MEDICATION USE IN THE OLDER POPULATION
Other Countries: Use of Drugs in the Community-Dwelling Older Population
Polypharmacy is common in the older community-dwelling population of other countries. This section of the report summarizes recent studies of medication use by older community-dwelling individuals in Spain (one study), Canada (two studies), Finland (three studies), Sweden (one study), and Denmark (one study).
In one study of 65 patients with heart failure in Spain (49 males and 16 females age 60.5 ± 12.0 [mean ± SD] years), it was found that 74 percent of the subjects were taking 6 or more pills per day and that 28 percent were taking 11 or more pills per day (Martínez-Sellés et al., 2004). One key finding of the study was the high rate of homeopathic and alternative medicine use among female patients, one-third of whom reported the use of such treatments.
In a sample of 1,216,000 community-dwelling adults 66 and older in Ontario, Canada, 40,307 (3.31%) were prescribed a drug that was on the Beers list and categorized by the study authors as “drugs to always avoid” or “drugs that are rarely appropriate” (Lane et al., 2004). For the 12,162 older people who received a drug on the “always-avoid list,” the majority (63%) were prescribed flurazepam.14 Meperidine15 accounted for 19 percent of the always-avoid prescriptions, chlorpropamide16 accounted for 16 percent, and barbiturates accounted for 4 percent. For the 28,985 older people who received a drug categorized as “rarely appropriate,” diazepam17 accounted for 85 percent of the prescriptions and chlordiazepoxide18 accounted for 15 percent of the prescriptions.
Hogan, Maxwell, Fung, and Ebly (2003) reported on the prevalence of benzodiazepine use in Canada. In the mid 1990s, approximately 25 percent of senior citizens in Nova Scotia annually received a prescription for a benzodiazepine. In 1991, 16 percent of Saskatchewan seniors received a benzodiazepine prescription. They conducted a study to evaluate the change in benzodiazepine use in a group of 1,181 subjects 65 and older in a 5-year longitudinal study between 1990/91 (T1) and 1996 (T2). They found that although the average number of medications (prescribed, non-prescribed, and nutritional supplements) increased from 3.9 to 5.8, the proportion of subjects using benzodiazepines at T1 and T2 was similar (26.4% versus 25.2%). The most commonly used classes of medications (and percentage of subjects using them) were: analgesics (58.5%); diuretics (32.7%); anxiolytics (29.5%); cardiac drugs (24.5%); and antihypertensives (17.2%). With regard to use of benzodiazepines, at T2, 74 percent of users were taking benzodiazepines regularly and 26 percent were using them only as needed. Approximately half of the T2 benzodiazepine users had been consuming benzodiazepines at T1. A key finding between T1 and T2 was a decline in the proportion of subjects consuming long half-life benzodiazepines (nonsignificant) and triazolam19 (significant), which, the authors suggest, may indicate better prescribing practices as physicians have been cautioned about the use of both.
It should be noted here that with hypnotic drugs such as benzodiazepines, the duration of hypnotic effect and the profile of unwanted effects may be influenced by the distribution and elimination half-lives of the administered drug and any active metabolites formed. When half-lives are long, the drug or metabolite may accumulate during periods of nightly administration and be associated with impairments of cognitive and motor performance during waking hours. If half-lives are short, the drug and metabolites will be cleared before the next dose is ingested, and carry-over effects related to sedation or CNS depression should be minimal or absent (RxMed, 2005). Furthermore, benzodiazepines or active metabolites with very long elimination half-lives can accumulate with chronic dosing and produce prolonged effects, especially in elderly or obese patients, those with liver disease, or with concurrent use of other drugs that compete for hepatic oxidation. The updated Beers criteria indicate that short- and intermediate-acting benzodiazepines are preferred over long-acting benzodiazepines, if a benzodiazepine is required (Fick et al., 2003).
