33: Geriatric Principles

The population is aging steadily across the world. In the United States, people older than 65 years of age comprise not only an increasing proportion of the population at large (13%) but also an increasing proportion of patients seen in medical practices. Compared with all other age groups, patients older than 65 years of age account for one-fourth of emergency department (ED) ambulance arrivals with the highest proportion of patients in EDs triaged as emergent¹⁰² and the highest number of hospital and intensive care unit admissions.¹¹⁷

Although the elderly account for only a small minority of toxicologic exposures, when exposed, they have a high mortality rate. Among exposures reported to the American Association of Poison Control Centers (AAPCC), the fatality ratio for adults (ie, number of cases divided by number of deaths) exhibits a bimodal pattern, declining after age 30 until age 60, when it again rises (Chap. 136). The factors associated with the increased fatality ratio after age 60 are complex but are likely due in part to physiologic vulnerability.

In a separate study, seven specific pharmaceuticals were selected from the AAPCC database for analysis based on their prevalent use and potential toxicity from 1995 through 2002. These pharmaceuticals were theophylline, digoxin, benzodiazepines, tricyclic antidepressants (TCAs), calcium channel blockers, acetaminophen (APAP), and salicylates.¹¹⁴ The death rate from intentional or unintentional exposure to these pharmaceuticals was found to increase by 35% for each decade of life after age 19 years.¹¹⁴ Although prescribing for some of these drugs has dramatically decreased over time, specific categories continue to pose problems for patients in the latest decades of life.^14,62

Exposures reported to the AAPCC may underestimate the serious consequences for elderly people exposed to xenobiotics that are toxic or potentially toxic. Data from the National Electronic Injury Surveillance System–Cooperative Adverse Drug Event Surveillance Project (NEISS-CADES) indicate that patients aged 65 years and older accounted for 25% of estimated visits related to adverse drug events (ADEs) and almost 50% of such visits requiring hospitalization or prolonged monitoring in the ED during 2004–2005.¹⁵ Furthermore, the incidence of ADEs increases steeply from age 65 years through the tenth decade of life.¹⁵ More recent NEISS-CADES data indicate that almost 50% of elders who required emergency hospitalization as a result of ADEs are people 80 years of age or older.¹⁴ The problem may be even greater than the NEISS-CADES study suggests, because their data did not capture ADEs in patients treated or dying outside of EDs, ADEs that could only have been recognized after admission, or ADEs that might be erroneously diagnosed as non–drug-related problems.

Toxic exposures in the elderly may be underrecognized for several reasons. First, because of pharmacokinetic and pharmacodynamic changes that occur with aging,³³ which a “standard” therapeutic dose may produce as an unexpected serious effect. Second, the presentation of disease, including drug toxicity, is often atypical in the elderly.⁷¹ For example, falls may be a presenting sign of toxicity in the elderly, with prescribed sedative–hypnotics, opioids, antipsychotics, antidepressants, and certain cardiovascular xenobiotics among the most often associated with an increased risk of falling.^{46,107,115,133}

If the patient is cognitively impaired and the fall is unwitnessed, the immediate consequences of the fall may be adequately addressed, but the xenobiotic and primary etiology causing it may not.⁸⁰ Another important syndrome in older patients is delirium, which may be caused by many factors but is a common presentation of xenobiotic toxicity.⁵⁹ A large variety of xenobiotics may cause mental status changes, which may mistakenly be attributed to non–xenobiotic-related causes in the elderly.¹²⁸ Conversely, altered mental status may be unrecognized^67,89 or misdiagnosed—for example, as Alzheimer’s disease or psychiatric illness.¹¹¹ A striking case of drug-induced mental status changes is represented by a previously normal 66 year-old man who developed psychosis and attempted suicide after taking increasing doses of dextromethorphan-containing cough syrup for a respiratory infection. The patient was treated with psychiatric medications, and several months elapsed before providers realized that his behavior was xenobiotic-induced rather than due to a primary psychiatric condition.⁹³

Finally, the presentation of drug toxicity may be delayed in elderly individuals. Drugs with long half-lives may not reach a steady state and hence not achieve peak effects until days after the drug therapy is initiated. In some older patients, the active metabolite of flurazepam, desalkylflurazepam, has a half-life of up to 100 hours or longer, which requires days to achieve a steady state.⁵³ When peak effects are delayed in this way, drug ­toxicities may easily be mistaken for non–xenobiotic-related illnesses.

Table 33–1 lists xenobiotics commonly responsible for toxicity in the elderly.

