Reducing Adverse Drug Events

Reducing Adverse Drug Events 20

DeAnne Zwicker and Terry Fulmer

EDUCATIONAL OBJECTIVES

On completion of this chapter, the reader should be able to:

1. Identify persons at high risk of adverse drug events (ADEs)

2. Conduct a comprehensive medication assessment

3. Specify four medications or medication classes having a high potential for toxicity in older adults

4. Describe five reasons that older adults experience adverse drug events

5. Delineate strategies to prevent common medication-related problems in older adults

OVERVIEW

One in seven Medicare beneficiaries experienced an adverse event while being hospitalized in 2008. Between 2007 and 2009, almost 100,000 emergency hospitalizations in the United States were caused by ADEs in adults 65 years of age or older. That is, nearly half of those were aged 80 years and older (Budnitz, Lovegrove, Shehab, & Richards, 2011). This number is likely much higher as ADEs often go unreported with only 50% being reported in U.S. hospitals. ADEs increase the costs to patients, length of hospital stay, and mortality, yet often go unreported and nearly half are likely preventable (Levinson, 2010; Salvi et al., 2012). Older adults are the largest consumers of prescription and over-the-counter (OTC) medications and the most at risk of ADEs (Hamilton, Gallagher, Ryan, Byrne, & Mahony, 2011; Qato et al., 2008; Sansgiry, Nadkarni, & Doan, 2011).

BACKGROUND AND STATEMENT OF PROBLEM

Adverse Drug Event

An ADE is “an untoward medical occurrence that may appear during treatment with a pharmaceutical product but which does not necessarily have a causal relationship with the treatment” (World Health Organization [WHO], 2002, p. 5). An ADE may occur during normal use of medication, inappropriate use, inappropriate or suboptimum prescribing, poor adherence, self-medication, or harm resuting from a medication error. An adverse drug reaction (ADR) is harm directly caused by a drug at normal doses (Nebeker et al., 2004; Table 20.1). Most ADRs are not ADEs. For example, if a patient experiences an expected ADR, as in the case of hypokalemia secondary to furosemide therapy, this would not be an ADE unless it is not identified as a clinically significant event. Likewise, medication errors are common but most do not result in actual patient harm. The use of prescription drugs has increased significantly over the past 25 years with a 90% increase and a consequent increase in ADEs reported to the Food and Drug Administration (FDA; National Center for Health Statistics [NCHS], 2012). Between 2004 and 2008, there was a 52% increase in the number of ADEs reported in U.S. hospitals (Lucado, Paez, & Elixhauser, 2008). Older adults are the greatest consumers of medications with on average, older Americans taking three OTC medications and up to six or more prescription medications (Petrovic, van der Cammen, & Onder, 2012). Among older adults, up to 30% of all hospital admissions are secondary to an ADE—more than half could have been prevented (Figueiras, Herdeiro, Polónia, & Gestal-Otero, 2006; Steinman & Hanlon, 2010). Intrinsic factors, such as advanced age, frailty, and polypharmacy, place older adults at greater risk of adverse outcomes (Buck et al., 2009). Older adults are at high risk of further ADEs while in the hospital and often experience ADEs after discharge (Kanaan et al., 2013; Lucado et al., 2008; Salvi et al., 2012).

TABLE 20.1

Summary of Definitions Relevant to Drug-Related Harm

Term	Definition	Example
Harm occurred
Adverse event	Harm in a patient administered a drug but not necessarily caused by a drug	Traumatic death while taking lovastatin
Adverse drug reaction	Harm directly caused by a drug at normal doses Unexpected adverse drug reaction: An adverse drug event whose nature or severity is not consistent with the product information	Congestive heart failure from metoprolol
Adverse drug event	Harm caused by the use of drug Effective definition in common practice: Harm caused by a drug or the inappropriate use of a drug	Hematoma from tirofiban overdose
Harm may have occurred
Medication error	Inappropriate use of a drug that may or may not result in harm	Failure to renew prednisone order on transfer to medical ward
Side effect	A usually predictable or dose-dependent effect of a drug that is not the principal effect for which the drug was chosen; the side effect may be desirable, undesirable, or inconsequential	(This term should be avoided when considering adverse events.)
Harm did not occur
Potential adverse drug event	Circumstances that could result in harm by the use of a drug but did not harm the patient	Receipt of roommate’s felodipine but no resulting hypotension

Source: Courtesy of John W. Devlin, PharmaD, Northeastern University (personal communication, March 2015).

Persons older than 65 years experience medication-related events for five major reasons: (a) alteration in pharmacokinetics; a reduced ability to metabolize and excrete medications and pharmacodynamics, what the drug does in binding at the receptors (Ruscin & Linnebur, 2014; Steinman & Holmes, 2014); (b) polypharmacy, indicating multiple meds prescribed, which are often prescribed by multiple providers (Rogers et al., 2009); (c) therapeutic failures, for example, over or under therapeutic dosage (Marcum, Driessen, Thorpe, Gellad, & Donohue, 2014); (d) iatrogenic causes such as ADRs or inappropriate prescribing (Davies et al., 2009); medication errors (Hayes, Klein-Schwartz, & Gonzales, 2009); and using medication for treatment of symptoms that are not disease dependent or specific such as self-medication or prescribing cascades (Rochon, 2014; Rochon & Gurwitz, 1997); and (e) medication adherence (Steinman & Hanlon, 2010). Most ADRs are not ADEs. If a patient experiences an expected ADR (e.g., hypokalemia secondary to furosemide therapy), based on package insert information, this would not be an ADE unless it was not identified and a clinically significant event (e.g., cardiac arrest) occurred. Also, although medication errors are common, most do not result in actual patient harm.

Common OTC medications, such as aspirin and acetaminophen, often interact with prescription medications (Qato et al., 2008). Older adults often combine OTCs with prescription medications or herbal remedies yet do not report their use to health care providers. Likewise, providers often do not inquire about OTCs or herbal remedies. Underreporting may lead to ADEs, including unrecognized adverse drug–disease or drug–drug interactions (Lopez-Gonzalez, Herdeiro, & Figueiras, 2009). Patient safety in medication management is a major focus in older adults; however, current information on concurrent use of prescription medications, OTC medications, and dietary supplements is limited (Qato et al., 2008).

