Abstract
Objectives: to measure sex differences in the risk of receiving potentially inappropriate prescription drugs and to examine what are the factors that contribute to these differences.
Design: a retrospective cohort study.
Setting: community setting of British Columbia, Canada.
Participants: residents of British Columbia aged 65 and older (n = 660,679).
Measurements: we measured 2013 period prevalence of prescription dispensations satisfying the American Geriatrics Society's 2012 version of the Beers Criteria for potentially inappropriate medication use in older adults. We used logistic regressions to test for associations between this outcome and a number of clinical and socioeconomic factors.
Results: a larger share of women (31%) than of men (26%) filled one or more potentially inappropriate prescription in the community. The odds of receiving potentially inappropriate prescriptions are associated with several clinical and socioeconomic factors. After controlling for those factors, community-dwelling women were at 16% higher odds of receiving a potentially inappropriate prescription than men (adjusted odds ratio = 1.16, 95% confidence interval = 1.12–1.21). Much of this sex difference stemmed from women's increased odds of receiving potentially inappropriate prescriptions for benzodiazepines and other hypnotics, for tertiary tricyclic antidepressants and for non-selective NSAIDs.
Conclusion: there are significant sex differences in older adults' risk of receiving a potentially inappropriate prescription as a result of complex intersections between gender and other social constructs. Appropriate responses will therefore require changes in the information, norms and expectations of both prescribers and patients.
Keywords: inappropriate, Beers criteria, older adults, sex and gender, socioeconomic disparities, older people
Introduction
Despite known risks, potentially inappropriate prescribing is relatively common, with several studies across North America and Europe reporting 20% or greater prevalence of potentially inappropriate medication use among community-dwelling older adults [1–4]. Studies have found certain factors—such as patient age, health status and the number of prescription drugs they are taking—to be positively associated with the risk of receiving potentially inappropriate prescriptions [1–7]. Studies have also reported sex differences in exposure to potentially inappropriate medications; however, some studies have found that sex differences are moderated and in some cases reversed, when patient age health status or income are taken into account [7–13].
In this study, we document sex differences in older adults' risk of receiving potentially inappropriate prescriptions in British Columbia, Canada. We draw on comprehensive, population-based linked healthcare datasets that include information about residents' age, health, income and ethnicity. This allows us to measure sex differences in the risk of receiving potentially inappropriate prescriptions and to examine clinical and socioeconomic factors that contribute to these differences.
Methods
Study design and setting
This is a retrospective study of outpatient prescription drug purchases by residents of British Columbia who were aged 65 and older in 2013. All subjects were covered under British Columbia's universal, public health insurance program for medical and hospital care, and all were eligible for coverage under British Columbia's universal, public drug benefit plan, under which deductibles are set in relation to household income.
Data sources and cohort
We obtained de-identified linked health datasets from Population Data BC, with approval of relevant data stewards and the University of British Columbia's Behavioural Research Ethics Board [14–16]. The datasets included administrative records of all prescription drug dispensations, fee-for-service physician visits and hospitalisations for all residents aged 65 or older in 2013, except military veterans, registered First Nations, and inmates of federal penitentiaries (which collectively make up ∼4% of the population). To accurately measure period prevalence of medicine use, we excluded individuals who lived in British Columbia for <275 days in 2013.
Variables
We measured period prevalence of calendar year 2013 prescription dispensations satisfying the American Geriatrics Society's 2012 version of the Beers Criteria for potentially inappropriate medication use in older adults [17]. We implemented the Beers criteria for drug type, dose, duration and, where relevant, medical conditions (Table 2 of the 2012 publication).
Health status and medical conditions relevant to the Beers criteria were identified though diagnosis codes in medical and hospital records. A primary diagnostic code (ICD-10) is contained in records of every fee-for-service billing for primary and speciality care. Records of each hospitalisation contain up to 25 diagnostic codes. We gauged overall health status using counts of major Aggregated Diagnostic Groups (ADGs of the John Hopkins ACG case-mix adjustment system, version 10.0) that have been validated for studying health services and pharmaceutical use [18, 19].
We defined polypharmacy as the use of drugs from five or more different drug classes defined by the third level of the World Health Organisation's Anatomical Therapeutic Chemical drug classification system [20]. We defined patients who visited five or more different physicians in the year as having many providers of medical care.
Datasets contained validated household-specific income data for 78% of our study population and neighbourhood-based proxy incomes for the remaining 22% [21]. We assigned ethnicity using a validated algorithm to identify surnames of the dominant ethnic minorities in British Columbia: Chinese (40% of minorities) and South Asians (26%) [22, 23]. Finally, we categorised neighbourhood urbanisation based on the population density of the Local Health Area in which people lived.