The risk of suicide in men increases with age in a bimodal distribution, rising steadily after age 70. Although data for individual ethnic groups are limited, white men have a substantially higher risk of suicide than age-matched cohorts among the African American population.²⁵ Another high-risk group appears to be Native American and Alaskan Native men, although the number of deaths from all causes in elderly men and certain other groups is too low to make meaningful comparisons.²⁵ Among people 65 years of age and older in the United States, white men have the highest rate of suicide of all groups, with almost 32 suicides per 100,000. The rate in the next highest risk group, Asian/Pacific Islander men, is almost 15 suicides per 100,000.²⁵ In women, the suicide rate peaks in midlife and then declines. However, for all age groups, women have a much lower rate of suicide than men; in 2010, for all races combined, the rate for men was 29 suicides per 100,000 and for women was 4.2 suicides per 100,000.²⁵

In the United States, firearms account for far more suicide-related deaths than do poisonings or other methods, particularly in men. The proportion of deaths due to firearm increases steadily after age 50, among men; by age 70 more than 75% of male suicides are by firearms, when all races are combined.²⁵ Suicides by poisoning are much less common in men than women at all ages. In women, poisonings account for similar or greater proportions of suicide than firearms.

Xenobiotic overdose is an important factor in suicide attempts by the elderly of both sexes.⁴⁷ Although less likely than men to complete suicide, women are more likely to attempt it.²⁶ The male-to-female ratio of suicide attempts narrows with increasing age, and in the oldest age groups, men attempt suicide slightly more often than women, when all methods of attempted suicide are considered (Chap. 27).^26,124

Geographic and cultural factors also determine the method of suicide. In countries other than the United States, firearms are much less frequently used to commit suicide at any age, with suicide by oral overdose, toxic inhalation, or hanging comprising a much greater proportion.^2,130 Among the elderly, the pattern of xenobiotics responsible for suicidal deaths may be changing because selective serotonin reuptake inhibitors are increasingly prescribed for depression instead of TCAs.²⁰ In the United States, higher suicide rates are associated with greater use of TCAs than non-TCA antidepressants,⁵¹ which could be related to greater lethality of TCAs in overdose or poorer tolerability of TCAs leading to nonadherence and enhanced suicide risk. In a Swedish study examining autopsy-confirmed suicides that included analysis of legal xenobiotics, the incidence of suicidal fatalities attributed to benzodiazepines was reported to be increasing despite a marked decrease in benzodiazepine prescriptions.²⁰ Overdose of benzodiazepines is rarely fatal unless it is accompanied by alcohol or another toxin or occurs in the presence of serious medical conditions.⁴⁰ However, the likelihood of fatality from an overdose of benzodiazepine taken alone increases markedly with each decade of life,¹¹⁴ perhaps because benzodiazepines in frail elderly people commonly lead to secondary morbidity and mortality from aspiration pneumonia, falls including hip fracture, and other medical complications that may be the proximate cause of death in the case of overdose.

Notably, flunitrazepam and nitrazepam, the two most frequently implicated in single benzodiazepine suicides in Sweden,²¹ are not available in the United States. However, single-drug suicides have been reported with other benzodiazepines that are available, such as flurazepam, triazolam, diazepam, and oxazepam.^21,92,94

Substance abuse declines with age⁴² but is important to consider in older adults under relevant clinical circumstances. Alcohol is the most common substance of abuse in people older than age 65 years. However, in recent years, an increase in the use of illicit xenobiotics and prescription drug misuse has been documented among older adults.¹⁴⁷

Currently, this change is largely noted in adults 50 to 64 years of age, as opposed to those 65 and older;¹²⁷ however, with increasing numbers of people entering the seventh decade of life (including those with past history of abuse disorders), illicit and nonmedical use is expected to increase substantially in older age groups. Unfortunately, a strong case can be made that middle-aged substance abusers have an “accelerated rate of biologic aging” than their nonabusing peers⁷³ and may be less likely to survive into the seventh decade, or, if they do, may be frailer than their same-age peers. However, approximately 22% of people 65 years of age and older use one or more prescription medications with abuse potential, including opioids and other central nervous system depressants.¹²¹

Analgesics are among the most commonly prescribed group of drugsamong elderly people, second only to cardiovascular drugs. In one survey of community residing elderly, 23%, 21%, and 52% of those aged65 to 74, 75 to 84, and 85 and older, respectively, had some exposure to opioids.⁹⁹ Although only a small proportion of people who take opioids are likely to abuse them, concern exists that long-term use could ensue once a prescription is written. In a recent study of elderly opioid-naive patients undergoing low-risk surgery, those who received a prescription within a week of surgery were 44% more likely to become long-term users within the first year than those who were not given an opioid.³ Although the surgeries themselves were largely for conditions not associated with chronic pain, one cannot conclude that opioid use constituted “misuse,” because the study did not report specific reasons for the ongoing analgesic use. Notably, new users of nonsteroidal antiinflammatory drugs (NSAIDs) were far more likely than opioid users to continue analgesic use in this elderly group of patients,³ who would have been likely to experience chronic pain from a variety of causes.²⁸

Abuse of alcohol or other xenobiotics may be a continuation of long-term habits, but some individuals may first begin later in life.⁷⁵ Substance abuse in late life is probably underrecognized and underreported.¹¹² In one Illinois study of patients presenting to a trauma facility, only 5% of those aged 65 years and older were tested for alcohol or other substances, but among those younger than age 65, 22% were tested for alcohol and 29% for other xenobiotics.¹⁴⁹ Among elderly patients presenting to trauma facilities a large majority have a blood alcohol concentration (BAC) of 80 mg/dL or above.^118,149 Such results, while similarly found in nonelderly trauma patients,¹⁴⁹ have different implications in the older adult, reflecting a smaller amount of alcohol ingested,^134,138 owing to changes in body composition, and a greater impact on cognitive and motor function, owing to a diminished tolerance for alcohol (Table 33–2).