Iatrogenic Causes of ADE

The term iatrogenic may be defined as an ADR or complication induced by a nondrug medical intervention (Atiqi, van Bommel, Cleophas, & Zwinderman, 2010). An iatrogenic ADE is disease induced by a drug prescribed by a medical provider (WHO, 2002). An iatrogenic medication event is one that is preventable, such as the wrong dose of a medication given that resulted in an adverse outcome. ADRs, inappropriate prescribing of high-risk medications, and medication errors are all considered iatrogenic. An ADR is any noxious or unintended and undesired effect of a drug that occurs at normal human doses for prophylaxis, diagnosis, or therapy. Inappropriate prescribing of high-risk medication to older adults and medication errors are all considered iatrogenic (WHO, 1972).

Risk factors of iatrogenic complications in older adults include drug-induced iatrogenic disease, multiple medications and conditions, multiple physicians, hospitalizations, and medical or surgical procedures (Davies et al., 2009). Onder, van der Cammen, Petrovic, Somers, and Rajkumar (2013) developed and evaluated an ADR risk score based on prior evidence. The researchers reported that the number of drugs and history of a previous ADR were the strongest predictors of ADRs, followed by heart failure, liver disease, the presence of greater than or equal to four conditions, and renal failure. They determined that a GerontoNet ADR risk score of four out of 10 showed a good balance between specificity and sensitivity. In a sample of 513 acutely ill patients aged greater than or equal to 65 years, the GerontoNet ADR risk score missed almost 40% of those at risk of ADRs, although more research is needed to identify further risk factors (Onder et al., 2013). So far, this is the only tool available to determine potential ADR risk. Frail older adults with multiple medical problems, memory issues, and multiple prescribed and nonprescribed (OTC and herbal) medications are at highest risk of ADRs (Rochon, 2014).

Adverse Drug Reactions

It is important to differentiate an ADR from an ADE (Figure 20.1). An ADE may be an event caused by a number of different adverse outcomes and may be related to normal or inappropriate use of medications, suboptimum prescribing, poor adherence, self-medication, or harm resulting from a medication error (WHO, 2002). The WHO defines an ADR as “a response to a drug which is noxious and unintended, and which occurs at doses normally used in man for the prophylaxis, diagnosis, or therapy of disease, or for the modifications of physiological function” (p. 5). Thus, an ADR is attributable only to a drug, whereas an ADE includes many different types of adverse outcomes rather than just a drug reaction.

FIGURE 20.1

Relationship between ADEs and ADRs.

ADEs, adverse drug events; ADRs, adverse drug reactions.

Source: Nebeker et al. (2004).

ADRs are the result of some action of a drug and may be classified as dose related (digoxin > 0.125 mg), nondose related (hypersensitivity), dose related and time related (cumulative dose), time related, unexpected therapy failure, or the result of drug withdrawal (Salvi et al., 2012). In a systematic review, Kongkaew, Noyce, and Ashcroft (2008) reported a 10.7% prevalence rate of hospital admissions caused by ADRs in older adults; however, a confounding factor in the accuracy is that different methods and studies were employed to gather the data (Alomar, 2014). ADRs may occur because of drug interactions, duplication, additive effects, discontinuation of therapy, skipping medication, changing dose to costs, and physiologic antagonism (Alomar, 2014). Polypharmacy can also make it difficult for older adults to keep track of their multiple medications, particularly with added instructions such as taking meds with food, after meals, or at bedtime. Chronic side effects, such as fatigue, constipation, rashes, falls, and anxiety, can also lead to skipping medications. Finally, many process-related issues, such as lack of thorough medical reconciliation at each step of transition through the health care system, lead to error.

Frail older adults with multiple chronic medical problems requiring multiple medications are at high risk for ADRs (Hubbard, O’Mahony, & Woodhouse, 2013). Drug–drug and drug–disease interactions are the most common ADRs. Causes of drug interactions are multifaceted and include drug dosage, serum drug level, administration route, drug metabolism, therapy durations, and patient factors (Heuberger, 2012). Gray and Gardner (2009) reported that polypharmacy and multiple prescribers tend to be key factors in adverse reactions. Debate continues as to whether advancing age causes an increased ADR risk or if it is a marker for comorbid illness, change in pharmacokinetics, and polypharmacy. Many studies have shown that the key risk factors of adverse reactions are related to polypharmacy, multiple providers, and inappropriate use of medications (Gray & Gardner, 2009).

In hospitals, medications that are commonly associated with ADRs include diuretics, antihypertensives, anticoagulants, and antineoplastics. Drug–drug interactions occur when one therapeutic agent either alters the concentration (pharmacokinetic interactions) or the biological effect of another agent, a pharmacodynamic interaction. An example of drug–drug interaction is bleeding as a result of a combination of warfarin and nonsteroidal anti-inflammatory drugs (NSAIDs) or warfarin and aspirin. Drug–disease interactions occur when the administration of a medication alters a preexisting disease state. For example, the administration of aspirin to a patient with peptic ulcer disease will increase the risk for gastrointestinal (GI) bleeding (Salvi et al., 2012).

More than 80% of ADRs that occur in the hospital or cause admission are dose related, and therefore are predictable and potentially avoidable (Atiqi et al., 2010). Marcum et al. (2012) reported that inappropriate medications and therapeutic failures (a failure to accomplish treatment goals because of inadequate or inappropriate drug therapy) as well as adverse drug-withdrawal events (symptoms or signs related to removal of a drug) have not been studied as much as ADRs and often are a cause of hospitalizations (Z. A. Marcum et al., 2012).

Medication Errors

There is no international definition for medication errors; however, the most commonly used definition by the National Coordinating Council for Medication Error Reporting and Prevention (2015, p. 4) is “any preventable event that may cause or lead to inappropriate medication use or patient harm while the medication is in the control of the health care professional, patient, or consumer. Such events may be related to professional practice, health care products, procedures, and systems, including prescribing, order communication, product labeling, packaging, and nomenclature, compounding, dispensing, distribution, administration, education, monitoring, and use.”