Statistical analyses
We compute study population characteristics and χ2 tests for significance across groups. We ran sex-stratified and sex-pooled logistic regressions to test for associations between the binary exposure measure (prevalence of one or more potentially inappropriate prescription) and explanatory variables selected based on established models of health services utilisation and sex- and gender-based analyses [24–27]. After testing for collinearity between explanatory variables and goodness of fit, our models included measures of sex, age, health status, concomitant drug use (polypharmacy), the number of physicians providing care, income, marital status, ethnicity and level of neighbourhood urbanisation. We also tested interactions between an individual's sex and other explanatory variables that theory predicts may have sex-specific effects: specifically, age, health status, income and ethnicity [26]. For all analyses, a value of P < 0.05 was considered statistically significant. All analyses were performed using Stata version 13.1 (College Station, TX, USA) and SAS version 9.3 (SAS Institute, Cary, NC, USA).
Results
Table 1 describes the characteristics of the study population. A total of 660,679 persons aged 65 and older resided in British Columbia for at least 275 days during 2013. Women made up just over half of this population (54%). A larger share of women (31%) than of men (26%) filled one or more potentially inappropriate prescription in 2013. Women in our study population were more likely to be over age 85, reside in a long-term care facility, fill prescriptions for five or more different types of drug and have incomes in the lowest quintile—all of these characteristics were associated with higher crude prevalence of filling one or more potentially inappropriate prescription. Men in our study population had relatively poor health status, which was associated with higher crude prevalence of potentially inappropriate prescriptions.
Table 1.
Study population characteristics, older British Columbians, 2013
| Variable | Women |
Men |
Women and men |
||||||
|---|---|---|---|---|---|---|---|---|---|
| n | % | Prevalence of PIP (%) | n | % | Prevalence of PIP (%) | n | % | Prevalence of PIP (%) | |
| Population | 357,165 | 100 | 31 | 303,514 | 100 | 26 | 660,679 | 100 | 28 |
| Agea | |||||||||
| 65–74 | 186,947 | 52 | 29 | 177,307 | 58 | 24 | 364,254 | 55 | 27 |
| 75–84 | 110,563 | 31 | 33 | 94,723 | 31 | 29 | 205,286 | 31 | 31 |
| 85+ | 59,655 | 17 | 30 | 31,484 | 10 | 27 | 91,139 | 14 | 29 |
| Health statusa | |||||||||
| 0 Major ADGs | 147,192 | 41 | 20 | 110,343 | 36 | 15 | 257,535 | 39 | 18 |
| 1 Major ADG | 106,297 | 30 | 33 | 91,219 | 30 | 27 | 197,516 | 30 | 30 |
| 2–3 Major ADGs | 87,272 | 24 | 41 | 84,065 | 28 | 35 | 171,337 | 26 | 38 |
| 4+ Major ADGs | 16,404 | 5 | 52 | 17,887 | 6 | 45 | 34,291 | 5 | 48 |
| Residing in long-term carea | 13,462 | 4 | 42 | 6,100 | 2 | 42 | 19,562 | 3 | 42 |
| Polypharmacy (5+ drug classes)a | 164,022 | 46 | 50 | 128,174 | 42 | 45 | 292,196 | 44 | 48 |
| Many providers (5+ physicians) | 212,080 | 59 | 39 | 180,179 | 59 | 33 | 392,259 | 59 | 36 |
| Income quintilea | |||||||||
| Lowest | 91,676 | 26 | 35 | 45,065 | 15 | 31 | 136,741 | 21 | 34 |
| Second | 70,849 | 20 | 32 | 56,699 | 19 | 28 | 127,548 | 19 | 30 |
| Third | 69,576 | 19 | 29 | 63,471 | 21 | 25 | 133,047 | 20 | 27 |
| Fourth | 68,129 | 19 | 27 | 70,154 | 23 | 24 | 138,283 | 21 | 26 |
| Fifth | 56,935 | 16 | 29 | 68,125 | 22 | 24 | 125,060 | 19 | 26 |
| Relationship statusa | |||||||||
| Single | 189,878 | 53 | 31 | 83,744 | 28 | 27 | 273,622 | 41 | 30 |
| Marriage-like relationship | 167,287 | 47 | 30 | 219,770 | 72 | 26 | 387,057 | 59 | 27 |
| Ethnicity | |||||||||
| European and other | 313,925 | 88 | 32 | 266,904 | 88 | 26 | 580,829 | 88 | 29 |
| Chinese | 31,045 | 9 | 20 | 25,784 | 8 | 21 | 56,829 | 9 | 21 |
| South Asian | 12,195 | 3 | 31 | 10,826 | 4 | 29 | 23,021 | 3 | 30 |
| Neighbourhood urbanisationa | |||||||||
| Metropolitan | 224,523 | 63 | 30 | 182,905 | 60 | 25 | 407,428 | 62 | 28 |
| Mixed urban/rural | 108,229 | 30 | 33 | 96,513 | 32 | 27 | 204,742 | 31 | 30 |
| Rural | 24,413 | 7 | 32 | 24,096 | 8 | 27 | 48,509 | 7 | 29 |
PIP, one or more potentially inappropriate prescription.
aDifferences in population characteristics statistically significant at P = 0.05.