Elderly patients who are tested for BAC may also have positive urine toxicology screening results, although this occurs less often than in younger adults.¹⁴⁹ In one study, the most common xenobiotics detected in the elderly were benzodiazepines, opioids, and barbiturates,¹⁴⁹ which, like alcohol, are more likely to impair older than younger adults. The failure to consider use of these xenobiotics, whether illicit, nonprescription, or prescription, may have serious consequences. When admitted to hospitals, if a careful history is not elicited, withdrawal from these xenobiotics may be missed or misdiagnosed and be inappropriately managed as a result. Given the ongoing “graying” of the “baby boom” generation, greater vigilance will be needed and screening used appropriately.

Some caveats are in order when it comes to the older alcoholic at risk for withdrawal. Studies suggesting a longer duration or greater severity of the syndrome in the elderly have been criticized for their small size, retrospective nature, and absence of middle-aged comparison groups.⁷⁹ However, vigilance is called for because delayed assessment could lead to greater morbidity.⁴⁵“Hypoactive” delirium, as opposed to typical delirium tremens, may occur, but this could be related to benzodiazepine use, because age-related enhanced sensitivity to these xenobiotics could lead to greater sedation and delirium in elderly who are being treated for withdrawal. Some clinicians may recommend shorter acting benzodiazepines, such as lorazepam rather than chlordiazepoxide,¹¹³ because the latter and its metabolites may be particularly long acting in the elderly and may cause progressive and excessive sedation as the drug and metabolites accumulate. Long-acting benzodiazepines, moreover, are more often associated with injury, including fractures, than short-acting benzodiazepines.¹³¹ Regardless of which benzodiazepine is used for alcohol withdrawal, careful monitoring with symptom-triggered dosing is generally preferable to a fixed-dose schedule in elderly patients, as in all other patients being treated for alcohol withdrawal.

Age-related pharmacokinetic changes have important clinical implications for elderly patients. The most consistent pharmacokinetic change that occurs with aging is a decrease in kidney function. Glomerular filtration rate (GFR) declines, on the average, by 50% between the ages of 30 and 80 years^37,116 and cannot be accurately predicted by serum creatinine, which does not increase significantly with age,¹¹⁶ because muscle mass, the source of serum creatinine, declines with age.^81,103

Because it is impractical and often difficult to measure 24-hour ­creatinine clearance before instituting therapy with a renally excreted xenobiotic, clinicians commonly estimate creatinine clearance using age-adjusted formulas or nomograms. Frequently applied formulas are fairly predictive of GFR when kidney function is stable.¹⁰¹ However, age-related declines in GFR are not universal, and limited data from longitudinal studies suggest that as many as 33% of elderly individuals do not experience this age-related decline.⁹¹ Conversely, predictive formulas could significantly overestimate actual creatinine clearance in chronically ill, debilitated elderly people, especially those with chronic kidney disease (CKD).⁸⁴ Modifications to the Cockcroft-Gault equation to correct for body surface area^69,119 and obesity¹⁴⁴ are proposed to better reflect kidney function (Table 28–7). However, anthropomorphic measurements are unreliable in the elderly,^78,122 and physiologic variability is great. Alternative measurements, such as the Modification of Diet in Renal Disease (MDRD) Study Group equation, serum cystatin C, and others are also proposed, but no current method of estimating kidney function in the elderly appears to be superior for use in the clinical setting.^125,136 For all of these reasons, it is difficult to accurately predict the renal elimination of xenobiotics or their metabolites in the elderly. A practical solution is to assume that kidney function has declined significantly and to exercise caution when prescribing maintenance doses of drugs with a narrow therapeutic-to-toxic ratio (Table 33–3). Failure to do so is an important cause of toxicity.⁸⁷

The nature and degree of impact of age-related hepatic changes on drug elimination is controversial. Liver mass decreases with an associated decrease in hepatic blood flow,¹⁴⁶ which results in decreased efficiency of hepatic extraction. Enzymatic processes are often unpredictable,⁷⁷ and although hepatic oxidation appears to decline with age, this change is difficult to demonstrate. There is substantial genetic variability among cytochrome P450 ­isoenzymes, making it difficult to interpret studies of age-related alterations in oxidative metabolism. Some studies that consider cofactors that could affect hepatic enzymes, such as concurrent xenobiotics or cigarette smoking, or that determine isoenzyme genotype may demonstrate that there is no age-related change in hepatic oxidative enzyme function.¹⁰

Hepatic conjugation does not decline significantly with age, so xenobiotics such as temazepam and oxazepam that are metabolized by these processes do not have prolonged elimination half-lives. In contrast, xenobiotics such as diazepam and flurazepam, which are metabolized by hepatic oxidative enzymes, are eliminated more slowly with age.⁵³ Similar to many oxidized xenobiotics, metabolites of diazepam and flurazepam are active, undergo further metabolism, and remain in circulation after the parent xenobiotic has been metabolized. The active metabolites of some xenobiotics, such as opioids, are renally eliminated, and excretion may be prolonged because of an age-related decline in kidney function (Table 33–3).