A large percentage of errors is caused by administration of the wrong medication or the correct medication with the wrong dose or at the wrong time interval between dosing. A medication error may be the result of negligence or prescribing errors. A reduction of these types of prescribing errors has been a high priority for health care policy in order to improve the safety profile of the health care delivery system (Aspden, Institute of Medicine [IOM], & Committee on Identifying and Preventing Medication, 2007).

In a retrospective review of 140,786 older adults, 49,320 experienced therapeutic errors resulting in adverse medical outcomes (Hayes et al., 2009). One example of therapeutic error may be taking twice the dose of prescribed medication, for example, which may or may not result in an adverse outcome. The majority of patients reporting ADEs (82%) were in the home environment or other non–health care facility (HCF). In those with a known adverse outcome, no adverse effect occurred in 62.9%, minor effects occurred in 25.2%, moderate effects in 10.7%, major effects occurred in 1.0%, and death in 0.2%. Older adults aged 65 to 74 years constituted 45% of major effects, 75 to 84 years 40%, and 85 years and older 14.5% of the majority of cases. Serious outcomes occurred more frequently in the older age groups (Hayes et al., 2009).

Medication Adherence

Costs of nonadherence are estimated at $100 billion annually (Ho, Bryson, & Rumsfeld, 2009). Medication adherence (or compliance) with a medication regimen is defined as “the extent to which a person’s medication-taking behavior corresponds with agreed recommendations of a health care provider” (WHO, 2003, p. 3). A study by Marcumet al. (2014) revealed that older adults were nonadherent to treatment with chronic medications, such as 26% not taking calcium channel blockers. Those who used multiple pharmacies were more likely to be nonadherent to medications. There was no difference when demographics, health status, or access to care were considered. In a study providing elimination of copays for medications prescribed after a myocardial infarction, those with the usual coverage had an adherence rate from 35% to 49%; the full-coverage group was four to six points higher (Choudhry et al., 2011).

In a study of 7,108 patients enrolled in a pharmacy assistance program evaluating medication adherence, 40% to 50% adhered to their chronic medications with 80% of those patients reporting no income (Roberts et al., 2014). Older age and adherence to antihypertensive and statins (alone and combined) were significant predictors of adherence to medications. In patient admissions were much higher for the nonadherence population. Those who spoke English and those who took antihypertensives and oral hypoglycemic medication also had lower admission rates (Roberts et al., 2014).

Gellad, Grenard, and Marcum (2011) found the following barriers to medication adherence: health literacy, lack of knowledge of chronic disease, cognitive function, and adverse effects of drugs and polypharmacy; although they report there is a paucity of evidence on medication adherence in the literature and a need for using standardized measurements in research. Fried, Tinetti, Towle, Leary, and Iannone (2011) revealed that the presence of ADEs influenced the patient’s decision whether to take a medication or not. Patients are often reluctant to admit nonadherence; however, pill counts and refill history can aid in determining this issue (Gearing, Townsend, Elkins, El-Bassel, & Osterberg, 2014; Steinman & Hanlon, 2010). Acute care nurses are ideally positioned to educate older adults and aid in evaluating reasons for nonadherence in hospitalized older adults.

EVALUATION OF THE PROBLEM

Assessment Tools

Geriatric assessment tools are used to evaluate an older adult’s ability to self-administer medications, including physical and cognitive functional capacity (see Chapter 7, “Assessment of Physical Function” and Chapter 14, “Preventing Functional Decline in the Acute Care Setting;” review of the medication list for potentially inappropriate medications (PIMs; Campanelli, 2012), medications that are indicated but underused; medication reconciliation (MR), evaluation for potential drug–drug or drug–disease interactions; assessment of renal function; and encouraging the brown-bag method to augment reconciliation. Each of these methods should be patient/family centered and include collaboration with interprofessional team members. Common assessment tools include the following:

Beers Criteria: American Geriatrics Society Beers Criteria for Potentially Inappropriate Medication Use in Older Adults (Campanelli, 2012)—This is used to assess the medication list for medications that should generally be avoided in older adults.

STOPP Criteria: Screening Tool of Older Persons’ Prescriptions—This tool identifies potentially inappropriate medications; identifies many more than the Beers Criteria and has identified greater number of older adults requiring hospitalization because of ADEs (Gallagher, O’Connor, & O’Mahony, 2011).

START Criteria: Screening Tools to Alert Doctors to the Right Treatment—This is an evidence-based screening tool used to detect potential prescription omissions; recommends important medications for specific chronic conditions that are often omitted. This is intended to be used simultaneously with STOPP Criteria (Gallagher et al., 2011).

Drug–Drug Interactions: Table 20.2 provides a list of some common medications known to interact with other medications. This is done using a computer order-entry program (COE) and computer decision support that includes drug–drug interaction alerts (Clyne, Bradley, Hughes, Fahey, & Lapane, 2012).

Cockroft–Gault Formula—This is useful for estimating creatinine clearance based on age, weight, and serum creatinine levels (Terrell, Heard, & Miller, 2006). A creatinine clearance of less than 50 mL/min places older adults at risk of ADEs (Fouts, Hanlon, & Pieper, 1997) and virtually all people older than 70 years have a creatinine clearance of less than 50. An important point to note is that for older adults, particularly those of lower weight, this formula may estimate a creatinine clearance that is far higher than the actual glomerular filtration rate (GFR; Table 20.3).

Brown-Bag Method (Nathan, Goodyer, Lovejoy, & Rashid, 1999)—This method is used to assess all medications an older adult has at home, including prescriptions from all providers, OTCs, and herbal remedies. All medications at home are placed in a bag and brought to hospital/other care setting. This should be used in conjunction with a complete medication history.

Drugs Regimen Unassisted Grading Scale (DRUGS) Tool—This is a standardized method for assessing potential medication adherence problems. This requires a higher level of patient functioning, used at transfer to other levels of care (Edelberg, Shallenberger, & Wei, 1999; Hutchison, Jones, West, & Wei, 2006).

InterMed-Rx computer software—This is new software that identifies cytochrome p-450 drug interactions. The cytochrome p-450 enzyme system is where drug–drug interactions often occur. It was evaluated in 100 patients with polypharmacy older than 82 years with a mean of 12 drugs prescribed. The software identified at least one drug–drug interaction and enabled the patients to have immediate pharmacist intervention, and 56% required further follow-up and medication adjustment. This software may aid in early identification of persons at risk of ADRs (Zakrzewski-Jakubiak et al., 2011).