Another key finding by Hogan et al. (2003) at T2 was that older people with a depressed mood (Geriatric Depression Scale score greater than 5) were more likely to be prescribed a benzodiazepine (37%) than an antidepressant (26.9%). The authors state that benzodiazepines are often used by depressed individuals, despite the fact that with the exception of alprazolam, they are believed to be ineffective for the primary treatment of depression. They cite their earlier findings that physicians appear to overprescribe benzodiazepines and underprescribe antidepressants in depressed older patients (Hogan, Ebly, and Fung, 1995). Their 2003 data showing a higher rate of benzodiazepine than antidepressant use in seemingly depressed patients suggests potentially inappropriate psychotropic prescribing.
People over 64 years of age represent 15 percent of the Finnish population, but their medication costs comprise 40 percent of the expenditures of the total Finnish population (Linjakumpu et al., 2002). These authors conducted two cross-sectional surveys among community-dwelling people 64 and older in 1990-1991 and 1998-1999 to investigate changes in the use of prescription drugs and polypharmacy. The mean age was the same in both surveys (72 for men and 73 for women). In this study, polypharmacy was defined as the concomitant use of over 5 medications. Nonprescription drugs were not included in this study. The number of medications per person increased from 3.1 to 3.8, and polypharmacy increased from 19 percent of the sample to 25 percent of the sample. These changes were most prominent among the population 85 and older, especially women. Medication use grew from 93 percent to 97 percent in subjects 84 and older, between the two surveys. In 1998/99, women 84 and older used an average of 6.8 medications. Drug users with polypharmacy were older than those without polypharmacy, and two-thirds of those with polypharmacy were women. Cardiovascular and central nervous system (CNS) medications were used most commonly in both surveys, and CNS medication use increased significantly between surveys. Of CNS medication users, 84 percent used psychotropics or psychostimulants in 1990/91 and 72 percent in 1998/99. Main categories of medications among the elderly with polypharmacy included cardiovascular (90% of the elderly in both surveys), followed second by alimentary tract/metabolic system medications (67% in the first survey), and CNS medications (63% in the second survey). The use of medications for blood/blood-forming organs and the genitourinary system grew most significantly from the first to the second survey among people with polypharmacy (42% to 56% and 10% to 29%, respectively). Of the various subgroups of cardiovascular medications, the use of beta-blocking agents, calcium-channel blockers, agents acting on renin-angiotensin system and statines grew significantly between surveys. For those using musculoskeletal system medications, anti-inflammatory and antirheumatic medication use grew from 80 to 84 percent between surveys.
The same populations were evaluated for their use of psychotropics by Linjakumpu et al. (2002). In the data analysis, psychotropics were divided into four groups: anxiolytics, hypnotics, antipsychotics (including lithium and neuroleptics) and antidepressants. Twenty-five percent of the sample was taking at least one psychotropic drug in both surveys. Fifty percent of subjects 85 and older used psychotropics in 1998/99. Hypnotics/sedatives, which mainly consist of benzodiazepines, were the most commonly used psychotropics in both surveys, and most users were taking them regularly.
In the Linjakumpu et al. (2002) study, the use of hypnotics and antidepressants increased most during the study period. Hypnotics increased from 11 percent to 15 percent between the surveys. The use of short-acting benzodiazepines (midazolam, triazolam) decreased from 7 percent to 2 percent, while the use of non-benzodiazepine sleeping pills (zopiclone, zolpidem) increased from 1 percent to 9 percent. Eleven percent used long-acting benzodiazepines in the first survey and 8 percent in the second survey. They were taken regularly by 57 percent of the users in the first survey and 60 percent of the users in the second survey. The use of benzodiazepines increased only among the oldest subjects, from 25 percent to 28 percent. Concomitant use of two or more psychotropics increased significantly from 7 percent to 10 percent between the surveys. Approximately one-third of all psychotropic users were taking at least two psychotropic medications concomitantly; and the proportion of users of three or more psychotropics was 6 percent in 1990/91 and 8 percent in 1998/99. Of all psychotropic users, 1 percent were taking four or more psychotropics concomitantly in both surveys. Polypharmacy and the use of psychotropics were most prevalent among those 85 and older, with women predominating. Concomitant use doubled from 11 percent to 22 percent among the oldest subjects. Drug users with polypharmacy used psychotropics more commonly (53%) than those without polypharmacy (21%) across both study periods.