Other changes in enzyme systems may occur late in life. For example, a decline in gastric alcohol dehydrogenase activity increases the bioavailability of ingested ethanol in the elderly.¹⁰⁸ The decline in this enzyme is attributed to the increased incidence of gastric atrophy with age. The etiology of the age-related gastric mucosal changes is uncertain, with the relative contributions of underlying Helicobacter pylori infection^7,38 and antiparietal cell antibodies the subjects of research.¹⁵⁰ Whether age-related changes occur in metabolic enzymes that are present in the intestines, brain, ­kidneys, and other organ systems and what impact such changes have on drug disposition and drug interactions are also likely to become active areas of research. Studies to date have not demonstrated a substantial effect of age on P-glycoprotein, a transmembrane transport protein involved in ­xenobiotic interactions in the kidneys, intestines, and other organs.¹¹

Age-related alterations in body composition may affect xenobiotic ­disposition in later life (Table 33–2). For example, lean muscle mass and total body water decline, and the fat-to-lean ratio increases with advancing age.^81,103 Thus, highly lipid-soluble xenobiotics tend to have an increased volume of distribution (Vd). As a result, there may be a delay before steady state is reached, and peak effect and toxicity may occur later than expected as demonstrated with amiodarone and certain benzodiazepines. In contrast, xenobiotics that distribute in water, such as ethanol, have a smaller Vd, leading to higher peak concentrations, accounting at least in part for the more pronounced peak effect of ethanol in the elderly. This may also account for the increased BAC attained in an older adult who drinks an equivalent amount of alcohol as a younger person, as noted above.^134,138

Protein reserve diminishes with age as a consequence of decreased muscle mass and decreased protein synthesis.¹¹⁰ Although serum albumin concentration remains in the normal range in healthy elderly individuals,¹⁹ elderly people are more likely to experience a rapid decrease in albumin concentrations when experiencing acute or chronic illness or when their protein intake diminishes.^34,148 A decline in serum protein concentration increases the free or active fraction of xenobiotics that are otherwise highly protein bound. Free xenobiotic is able to travel more readily to the liver and kidney for metabolism or excretion, so a gradual change in the serum protein concentration is unlikely to lead to a change in the patient’s response to the xenobiotic. However, these changes may be clinically important for interpreting serum concentrations of highly protein-bound xenobiotics. Clinical laboratories typically measure total xenobiotic concentrations, which include both free and bound xenobiotic. Because most xenobiotic is bound, the reported value reflects mostly bound xenobiotic; therefore, the total xenobiotic concentration may be in the therapeutic range even though the active unbound fraction is actually elevated. Phenytoin, which is highly bound to albumin, serves as an illustrative example. If the serum concentration of phenytoin is reported as subtherapeutic, the physician might order a dose increase even though the free fraction of phenytoin actually is in the therapeutic range. With a dose increase, the free or active fraction may increase to toxic concentrations. Basic xenobiotics are not bound to albumin but to α-1-acid glycoprotein, an acute-phase reactant that tends to increase, rather than decrease, with age.¹ However, the increase attributed to age is most likely related to underlying disease. These unpredictable changes would be expected to have the reverse effect on the ratio of bound to unbound xenobiotic in any laboratory report. However, the correlation between clinical effect and free xenobiotic concentrations requires further study because there may be complex factors involved—for example, alterations in Vd, tissue-specific xenobiotic concentrations, and the contribution of other serum proteins to xenobiotic binding.

The precise contributions of age-related changes in the gastric and intestinal tract to toxicity have not been adequately determined. Changes in gastric absorption are not substantial enough to have an important clinical impact, and if anything, there is a modest decline, partly because of delayed gastric emptying, which could cause a delay in absorption of certain xenobiotics. However, age-associated changes in the gastric mucosa may account for enzymatic changes, as demonstrated in the case of alcohol dehydrogenase, noted above.

Pharmacodynamic factors may also affect a patient’s response to a particular xenobiotic. In general, age-related physiologic changes in target or nontarget organs lead to increased sensitivity to most xenobiotics, although sensitivity to some xenobiotics may also be decreased. For example, there is evidence that the sensitivity of the β-adrenergic receptor declines with aging, leading to a diminished response to both β-adrenergic agonists and antagonists.^31,43,139 However, these experimental findings may not necessarily be demonstrated clinically.^5,6 Observed enhanced sensitivity to xenobiotics is probably due to altered pharmacokinetics in many, if not most, cases.⁵⁶ Proving that enhanced sensitivity is related to altered pharmacodynamics would require demonstrating that the concentration of drug at the tissue site was not increased as the result of diminished elimination. Regardless of the mechanism, it is important to recognize that the response to a given xenobiotic might be altered in specific ways among the elderly. These altered responses are probably caused less by chronologic aging and more by an increased prevalence of disease.^28,56Table 33–4 provides examples of pathologic or physiologic changes that frequently occur in the elderly and disorders that are worsened or unmasked by xenobiotics.