TABLE 20.2

Common Drug–Drug Interactions

TABLE 20.3

Cockroft–Gault Formula for Estimation of Creatinine Clearance (CrCl)

ASSESSMENT STRATEGIES

Changes With Aging in Pharmacotherapy

Nurses and team members must first be aware of aging changes in pharmacokinetics and pharmacodynamics when assessing medications in older adults (Steinman & Holmes, 2014). New research is directed at pharmacogenetics and pharmacogenomics and has some application to ADRs. Interindividual variations of these genes may account for the differences observed in drug efficacy and the appearance of ADRs in elderly people (Cardelli, Marchegiani, Corsonello, Lattanzio, & Provinciali, 2012). The following describes aging changes.

Pharmacokinetics is best defined as the time course of absorption, distribution across compartments, metabolism, and excretion of drugs in the body. As the body ages, the metabolism and excretion of many drugs declines and physiological changes require dosage adjustment for some drugs. The vast majority of drugs are cleared either through liver clearance or renal clearance (Steinman & Holmes, 2014).

Changes in drug absorption (e.g., increased gastric pH and decreased GI motility in an absorptive surface) or underlying disease states such as diabetes. Older adult patients taking acid-suppressive agents, such as a proton pump inhibitors (e.g., Prilosec or Pepcid) or digoxin, may have decreased absorption (Steinman & Holmes, 2014).

Drug distribution changes include decreased cardiac output, reduced total body water, decreased serum albumin (which is more likely to be related to malnutrition or acute illness than aging), and increased body fat. Changes in tissue or plasma binding can change the apparent volume of distribution (Vd) determined from plasma concentration measurements. Older people have a relative decrease in skeletal muscle mass and tend to have a smaller Vd, such as Vd of digoxin, which binds to muscle proteins. Reduced total body water creates a potential for higher serum drug levels because of a low Vd and occurs with water-soluble drugs (hydrophilic) such as alcohol or lithium. Decreased serum albumin results in higher unbound drug levels with protein-bound drugs, such as warfarin, phenytoin, digoxin, and theophylline. Drugs that distribute in fat (e.g., diazepam) may thus have a larger Vd. These lipophilic drugs (e.g., long-acting benzodiazepines [BZDs]) are stored in the body fat and slowly leech out, resulting in increased half-life and the drug staying around longer (Salvi et al., 2012; Steinman & Holmes, 2014).

A significant change in drug metabolism is a reduction in the cytochrome p-450 system, which affects metabolism of many drugs cleared by this enzyme system. Drug metabolism includes the processes of absorption, distribution, metabolism, and elimination. These processes may change with aging; however, they are typically more influenced by genetic factors and by an individual’s diseases, environment, and other medications (Steinman & Holmes, 2014). Many classes of drugs are cleared by the cytochrome p-450 enzyme system, including cardiovascular drugs, analgesics, NSAIDs, antibiotics, diuretics, psychoactive drugs, and others. Drugs, such as beta blockers, that go through the first-pass effect in the liver may be effective in lower doses in older adults (Steinman & Holmes, 2014). Drug metabolism may also be affected by alcohol abuse. Alcohol is metabolized and cleared slower in older adults thus stays around longer increasing the ADE risk. Metabolism may be affected by disease states common in older individuals (e.g., thyroid disease, congestive heart failure [CHF], and cancer) or drug-induced metabolic changes. Several drugs are cleared by multistage hepatic metabolism, which is more likely to be prolonged in older persons. Some drugs undergo hepatic metabolism and then renal clearance. BZDs have enormously longer half-lives in older adults because both systems are impaired, particularly with diazepam, clonezapam, and temazepam.

Elimination or renal clearance of medications from the body may be slowed because of decline in GFR, renal tubular secretion, and renal blood flow that naturally decreases with age (Steinman & Holmes, 2014). A decrease in clearance prolongs drug half-life and leads to increased plasma concentrations. A decrease in glomerular filtration is usually not accompanied by an increase in serum creatinine because of decreasing lean muscle mass with aging and a subsequent decline in creatinine production. Lack of dosage adjustment for renal insufficiency is a common reason for ADEs. Changes in renal function have the greatest impact on pharmacokinetics (Steinman & Holmes, 2014). Therefore, serum creatinine is not an accurate measure of renal function in the older adult. Instead, assessment of renal function using the Cockroft–Gault formula (see Table 20.3) should be made before initiation of renal clearing medications. Another instrument is the Modification of Diet in Renal Disease (MDRD) formula. This may be a more accurate reflection of renal function compared with the Cockroft–Gault equation; however, both are an estimate (Steinman & Holmes, 2014).

Pharmacodynamic problems occur when two drugs act at the same or interrelated receptor sites, resulting in additive, synergistic (toxicity is greater than the sum of either agent used alone), or antagonistic effects (opposite effects). Many interactions of drugs are multifactorial, with a sequence of events that are both pharmacokinetic and pharmacodynamics (see Figure 20.2) (Steinman & Holmes, 2014).

Beers Criteria

The most recent Beers Criteria (Campanelli, 2012) were developed by experts in geriatrics and pharmacotherapy with a higher level of rigor using the IOM standards. The rigorous review includes the rating of quality and strength of the evidence. Medications included in this list should be avoided in older adults because of toxic effects; for example, avoidance of anticholinergic drugs or avoiding digoxin dosing greater than 0.125 mg. Other recommendations for specific problems, such as reducing doses for patients with a decline in kidney function or the potential of adverse effects from specific high-risk medications (BZD associated with falls). Avoiding PIMs is a very simple method to reduce inappropriate medications and thus ADEs in older adults.

Education and quality measures have been developed to inform and encourage all interprofessional providers to focus on PIMs to reduce the number of medications (Pugh et al., 2013). Although these criteria have been recommended as a standard by many regulatory and quality-improvement bodies for many years, PIMs continue to be prescribed in older adults. All members of the interprofessional team can aid in ensuring that PIMs are avoided.

FIGURE 20.2

Multifactorial interactions leading to adverse drug events.

ADE, adverse drug event; BZDs, benxodiazepines; ICU, intensive care unit; NSAIDs, nonsteroidal anti-inflammatory drugs; OTC, over the counter.

