convened a group of experts in geriatrics and geriatric pharmacotherapy to revise and update the Beers criteria by the use of a modified Delphi method. The panel eliminated 15 medications or medication classes from the 1997 Beers criteria. The panel added 25 medications or medication classes independent of diagnosis and another 19 medications or medication classes considering diagnosis. This update has not been adopted by the Centers for Medicare and Medicaid Services, but it is a valuable tool for clinicians.
■ Goulding MR. Inappropriate medication prescribing for elderly ambulatory care patients. Arch Intern Med. 2004;164(3):305–312.
The author applies the 1997 Beers criteria and another modified version of the Beers criteria referred to as the Zhan list to the National Ambulatory Medical Care Survey and the National Hospital Ambulatory Medical Care Survey for the years 1995 to 2000. The findings show that over the 5-year period the prevalence of inappropriate prescribing did not change, that women were more likely and patients aged > 80 years were less likely to be prescribed an inappropriate medication. The odds of being prescribed an inappropriate medication increased with each prescription drug added to the patient’s regimen. Propoxyphene, diazepam, amitriptyline, hydroxyzine, and oxybutynin were the most frequently prescribed inappropriate medications. Family and general practice physicians were more likely than other specialists to prescribe inappropriate medications.
■ Gurwitz JH, Field TS, Harrold LR, et al. Incidence and preventability of adverse drug events among older persons in the ambulatory setting. JAMA. 2003;289(9):1107–1116.
This cohort study was designed to determine the incidence and prevalence of adverse drug events in all ambulatory Medicare enrollees in a multispecialty practice. The frequency of adverse events was 50.1 per 1000 person-years; the rate of preventable adverse events was 13.8 per 1000 person-years. More than half of preventable events were due to errors in prescribing and monitoring. This study serves as evidence that a portion of adverse drug events are preventable.
■ Murray MD, Callahan CM. Improving medication use for older adults: an integrated research agenda. Ann Intern Med. 2003;139(5 Pt 2):425–429.
In this overview of the premarketing and postmarketing phases of drug development and study, the authors describe how pharmaceutical advances are an integral part of successful aging. Ways that medication use can be improved through a research agenda are discussed and include drug discovery and delivery, drug efficacy and safety, pharmacoepidemiology and drug policy, along with improved access to and use of drugs.
■ Rollason V, Vogt N. Reduction of polypharmacy in the elderly: a systematic review of the role of the pharmacist. Drugs Aging. 2003;20(11):817–832.
This systematic review covers the definitions and prevalence of polypharmacy, its causes and consequences, and the impact of pharmacists on its reduction. Fourteen clinical trials were identified that included an intervention by a pharmacist. Interventions most often included a review of the patient’s medication profile followed by the communication to the patient’s prescriber of recommendations to reduce polypharmacy. Seven of the fourteen trials were shown to significantly reduce the number of medications taken by patients in the intervention groups. Other trials did not reduce the number of medications but improved overall care by eliminating unnecessary medications and adding medications for previously undiagnosed or undertreated conditions. The authors review the limitations of these trials and offer recommendations for further research on interventions to reduce polypharmacy.
Todd P. Semla, PharmD, MS
Paula A. Rochon, MD, MPH