Although the elderly account for numerically far fewer poisoning exposures overall, the exposure rate per 100,000 population remains substantial. The contribution of ADEs and therapeutic errors may pose a particular problem for older patients and have become an important part of poisoning exposure analysis in recent years.

The incidence of ADEs among adults increases steadily with age.¹⁵ An ADE is defined as “one that occurs with normal, prescribed, labeled or recommended use of the drug.”¹³ ADEs are more likely to be serious in the elderly than in younger adults. Serious ADEs are those resulting in death, hospitalization, life-threatening outcome, disability, or other serious outcome²³ and are most prevalent among people 85 years of age and older.¹⁸

Likewise, exposures in persons older than age 60 years are often the result of therapeutic errors.^29,36,62 A therapeutic error is defined as “an unintentional deviation from a proper therapeutic regimen [of a medication or product used as a medication]⁶² that results in the wrong dose, incorrect route of administration, administration to the wrong person, or administration of the wrong substance;” therapeutic errors also include “unintentional administration of drugs or foods which are known to interact.”¹³ In the NEISS-CADES study, unintentional overdoses (defined as “excessive doses or supratherapeutic drug effects”) accounted for 65.7% of ADE-related hospitalizations in older adults.¹⁴ Risk factors for this age-related vulnerability include physiologic changes that affect xenobiotic disposition or effect, patient errors caused by cognitive or visual impairment, or lack of provider proficiency in geriatric prescribing principles.

In contrast to serious consequences of therapeutic errors or unintentional overdoses, serious reactions can occur with appropriate doses of certain medications. An important example is the neuroleptic malignant syndrome (NMS), which is potentially life threatening.²² Although one might predict it would be more common to occur in late life when abnormalities in thermoregulation are more common,¹² the risk of developing NMS has not been linked to advanced age per se.²² However, antipsychotic drugs are often prescribed to frail elderly, especially those with dementia, and NMS can be overlooked in a patient with comorbidities that could obscure a precise diagnosis. For example, NMS must be distinguished from such geriatric syndromes as cerebrovascular events, heat stroke, or neuroleptic sensitivity syndrome that occurs in Lewy body dementia.²²

Syndromes such as NMS are unpredictable and relatively rare at any age. Among elderly patients, serious or life-threatening ADEs are more typically caused by commonly prescribed drugs, such as anticoagulants, insulin, cardiovascular and insulin secretagogues and drugs that cause delirium, such as sedative–hypnotics, opioids, and anticholinergics (Table 33–1).

As many as 42% of serious ADEs reported in elderly patients are potentially preventable.^8,57,58 NEISS-CADES data reveal that drugs requiring careful outpatient monitoring are disproportionately represented among ED visits attributed to unintentional drug overdose.^14,16 Data from 2007 to 2009 indicate that warfarin, insulins, antiplatelet agents, and insulin secretagogues accounted for an estimated 67% of ADE-related hospitalizations in people 65 years of age and older, with digoxin and anticonvulsants contributing another 5.2%.¹⁶ The previously mentioned medications can be supervised with therapeutic as well as clinical monitoring. For certain other medications, ADEs are often best prevented by not prescribing them to vulnerable patients or by cautioning against their use if they are available without prescription. Criteria have been developed to guide clinicians regarding these “potentially inappropriate medications” (PIMs) in elderly patients.⁴ Xenobiotics deemed potentially inappropriate in the elderly have caused numerically fewer ADEs than more frequently prescribed xenobiotics,^14,16,100 but consensus as well as evidence exist that they are strongly associated with adverse outcomes in elderly patients, including hospitalization and mortality.^4,48,85 PIMs are also linked to enhanced risk of specific ADEs. For example, the elderly are at enhanced risk of severe upper gastrointestinal (GI) bleeding from NSAIDs.¹⁴⁵ Although this could be partly explained by increased exposure to and regular use of these drugs for chronic conditions, older adults may be more vulnerable to their effects because of underlying atrophic gastritis. The elderly are also at greater risk of prolonged hypoglycemia from certain sulfonylureas, particularly glyburide (glibenclamide).¹²⁰ Age-related mechanisms to explain this vulnerability to glyburide may include a decreased counterregulatory hormonal response to xenobiotic-induced hypoglycemia,⁹⁷ inappropriate stimulation of insulin if hypoglycemia is present,¹²⁹ or reduced excretion of a renally eliminated active metabolite of glyburide.⁷² Finally, as noted earlier, data gleaned from ED visits do not capture problems that occur and may or may not be managed in other settings. Other age-related factors involved in enhanced susceptibility to ADEs are given in Tables 33–2 and 33–4.