Courtesy of John W. Devlin, PharmaD, Northeastern University (personal communication, March 2015).

STOPP and START Criteria

Like the Beers Criteria, the STOPP criteria are intended to identify potentially inappropriate prescriptions (PIPs vs. PIMs). These criteria include several more medications than the Beers Criteria. The prevalence of inappropriate medications in hospitalized older adults using these criteria was determined to be 77% (Lang et al., 2012). The START criteria identify omissions of medications that should be prescribed in hospitalized adults. According to the START criteria, 41.9% to 66% of hospitalized older adults experienced either undertreatment or omission of an appropriate medication (Lang et al., 2012). In a recent randomized controlled trial (RCT), the STOPP and START criteria were employed to identify PIPs and potentially omitted prescriptions (POPs) as determined by a pharmacist who made recommendations based on the criteria. Researchers reported a significant decrease in PIPs and POPs in the intervention group but not in the control group at the 6- to 12-month follow-up. The intervention group also showed a decline in falls (Frankenthal, Lerman, & Kalendaryev, 2014). This study indicates that pharmacist–physician collaboration in using the STOPP and START criteria in older adults may aid in decreasing ADEs.

Adverse Drug Reactions

ADRs are defined as any drug response that is noxious, unintended, and occurs at doses normally used for prophylaxis, diagnosis, or therapy of a disease (WHO, 2002). ADRs in older adults often go unidentified because of their atypical presentation in older adults or because the conventional symptoms of disease seen in younger persons are not evident (Steinman & Holmes, 2014) such as no pain with an acute abdomen. These changes are often a prodrome (symptom(s) indicative of an approaching disease) of an acute illness, especially in frail older adults. These symptoms may include a change in functional capacity, falls, vague cognitive changes, changes in behavior, decrease in appetite or anorexia, or new-onset incontinence (Steinman & Holmes, 2014). It is essential to take patients’, families’, and nonprofessional care providers’ reports of subtle symptoms seriously so that they are not missed. Timely identification of acute illness with vague presentation enables early treatment of illness resulting in reduced morbidity and mortality and an enhanced quality of life in older adults. Atypical presentation makes ADRs difficult to accurately identify, thus nurses must be aware that ADRs may present atypically. One strategy for improving recognition of ADRs is to identify risk factors associated with ADRs. The most common risk factors include the following: increasing age (85 years or older), multiple medical conditions, polypharmacy or multiple medications, and physical or cognitive or functional impairment (Steinman & Holmes, 2014; Table 20.4).

Polypharmacy is common in older adults with multiple chronic conditions and multiple prescribed medications. The number of medications prescribed to older adults can be up to three OTC medications and up to six or more prescription medications (Petrovic et al., 2012). Research has shown that polypharmacy positively correlates with an increased risk for ADRs, such as drug–drug and drug–disease interactions, and increased risk for nonadherence to taking prescribed medications. Therefore, providers should perform regular reviews for unnecessary medications and consider whether any medication may be discontinued.

When a new symptom develops in a person taking multiple medications, it must always be considered to be an ADE first, particularly in the frail hospitalized patient. Providers should evaluate the relationship between the new symptom and the most recent medication given to identify the potential cause. Inappropriate treatment of a new symptom may result in a prescribing cascade of multiple drug treatment (see Figure 20.3) and possibly an ADR (Petrovic et al., 2012). In a systematic review conducted by Christensen and Lundh (2013), a medication review reduced emergency department visits in older adults. Although more research is needed on the efficacy of medication review, the authors suggest the review should be a part of regular geriatric care.

OTC Medications

Self-medication with OTC medications, herbal remedies, and dietary supplements may lead to adverse drug–drug or drug–disease interactions (Rochon, 2014; Tachjian, Maria, & Jahangir, 2010). Older adults are the largest consumer of OTC medications and dietary supplements. The seven top purchased herbal medicines include St. John’s wort, ginseng, gingko biloba, echinacea, saw palmetto, and kava, all of which can interact with OTCs and prescription medications, warfarin being the drug that most commonly interacts with herbals and OTCs (Izzo & Ernst, 2009). In the United States, community-dwelling older adults take about as many OTC drugs as prescription drugs. In a study of 6,887 patients, self-medication represented almost 4% of ADRs and prescribed and OTC drugs accounted for the rest of the ADRs (Schmiedl et al., 2014). ADRs related to self-medication occurred in women aged 70 to 79 years and in men aged 60 to 69 years. NSAIDs and acetylsalicylic acid (ASA or aspirin) were the most frequent ADRs, causing GI complaints. ADRs resulting from self-medication were caused by NSAIDs and OTC ASA most frequently.

TABLE 20.4

Risk Factors for Potential Adverse Drug Reaction in Older Adults

Medication-related factors

Class of medication

Anticoagulants and antiplatelets

Cardiovascular meds

Hypoglycemics (oral and insulin)

Benzodiazepines

Antipsychotics

CNS agents/sedative/hypnotics

Psychotropics

Anticholinergics

NSAIDs

TCAs

Opioid analgesics

Corticosteroids

Specific medication

Warfarin

Digoxin

Diuretics

Benadryl

Lithium salts

Chlorpropamide

Theophylline

Beta blockers

ACEIs/ARBs

ASA

Patient characteristics

Polypharmacy

Dementia/memory problems

Multiple chronic medical problems (> 4–6)

Reduced homeostatic mechanisms

Reduced lean body mass

Renal insufficiency (CrCl < 50 mL/min)

Recent hospitalization

Advanced age (85 years of age)

Frailty

Multiple prescribers

Regular use of alcohol (~1 fl oz/d)

Prior ADR

Nonadherence

Alcohol misuse/abuse

Provider factors

Prescribing potentially inappropriate medications

Individual clinical information for prescribing

Inadequate monitoring of adherence

Adding medication for nonspecific indication

Regular medication review

Monitoring: new medication, lab parameters, follow-up visit

Identifying functional changes—self-administration ability; cognition

ACEIs, angiotensin-converting enzyme inhibitors; ADR, adverse drug reaction; ARBs, angiotensin II receptor blockers; ASA, acetylsalicylic acid; CNS, central nervous system; CrCl, creatinine clearance; NSAIDs, nonsteroidal anti-inflammatory drugs; TCAs, tricyclic antidepressants;.