Linjakumpu et al. (2002) found that subjects age 64 to 71 used cyclic antidepressants equally commonly in both surveys. A key finding was that none of these subjects used the new atypical antipsychotics in 1998/99. In 1990/91, most users had a combination of a hypnotic/sedative and an antipsychotic medication (3%) while in 1998/99, the most common combination was a hypnotic/sedative and an antidepressant (5%). The authors note that concomitant use of two or more CNS medications (including psychotropics and opioids) is associated with an increased risk of falls more often than the use of CNS-active drugs in general. Linjakumpu et al. (2002) caution that long-acting benzodiazepines and cyclic antidepressants impair cognition, and cause drowsiness in the morning and tiredness in the daytime. Cyclic antidepressants also cause blurring of vision, and may cause symptoms such as parkinsonism.
Mantyselkaet al. (2002) studied multiple medication use among Finnish patients visiting their primary care physicians due to nonacute musculoskeletal pain. Study subjects included 358 adults with a mean age of 54 (37% were 60 or older). The most common pain locations were back, lower extremities, and neck. Half of the patients had scores on the General Health Questionnaire (Goldberg and Williams, 1988) and one-fourth had scores on the Beck Depression Inventory (Beck and Beck, 1972) indicating depression. Sixty percent of the patients had used prescription drugs and approximately half had used over-the-counter drugs for pain management during the week prior to the office visit. For most analgesics, patients in Finland need a prescription; however, ibuprofen, ketoprofen, paracetamol, and acetylsalicylic acid are available as OTC drugs. Approximately one-third used more than one drug simultaneously due to pain. Depression was associated with the use of multiple medications due to pain, as well as the daily use of medications due to pain. OTC drug use was more common among females and patients who lived alone, than among males and married patients. One of six patients used sedatives or anxiolytics due to pain. The use of anxiolytics or sedatives was more common among older and less educated patients as well as those who were not employed. Mental distress and depression were independently associated with the use of anxiolytics and sedatives. A statistically significant relationship was found between depression and multiple drug use. Depression appeared to be underdiagnosed among study participants, however, as depression was recorded as a secondary reason for the visit for two patients. The authors note that excessive use of anxiolytics and sedatives among patients with prolonged pain may be an indicator of undetected and untreated mental disorders.
The prescription drug use of older, community-dwelling people in Tierp, Sweden,in 1994 was studied by Jörgensen et al. (2001). Prescription drug and diagnosis information were obtained using a computerized research register for all people 65 and older living in the community. Prescriptions purchased outside of the community are not included in the register, resulting in a estimated loss of 5 percent of the data on medication use. There were 4,642 subjects 65 and older residing in Tierp in 1994, 78 percent of whom had at least one prescription filled at the local pharmacies in that year.
Cardiovascular agents were most commonly used among both genders (47.2% of the study sample), with diuretics being the largest subgroup, followed by beta-blockers. Women used more cardiac drugs than men, and the highest use of cardiac drugs was in patients age 75 to 84. Approximately 10 percent of the elderly used cardiac glycosides. Thirty-seven percent of the sample used nervous system drugs, with opioids (14.8%) and hypnotics/sedatives (14.2%) the most commonly used subgroups. Use of anxiolytics was higher among women (17.4%) than men (7.9%). The third largest group of prescription agents was gastrointestinal drugs (34.2% of the sample) with antiulcer drugs as the largest subgroup. Oral antidiabetic drugs were used by 7.4 percent of the subjects and 2.4 percent used insulin. Respiratory system medications were prescribed for 22.8 percent of the subjects, equally distributed between men and women except in the oldest age group (85+) where men had higher use. Cough preparations was the largest subgroup, with 15 percent of the elderly using prescribed cough suppressants and expectorants. Use of all pharmacologic groups was significantly lower for women 85 and older compared to those 75 to 84, with the exception of nervous system drugs. The largest difference in use was found for cardiovascular drugs and anti-infective agents. A similar trend was shown for the men, except that among the oldest men, use was significantly lower only for cardiovascular drugs (beta-blockers, anticoagulants, and lipid-lowering agents). The differences in prescription drug use between men and women were largest in the youngest age group (65-74), where significantly more women used 9 of the 14 pharmacologic groups than men. In this age group, more men than women used antithrombotic agents. Among the oldest subjects (85+), there were more pharmacologic classes where men had a significantly higher use than women (calcium-channel blockers, antibiotics, respiratory drugs).