A complicated medication regimen reduces adherence,⁹⁸ increases errors,and increases the risk of clinically important xenobiotic interactions. Geri­atric patients take more prescription and nonprescription xenobiotics than any other patient group.⁷⁴ The likelihood of experiencing an ADE increases with the increasing number of xenobiotics taken.⁶⁵ The problem can be amplified by the common practice of pharmacies to fill an ongoing prescription of a particular medication with different brands over time; this diversity is expected but often problematic, and it is not necessarily accompanied by an adequate discussion with the patient or caregiver, or in the case of a mail order pharmacy, without highlighting the change. The new drug will differ in size, shape, or color from the medication previously dispensed, sometimes also appearing similar or identical to a different medication on the regimen, and medication errors can occur. The same problem can occur when a generic ver­sion of the prescribed medication replaces the drug with the “brand” name.

Concurrent disease in target or nontarget organs may alter the patient’s sensitivity to a xenobiotic, resulting in a serious ADE even when the patient is given a standard or previously used dose. Coexistent disease is often subclinical, and the patient’s enhanced sensitivity may not be anticipated. For example, a patient with subclinical Alzheimer disease whose cognitive function is overtly normal may acutely develop delirium or symptoms of dementia when given standard doses of drugs such as sedative–hypnotics or TCAs. Delirium is a medical emergency and an important cause of ED visits by the elderly.⁶⁷

Another factor contributing to ADEs is physician’s lack of knowledge with regard to the principles of geriatric prescribing.^41,60,105 In addition to prescribing inappropriate doses or increasing the dose too rapidly, clinicians may prescribe drugs considered potentially inappropriate for the elderly at any dose, such as TCAs, anticholinergics, and long-acting benzodiazepines.⁴ At a minimum, and particularly in the case of potentially inappropriate medications, it is reasonable to attend to specific risk factors outlined here and elsewhere in an effort to prevent ADEs.

Another problem is that clinicians may be unduly ready to prescribe drugs recently on the market, despite availability of acceptable and often less expensive alternatives. Recently approved medications are sometimes promoted as being safer than older ones, but problems often become apparent only after marketing and use by large numbers of patients. For example, the hypnotic agent zolpidem was marketed as a safe alternative to benzodiazepines for the elderly. However, like benzodiazepines, zolpidem may cause confusion, global amnesia, memory loss, and falls.^140,142 Low-molecular-weight heparins (LMWHs), such as enoxaparin and dalteparin, are other examples. LMWHs have more predictable pharmacokinetics than unfractionated heparin and are associated with a lower rate of overall bleeding. However, because therapeutic monitoring requires measurement of antifactor Xa activity, which is not as readily available as standard tests such as activated partial thromboplastin time (aPTT), monitoring is usually not performed.⁶⁴ LMWHs such as enoxaparin and dalteparin are eliminated by the kidneys, and repeated doses lead to progressive increases in antifactor Xa activity when GFR is 30 mL/min or below,^27,90 a degree of CKD that is common in frail elderly patients. Less severe levels of CKD may also result in reduced enoxaparin clearance that might be avoided with lowered doses.⁶⁶ Most reported cases of serious, unexpected enoxaparin-induced bleeding occur in elderly patients who are receiving “standard,” not age-appropriate, dosing.^96,135,137 Similar problems should be anticipated with other new anticoagulants. For example, the release of dabigatran, a renally eliminated direct thrombin inhibitor approved in 2010 for management of atrial fibrillation, was soon followed by many reports of severe GI bleeding and hemorrhagic stroke in the United States⁴⁴ and internationally.⁶³ The latter source specifically noted the preponderance of elderly patients in the reports. This finding is consistent with its use in atrial fibrillation, which is most prevalent among the elderly. Whether the risk of dabigatran is greater than that of warfarin is not known, because direct reports to the Food and Drug Administration (FDA) and others in the postmarketing period are not subject to comparison with other xenobiotics.¹²³ Likewise, it is premature to conclude that lack of monitoring or failure to adjust for occult CKD is a cause; however, the clinical trial supporting approval of dabigatran, which randomized subjects with mean age of 71 years, specifically excluded those with estimated GFR ≤ 30 mL/min.³² Importantly, anticoagulants in general remain among the most important causes of ADEs, therapeutic errors, and hospitalizations.^14,62 For elderly patients in particular, it is essential for clinicians to be mindful of the possible presence of occult CKD and lesions with the potential to bleed in order to reduce untoward bleeding in this high-risk group of patients.