Sources: Buck et al. (2009); Chan, Hussain, Lawson, and Ormerod (2013); Kongkaew et al. (2013); Petrovic et al. (2012); Rochon (2014).

FIGURE 20.3

Prescribing cascade.

ADE, adverse drug event.

Reproduced from Rochon and Gurwitz (1997) with permission from BMJ Publishing Group Ltd.

According to the National Institute of Drug Abuse (NIDA, 2014), the combination of prescription medications and alcohol misuse is estimated to be 19% in older adults. The combination of alcohol and age-related renal insufficiency can worsen and result in chronic salicylate intoxication because of its water solubility. Cold remedies that include alcohol are a significant source of drug potentiation in aging adults. Indeed, alcohol consumption is frequently omitted from history taking of older adults, even though it interacts with OTC and prescription medications in frank and subtle ways to produce unintended drug harm (NIDA, 2014).

The OTCs most commonly implicated in hospital admissions are low-dose aspirin and NSAIDs (Schmiedl et al., 2014). In a study evaluating concomitant use of prescriptions, OTC, and herbal remedies (N = 1,000), more than half of major interactions involved the use of nonprescription therapies (Qato et al., 2008). Of those, almost half involved the use of anticoagulants (e.g., warfarin) or antiplatelet agents such as aspirin. Across all age groups, the concurrent use of aspirin and warfarin was significantly more common in men than women. Supplement use was typically nutritional products, including vitamins and minerals. Alternative treatments used for cardiovascular reasons included omega-3 fatty acids, garlic, coenzyme Q, and glucosamine-chondroitin (Qato et al., 2008). In this same study, major drug–drug interactions increased with age, with more than half of all major interactions caused by nonprescription medications. Half of the major interactions included warfarin and aspirin. The FDA has been evaluating OTC ingredients and encouraging active labeling of OTCs; although it has yet to be seen whether the FDA will be more specific on safety issues that relate to older adults. Health care providers need to specifically ask patients what OTC and herbal remedies they are taking and their frequency.

High-Risk Medications

Many studies have revealed common high-risk medications in older adults. Four medications or medication classes were implicated alone or in combination in 67% of hospitalizations: warfarin (33.3%), insulins (13.9%), oral antiplatelet agents (13.3%), and oral hypoglycemic agents (10.7%; Budnitz et al., 2011). Although warfarin is deemed high risk, it is often underprescribed because of provider fear of bleeding; however, older patients often gain a greater absolute reduction in risk of stroke (Steinman, Handler, Gurwitz, Schiff, & Covinsky, 2011). Purposeful monitoring and special attention given to these medications can significantly reduce the risk of serious adverse effects. Digoxin should also be closely monitored as one third of all emergency room visits are digoxin-related ADEs (Salvi et al., 2012; Steinman & Hanlon, 2010). According to a recent systematic review, monitoring of high-risk medications is more likely to reduce ADEs (vs. prescribing interventions) as providers can anticipate and address abnormal results immediately. Thus, proactive, appropriate monitoring is more likely than interventions to reduce ADEs. Other medications to monitor include BZDs, which are an independent risk factor for falls (Rochon, 2014). Prescription or OTC NSAIDs given with cardiac-dose ASA account for 30% of upper GI bleeding-related admissions. This may be prevented if NSAIDs were given with GI-protective agents such as a proton pump inhibitor or misoprostol. NSAIDs also lead to renal failure and heart failure (Salvi et al., 2012). Diphenhydramine (Benadryl) may lead to impaired cognition or urinary retention (in men) and antipsychotics may lead to falls, death, and are commonly associated with aspiration pneumonia (Steinman & Hanlon, 2010). These medications result in not only hospital admissions but also increased health care expenditures and morbidity (Bustacchini et al., 2009). Nurses should become familiar with high-risk medications prescribed for older adults and participate in monitoring and promptly reporting lab results to aid in reducing ADEs.

Hematological Agents

Anticoagulants and antiplatelet medications are commonly associated with ADEs resulting in hospital admissions in up to 42% of patients. Hemorrhage is the most common with GI bleeding causing around 85% of occurrences. Warfarin is more likely to cause bleeding than antiplatelet drugs, causing bleeding in 75% of all admissions (Salvi et al., 2012). However, RCTs have shown that some of the new anticoagulants are better than warfarin for prevention of stroke and systemic embolism, such as dabigatran (Salvi et al., 2012).

Warfarin has been identified throughout many research studies as among the highest risk medications, including prescription and herbal medications taken by older persons (Izzo & Ernst, 2009; Salvi et al., 2012). Warfarin is highly bound (approximately 97%) to plasma protein, mainly albumin. The high degree of protein binding is one of several mechanisms whereby other drugs interact with warfarin (Reine, Kongsgaard, Andersen, Thøgersen, & Olsen, 2010). Those with malnutrition and low albumin levels are at risk for unbound warfarin in the bloodstream and a higher risk of bleeding. Warfarin is metabolized by hepatic cytochrome p-450 isoenzymes predominantly to inactive metabolites excreted in the bile and by the kidneys. Warfarin metabolism may be altered in advanced age and in the presence of liver problems. Drug interactions with warfarin are extensive and include (a) drugs that inhibit warfarin metabolism and prolong prothrombin time (e.g., Cipro, phenytoin, amiodarone); (b) drugs that inhibit vitamin K activity (e.g., cephalosporins and high-dose penicillins); (c) additive effects with other anticoagulants such as aspirin, Lovenox, and others; and (d) drugs that reduce the effectiveness of warfarin such as phenytoin, barbiturates, cholestyramine, opioids, and others (Reine et al., 2010).