In the Jörgensen et al., (2001) study, the average number of prescription drugs per subject was 4.3 (range = 0 to 34). Multiple drug use (5 or more prescription drugs) occurred in 39 percent of the population, with a higher percentage among women than men (43% versus 34%). Almost 13 percent of the sample used 10 or more drugs. It is not known if all the drugs were used simultaneously. Also, the percentage of polypharmacy could be higher if nonprescription drugs were taken into account. A multivariate analysis showed that the number of primary care visits was the greatest determinant of multiple drug use. Visiting a physician 5 times or more during the year increased the risk of using 5 or more different drugs by approximately 15 times compared with not being seen. A large number of diagnoses also increased the risk for multiple drug use. The authors note that the strong relationship could imply that the elderly were seriously ill and needed frequent follow-up and extensive drug therapy. Since the data is cross-sectional, it could also be that multiple drug use had consequences that made the elderly seek care.
In a survey of 636 drivers stopped randomly by police in Denmark and asked to take home a survey to complete and return, 5.8 percent indicated that they had used a medicinal or illegal drug within the 24-hour period prior to being stopped (Behrensdorff and Steentoft, 2002). Among them, 3 percent had taken hypnotics, tranquilizers, or analgesic drugs (e.g., “hazardous medicinal drugs”). One driver reported using cannabis. The remaining 2.8 percent reported the use of other types of “non-hazardous”20 drugs such as drugs for high blood pressure, stomach ulcer, etc. In this study, 44 percent of the respondents were 45 or older; however, the ages of the drivers reporting having used a medicine in the prior 24-hour period were not reported. In response to the question, “Do you ever drive a few hours after having had both a hazardous medicinal drug and alcohol,” 31 of the 361 respondents (8.5%) reported “yes” or “occasionally.” Of these respondents, 19.3 percent were 18 to 24; 22.6 percent were 24 to 44; and 48.4 percent were 44 and older. The distribution by gender for driving after consuming alcohol and a hazardous drug showed that 84 percent of these respondents were men and 16 percent were women.
14 Flurazepam (brand name: Dalmane in the United States) is a benzodiazepine derivative. It is a hypnotic agent prescribed for insomnia. Brand names in Canada include Apo-Flurazepam ; Novo-Flupam ; PMS-Flupam; Somnol; and Som Pam..
15 Meperidine (brand names: Demerol, Isonipecaine, and Pethidine)is used to relieve moderate to severe pain. Meperidine is in a class of medications called narcotic analgesics, a group of pain medications similar to morphine. It works by changing the way the body senses pain.
16 Chlorpropamide (“Diabinese”) is used to treat type II (noninsulin-dependent) diabetes. It lowers blood sugar by stimulating the pancreas to secrete insulin and helping the body to use insulin efficiently.
18 Chlordiazepoxide (Brand names: Libritabs; Librium; Limbitrol (combination with amitriptyline); Mitran; Reposans-10; Sereen.) is used to relieve anxiety and to control agitation caused by alcohol withdrawal.
20 In Denmark, hazardous medicinal drugs include all types of drugs that are considered a potential hazard in relation to road safety or machine handling. Such drugs are labeled with a red triangle and have a package information insert.