It is not surprising that ADEs are first noted following drug approval in the postmarketing period when actual patients (as opposed to carefully selected research subjects) are exposed to the drug, because drugs and their effects are often inadequately studied in the elderly (Chap. 139).^17,86 Reactions occurring in a small percentage of patients in a special subgroup can easily be missed during the initial drug evaluations. Even when a substantial number of subjects older than age 60 years are studied, a much smaller proportion of patients older than age 70 years may be included in clinical trials.⁸⁶ Thus, the adults at highest risk for many forms of toxicity are in a population that is often the least studied. Xenobiotic testing is typically carried out in subjects who are young adults and disease free, so pharmacokinetic profiles do not reflect patterns of xenobiotic disposition characteristic of geriatric patients. Pharmacokinetic testing may be limited to a one-time dose, and frequently the evaluation takes place over a short period of time. On average, approximately five half-lives of a xenobiotic are necessary to achieve steady-state xenobiotic concentrations (Chap. 9). Thus, a xenobiotic with a half-life of 24 hours might not reach a steady state for 5 days, and in the presence of prolonged elimination associated with age-related factors, a steady state might not be reached for substantially longer. As a result, even if elderly subjects are included in a drug trial, the ultimate effect might not be appreciated during testing intervals designed for younger patients.

Another factor is the nature of pharmaceutical research itself. Morbidity and mortality in elderly patients as a result of specific drugs might be avoided if the responsible drugs were studied under the predictably high-risk conditions typically present in the elderly. For example, xenobiotics eliminated by the kidney need to be evaluated after repeated dosing in elderly subjects. Gatifloxacin was removed from the market only after several years of use when it was finally linked to both hypoglycemia and hyperglycemia in large numbers of elderly subjects.^55,106 Adverse events from xenobiotics that are studied in the elderly or others with chronic illness may be less obvious in the presence of comorbidity in the population at risk. The cyclooxygenase-2 (COX-2) inhibitor rofecoxib (Vioxx) was withdrawn from the market in 2004 after it was shown to increase the risk of myocardial infarction and stroke, especially in older adults.³⁹ Based on the complex actions of COX-1 and COX-2 in many tissues, the possibility existed that COX-2 inhibition might increase cardiovascular morbidity, for example, by leaving COX-1–mediated platelet aggregation unopposed while inhibiting prostacyclin-induced vasodilation or by leading to fluid retention and increased blood pressure via renal mechanisms.^61,109 Because the elderly typically have one or more chronic illnesses²⁸ as well as occult disease, extra vigilance is required when new xenobiotics are given, and clinicians and clinical investigators must be very mindful of the theoretical possibilities of adverse outcomes. In view of the vulnerability of older patients to many medications, the FDA now requires that sponsors of new drug applications present effectiveness and safety data for important demographic subgroups, including the elderly, in their FDA submission data.²⁴ However, exceptions to the rule are allowed. For example, when studies have included insufficient numbers of subjects older than age 65 years to determine whether the elderly respond differently to the drug, the labeling must state this, but the statement is a poor substitute for providing actual efficacy and safety data.³⁰ Unfortunately, geriatric recommendations in the package insert may thus be insufficiently specific to provide guidance for drugs commonly used in this population.¹²⁶

Drugs, such as digoxin, warfarin, and diuretics, commonly prescribed in the elderly, are frequently involved in serious drug interactions. This situation is complicated by the frequency with which elderly patients, who often have multisystem disease, visit multiple physicians, who prescribe medications without specific knowledge of, or attention to, the remainder of the patient’s drug regimen, thereby increasing the risk of inappropriate combinations.^52,132 Problems can also arise when patients obtain prescriptions from more than one pharmacy or mail order service. Warfarin is a particular problem, owing to its narrow therapeutic index and numerous pharmacokinetic as well as pharmacodynamic interactions, the potential for which (eg, with certain antibiotics) may often be ignored at the point of care.⁵⁰

Herbal preparations used by the elderly also may interact with prescription medications.⁷⁰ The use of herbal preparations has increased substantially in recent years, particularly in patients with illnesses such as cancer, dementia, and depression, which commonly affect the elderly. Very few patients voluntarily report use of these or other nonprescribed therapies to their physicians, and too often the physician fails to inquire specifically about such “alternative” or “complementary” therapies. Poisonings, herb–drug interactions, and other problems related to herbal preparations are discussed further in Chap. 45.

The use of nonprescription pharmaceuticals may also cause serious adverse effects. For example, excessive use of magnesium-containing preparations frequently causes severe toxicity in older individuals. Impaired GFR, decreased GI motility, and other medical comorbidities are just three risk factors that potentiate magnesium toxicity in the elderly. The source of ­magnesium in these cases may include the cathartics that contain magnesium hydroxide (Milk of Magnesia) and magnesium citrate, antacid preparations, and magnesium sulfate (Epsom salts).⁴⁹ Magnesium-containing laxatives are still sometimes used in hospital settings with serious outcomes.¹⁰⁴ Likewise, sodium phosphate formulations may harm ­kidney function in certain elderly patients despite normal baseline creatinine.⁷⁶ Virtually all of the most popular nonprescription medications⁷⁴ are more likely to produce problems in the elderly than in younger patients, including GI bleeding (aspirin and other NSAIDs), enhanced warfarin sensitivity (cimetidine), confusion and urinary retention (anticholinergic antihistamines), and cardiovascular symptoms (pseudoephedrine).