Warfarin must be monitored to evaluate the time it takes for blood to clot, using the international normalized ratio (INR), to ensure that the INR is within the target therapeutic range (2.0–3.0) although this range may vary depending on the reason for anticoagulation. This narrow therapeutic range is often difficult to achieve because of the many factors that can affect INR, including interactions with prescription and OTCs (Clarkesmith, Pattison, & Lane, 2013; Izzo & Ernst, 2009). In addition, the lab results must be acted on immediately as it will usually take 2 days to change the INR level; for example, if the level is too high the medication must be stopped to prevent bleeding. Evidence-based warfarin algorithms are available as is INR electronic monitoring. Older adults are at a significantly greater risk for bleeding on taking warfarin. Prescribing should be determined on accurate evaluation and the individual’s benefit-versus-risk ratio. In some instances, the benefit of a higher risk medication may be deemed optimal and thus appropriate with close monitoring (Salvi et al., 2012). The safety literature recommends that all providers and nurses become educated in proficient warfarin management to improve safety for patients.

Fifty-eight percent of older persons do not report use of herbal supplements. Commonly used herbal remedies—ginkgo biloba, garlic, and St. John’s wort—all interact with warfarin to increase its anticoagulant effect and may lead to serious bleeding problems (Clarkesmith et al., 2013; Izzo & Ernst, 2009; Tachjian et al., 2010). Many foods interact with warfarin, specifically those with high vitamin K content such as chickpeas, spinach, and green tea. It is imperative to identify older adults on warfarin who fall or are at risk of falling as their risk of serious injury increases. The risk of harm versus benefit must be weighed as a fall and head injury could lead to a serious outcome. The nurse should clarify the risk versus benefit with the primary prescriber when noting that a patient on warfarin is at high risk of a fall (Izzo & Ernst, 2009; Steinman & Hanlon, 2010).

Cardiovascular Drugs

Cardiovascular drug events represent up to 48% of all hospitalizations. Angiotensin-converting enzyme inhibitors (ACEIs), angiotensin II receptor blockers (ARBs), beta blockers, alpha blockers, calcium antagonists (all antihypertensive agents), as well as nitrates, digoxin, antiarrhythmics, and statins are of high risk for ADEs in older adults (Ho et al., 2009). In a systematic review, Salvi et al. (2012) reported that diuretics are involved in the larger number of hospital admissions across studies. Common diuretic interactions induce electrolyte disturbance, syncope, dehydration, hypotension, and falls. Hyponatremia and hyperkalemia caused ADEs related to loop and thiazide diuretics to be a significant problem in older adults and the risk increases if the person is also taking ACEIs and selective serotonin reuptake inhibitors (SSRIs; Salvi et al., 2012).

Digoxin is useful in treating CHF related to systolic dysfunction in the older adult but is not the recommended treatment for CHF from underlying diastolic dysfunction in older adults. Digoxin toxicity occurs more frequently in older adults, presents atypically, and may result in death. The dose of digoxin is often prescribed inappropriately (greater than 0.125 mg) in hospital patients, and toxic effects occur when digoxin is given in the presence of chronic kidney disease (Hamilton et al., 2011). Classic symptoms of digoxin toxicity (nausea, anorexia, and visual disturbance) may occur; however, symptomatic cardiac disturbance and arrhythmias are more common in the older adult. Older adults may experience toxicity symptoms even with normal plasma levels of digoxin (Hamilton et al., 2011).

Many older people will have some reduction in renal function with aging; therefore, monitoring for symptoms, especially atypical symptoms of digoxin toxicity, and monitoring renal function and potassium levels are key. Particular caution must be exercised when digoxin is prescribed with diuretics; this combination can cause hypokalemia and exacerbate renal impairment and potentiate digoxin toxicity. Because the therapeutic window for digoxin is narrow and it is water soluble, the drug has a smaller Vd, therefore, the plasma concentration is higher. Correct and safe dosing in older adults is challenging. The maximum recommended dose in older persons for treating systolic heart failure is 0.125 mg (Hamilton et al., 2011). Debilitated older adults who often have low serum albumin levels are at risk for elevated plasma levels resulting in digoxin toxicity.

Endocrine Agents

Antidiabetic medications and corticosteroids have been shown to be responsible for up to 28% of ADEs causing hospitalizations in older adults (Salvi et al., 2012). Older adults are at a higher risk of hypoglycemia because of age-related changes and changes in drug clearance and slowed hepatic metabolism (Lipska & Montori, 2013). Although diabetic complications have improved over the past 20 years, serious hypoglycemic events, which can result in rapid death, are on the rise and these events are more common than the adverse glycemic effects the tight control is treating; thus, the risk of harm may be considered greater than the benefit (Lipska et al., 2015).

Nearly 90% of all hospitalizations are caused by hypoglycemia with or without symptoms (Salvi et al., 2012). Older adults may present atypically with neurological symptoms, such as dizziness, confusion, delirium, and weakness, although seizures and loss of consciousness are also primary presentations in older adults. Falls, fractures, and cardiac events may also present as a hypoglycemic episode (Salvi et al., 2012). Hypoglycemic hospitalizations caused by insulin are more common than oral agents; yet sulfonylureas are commonly associated with hypoglycemia in older adults. In the hospital setting, many sulfonylureas prescribed with antibiotics have shown an increased risk of hypoglycemia, such as glipizide and glyburide. Antibiotics found to interact, including clarithromycin, levofloxacin, Bactrim, metronidazole, and ciprofloxacin, were associated with increased odds of severe hypoglycemia (Lipska et al., 2015).

Recently, sliding scale therapy was determined to be an ineffective treatment and it increased the risk of prolonged hypoglycemia, particularly in long-term care. Basal insulin, or basal plus rapid-acting insulin, with one or more meals (often called basal/bolus insulin therapy) is recommended as it most closely mimics normal physiologic insulin production. It also controls blood glucose more effectively. End-stage renal disease can be prevented or delayed with aggressive treatment in those at high risk; however, tight glycemic control must be assessed by closely evaluating the individual’s risk of hypoglycemia. When the likelihood of benefit is examined in the context of limited life expectancy, the risk would be greater than the benefit. Intensive blood pressure control (less than or equal to 140/80 mmHg) is paramount to preserve renal function in diabetes mellitus (DM). Since the landmark United Kingdom Prospective Diabetes Study (UKPDS) in 2000, intensive blood pressure control has shown to preserve renal function and to reduce both diabetes-related morbidity and mortality, such as a significant reduction in microvascular and macrovascular complications, including strokes and heart failure, which were reduced in half with tight control. This also includes smoking cessation, glucose control, and lipid management (Steinman & Holmes, 2014). Again, limited life expectancy may make the risk of tighter blood pressure control greater than the benefit. ACEIs or ARBs are the primary choice for hypertensive treatment, although renal function must first be evaluated (Jarred & Kennedy, 2010; Steinman & Holmes, 2014).

Older adult diabetics are also susceptible to geriatric syndromes such as depression, cognitive impairment, urinary incontinence, chronic pain, polypharmacy, and injurious falls. The American Geriatrics Society (AGS) recommends that older adults be evaluated for atherosclerotic heart disease, functional status, and geriatric syndromes. Older adults should also be evaluated for their understanding of diabetes management, symptoms of hypoglycemia and, if they have any symptom, they should know the common presentation and treatment for them. Continual evaluation of this ability and support should be provided to older adults to avoid hypoglycemia. Diabetes management should be individualized. Annual self-management training is covered under Medicare Part B.

Factors contributing to polypharmacy in diabetes include tight glycemic control that often requires a multiple drug regimen. Likewise, the multiple comorbidities that tend to coexist with diabetes add to the number of medications and risk. Other co-occurring illnesses that may require more than one medication to control the disease include hypertension, coronary artery disease, early renal disease, glaucoma, and neuropathy. Polypharmacy, drug interactions (oral hypoglycemics with insulin or sulfonyureas), renal failure (decrease clearance of oral agents), and cognitive impairment (reduce ability to manage) increase the likelihood of hypoglycemia (Salvi et al., 2012). Multiple medications add to the cost for patients and may also lead to nonadherence.

Central Nervous System Agents

ADEs are often related to the following central nervous system (CNS) agents: BZDs; SSRIs and tricyclic antidepressants; antipsychotics; mood stabilizers such as lithium; and antiparkinson, antiseizure meds (especially Keppra), and dementia agents (Salvi et al., 2012). These drugs are responsible for up to 20% of all hospitalizations related to drugs. Change in mental status, syncope, falls, GI and respiratory disturbances, neuropsychiatric problems, and hyponatremia are often reported in older adults. Fall injuries are a leading injury related to psychotropic drug use, including hypnotics, antidepressants, and BZDs. Cholinesterase inhibitors and memantine, given for Alzheimer’s disease, are associated with a high risk of bradycardia, syncope, and hip fractures. Treatment of behavioral problems is often inappropriately done with antipsychotics in patients with dementia. These medications have been associated with an increased risk of mortality, aspiration pneumonia, stroke, and sudden cardiac death secondary to arrhythmia. Psychotropic drugs have been associated with falls, including sedatives, hypnotics, and antidepressants. Recent literature suggests that SSRIs have a higher risk than tricyclics (Salvi et al., 2012).

INTERVENTIONS AND CARE STRATEGIES

Comprehensive Evaluation and Medication Management

Interventions to reduce ADEs in older adults include comprehensive geriatric assessment, evaluation of patient risk for ADEs; identifying high-risk drugs or classes of drugs to monitor closely; evaluating benefit versus risk of harm to prescribing; computerized decision-making support and prescribing systems with alerts for potential drug interactions; MR; comprehensive medication review with pharmacist support; geriatric medicine services and comprehensive geriatric evaluation and multidisciplinary team drug reviews; initiation of nonpharmacological interventions; multifaceted interventions to reduce polypharmacy (unnecessary medications and PIMs); and patient outreach and activation to be involved in education and self-management safety (Kongkaew et al., 2013; Salvi et al., 2012; Steinman et al., 2011; Topinková, Baeyens, Michel, & Lang, 2012).

Prevention of ADEs

It is prudent that health care team members are aware of the risk factors for ADEs in older adults and which medications put older adults at high risk of adverse events. In a prospective observational study, Kongkaew et al. (2013) reported that patient age, length of time since starting a new drug, total number of prescription drugs, and hospitalization were significantly associated with ADEs. The most common preventable ADEs were antiplatelet drugs, anticoagulants, diuretics (loop and thiazide diuretics), ACEIs, and antiepileptic drugs (Kongkaew et al., 2013). Knowing and recognizing these drug classes as being high risk for ADEs support the importance of regular INR monitoring and ongoing evaluation for anticoagulants or chemistry panels for diuretics to monitor electrolytes or potential orthostatic hypotension, particularly in frail older adults. In initiating ACEIs, serum potassium and increasing creatinine levels need to be monitored closely. It is important for nurses to identify that hypotension and/or hyponatremia can occur when a patient is dehydrated or volume depleted. Patients taking potassium-sparing diuretics may experience hypokalemia even in the presence of loop diuretics. These high-risk drugs should be on the radar for potential ADEs for all providers (Kongkaew et al., 2013).

Interventions for Reducing ADEs

In older adults with complex medical problems and needs, a global evaluation obtained through a comprehensive geriatric assessment (CGA) may be helpful in simplifying drug prescriptions. It will also aid in prioritizing pharmacological and health care needs, resulting in an improvement in quality of prescribing. Prescriptions should be evaluated for benefit versus harm choosing the safest and best medication for the individual. Before prescribing, providers should consider the individual’s comorbidities, such as liver or kidney disease or a disease that may interact with a drug (Salvi et al., 2012). Likewise, consideration of life expectancy and health status should be evaluated in addition to the presence of frailty, which may increase the risk of ADEs and affect quality of life. Although many researchers have looked at reducing PIMs in research, the yield has been low. Computerized decision-support systems, pharmacist interventions, and CGA have shown more positive outcomes (Salvi et al., 2012). Nonpharmacological interventions and complementary and alternative medicine (CAM) therapies are underused in health care and should be considered before prescribing in many circumstances or used as an adjunctive therapy to prescribing.

Comprehensive Geriatric Assessment

An accurate medication history must be obtained from each patient and validate that the medication history is true. The amount and types of medication typically consumed, including OTCs and herbal remedies, should be examined as well as an estimate of how long each has been taken. The brown-bag method for medication review has shown to be a more accurate evaluation of a person’s current medication regimen (Topinková et al., 2012). See the Assessment Tools section. Providers can also check the refill date and the number of medications remaining in the bottle or consult the pharmacy to aid in accuracy. FitzGerald (2009) recommends the following medication history:

Currently prescribed drugs, doses, route of administration, frequency taken and duration of treatment