Outdated and discontinued xenobiotics are an additional problem for the elderly, who often retain unused or partially used products in their homes for decades or may continue to request prescription renewals when safer or more effective alternatives are available. Patients may be unwilling to change or physicians may continue to renew the prescription without sufficiently reevaluating the patient. Potential examples include digoxin and theophylline, which have been responsible for large numbers of adverse events diagnosed in EDs,^14,15 as well as sedating antihistamines, other anticholinergics, and diazepam.

Other age-related factors may increase the risk of unintentional poisonings in geriatric patients; impaired vision, hearing, and memory may lead to misunderstanding or an inability to follow directions concerning the use of prescription and nonprescription drugs. Dementia is another important risk factor in unintentional poisonings. In addition to cognitive impairment, patients with dementia sometimes exhibit abnormal feeding behaviors, which may lead to ingestion of toxic xenobiotics.¹⁴³

Management decisions must be made with the foregoing principles in mind. GI decontamination should proceed as in younger patients. However, because constipation is a more frequent problem in the elderly, when multiple-dose activated charcoal is used, particular attention must be paid to GI function and motility. The specific precautions and contraindications in the basic management and GI decontamination detailed in Chap. 8 are particularly pertinent for the geriatric population.

The presence of clinical or preclinical congestive heart failure or CKD may increase the risk of fluid overload when sodium bicarbonate is used. In the elderly, extracorporeal removal may be indicated earlier in cases of salicylate, lithium, or metformin poisoning, in which elimination may be hampered by a decreased GFR.

A situation that may go unrecognized in geriatric patients is the problem of alcohol or other substance withdrawal syndromes. Because elderly patients are typically not perceived as substance abusers, or because a ­physician evaluating an unfamiliar patient may not be aware of the patient’s chronic use of prescribed benzodiazepines or opioids, the possibility of ­substance withdrawal may not be considered when unanticipated complications occur during hospitalization.

Strategies to limit unintentional toxic exposures in elderly patients with cognitive or sensory impairment should be similar to those used in young children, who are also at high risk of ingesting toxic substances or pharmaceuticals prescribed for others in the household. The strategies should include the removal of potentially dangerous and unnecessary xenobiotics from the patient’s environment. The physician should request that the patient or caregiver bring all medications to the office in the original bottles. The physician should then make an effort to limit the number of medications prescribed or to seek alternative medications with a safer therapeutic index or appropriate geriatric safety profile. Administration and control of the medications by directly observed therapy may, of necessity, become the responsibility of the caregiver rather than the patient.

When geriatric patients are evaluated in the ED for poisonings or serious ADEs, the need for hospital admission should be weighed carefully against the known hazards of hospitalization for the elderly.³⁵ The physician should be particularly alert to certain situations that might mandate admission, including unexplained mental status changes, overdose of a prescribed medication with a prolonged duration of action, or evidence of inadequate home care or elder abuse/neglect, such as poor hygiene or unexplained injury.

When there is concern that the established caregivers at home may be abusing or neglecting the patient, the patient requires further observation, removal from the environment, and possibly hospitalization. Signs of actual physical abuse may be more obvious than signs of neglect.⁸² Vulnerable elderly who are physically disabled or cognitively impaired may be brought to the hospital because of presumed illness, but the source of the problem may actually be the caregiver. The caregiver, frequently but not necessarily a family member, may be depleting the patient’s funds for personal use. Patients may become ill because funds were diverted from the purchase of food or because the patient’s prescription drugs were sold on the street. More direct abuse may take the form of intentional poisoning or oversedation of the patient by overdose of the patient’s own prescription drugs, or selling the patient’s medications (diversion), resulting in clinical deterioration. Provision for follow-up care, such as an alternative place to live and guardian appointment, is essential because abuse and neglect are associated with serious outcomes.⁸³

Unresolved mental status changes may require close observation and hospitalization. Elderly patients who are confused or unable to walk are sometimes mistakenly assumed to be chronically impaired. However, incomplete explanation of an altered mental status or physical impairment should prompt careful inquiry into the patient’s baseline functional status. Poor function should never be assumed to be “normal aging.” Many very elderly patients are cognitively normal, physically robust, and independent in all activities of daily living.

A patient whose problem has been attributed to overdose with or accumulation of a long-acting medication requires careful monitoring. Because the durations of action may be markedly prolonged among geriatric patients, a higher degree of vigilance is required. An important example is chlordiazepoxide, which is commonly used for alcohol withdrawal and exhibits very long half-life in older patients, especially after repeated dosing. The ultra–long-acting sulfonylurea chlorpropamide is rarely used today, but glyburide may cause relatively prolonged hypoglycemia compared with other insulin secretagogues, as noted above.

• Older patients may account for only a small fraction of total poisoning victims, but when poisoned, they have a high mortality rate.

• The elderly are much more likely to experience serious ADEs as a consequence of both appropriate and inappropriate use of medications.

• Attention to risk factors is essential in this vulnerable population.

• Important risk factors include pharmacokinetic and pharmacodynamic changes; the presence of overt or subclinical disease, including dementia; patient and provider error; suicide risk; complex therapeutic drug regimens; and a general lack of knowledge about the principles of prescribing for the geriatric population.

References