Tables 2 and 3 list adjusted odds ratios from logistic regression analyses for the population living in the community setting (results for residents of long-term care facilities are in the Supplementary data, Appendix, available in Age and Ageing online). Excluding interaction terms, women had 23% higher odds of receiving one or more potentially inappropriate prescription than men after adjusting for all other clinical and socioeconomic factors (AOR = 1.23, 95% CI = 1.22–1.25). Including interaction terms that allow sex to modify the effects of age, health status, income and ethnicity, women had 16% higher adjusted odds of receiving a potentially inappropriate prescription than men (AOR = 1.16, 95% CI = 1.12–1.21). Model specification tests favoured the inclusion of the interaction terms.
Table 2.
Adjusted odds ratios for the likelihood of filling at least one potentially inappropriate prescription by community-dwelling British Columbians aged 65 and older, sex-stratified and pooled results
| Variable | Women |
Men |
Women + men, no interactions |
Women + men, with interactions |
||||
|---|---|---|---|---|---|---|---|---|
| AOR | 95% CI | AOR | 95% CI | AOR | 95% CI | AOR | 95% CI | |
| Sex | ||||||||
| Men (ref.) | – | – | 1.00 | 1.00 | ||||
| Women | – | – | 1.23 | 1.22–1.25 | 1.16 | 1.12–1.21 | ||
| Age | ||||||||
| 65–74 (ref.) | 1.00 | 1.00 | 1.00 | 1.00 | ||||
| 75–84 | 0.94 | 0.92–0.95 | 0.99 | 0.97–1.01 | 0.96 | 0.94–0.97 | 0.99 | 0.97–1.01 |
| 85+ | 0.89 | 0.87–0.92 | 0.93 | 0.90–0.96 | 0.91 | 0.89–0.92 | 0.93 | 0.90–0.97 |
| Sex and age interaction | ||||||||
| Women × 65–74 (ref.) | – | – | – | 1.00 | ||||
| Women × 75–84 | – | – | – | 0.95 | 0.92–0.97 | |||
| Women × 85+ | – | – | – | 0.95 | 0.91–0.99 | |||
| Health status | ||||||||
| 0 Major ADGs (ref.) | 1.00 | 1.00 | 1.00 | 1.00 | ||||
| 1 Major ADG | 1.01 | 0.99–1.03 | 1.00 | 0.97–1.02 | 1.01 | 0.99–1.02 | 1.00 | 0.97–1.02 |
| 2–3 Major ADGs | 1.07 | 1.04–1.09 | 1.03 | 1.00–1.06 | 1.05 | 1.03–1.07 | 1.03 | 1.01–1.06 |
| 4+ Major ADGs | 1.32 | 1.27–1.38 | 1.26 | 1.21–1.31 | 1.29 | 1.26–1.33 | 1.27 | 1.22–1.32 |
| Sex and health status interaction | ||||||||
| Women × 0 Major ADGs (ref.) | – | – | – | 1.00 | ||||
| Women × 1 Major ADG | – | – | – | 1.01 | 0.98–1.05 | |||
| Women × 2–3 Major ADGs | – | – | – | 1.03 | 1.00–1.06 | |||
| Women × 4+ Major ADGs | – | – | – | 1.04 | 0.98–1.09 | |||
| Polypharmacy | ||||||||
| 5+ Drug classes | 4.20 | 4.13–4.28 | 4.14 | 4.06–4.23 | 4.17 | 4.12–4.23 | 4.18 | 4.12–4.23 |
| Number of providers | ||||||||
| 5+ Providers | 1.14 | 1.12–1.16 | 1.16 | 1.14–1.19 | 1.15 | 1.13–1.17 | 1.15 | 1.13–1.17 |
Table 3.
Adjusted odds ratios for the likelihood of filling at least one potentially inappropriate prescription by community-dwelling British Columbians aged 65 and older, sex-stratified and pooled results
| Variable | Women |
Men |
Women + men, no interactions |
Women + men, with interactions |
||||
|---|---|---|---|---|---|---|---|---|
| AOR | 95% CI | AOR | 95% CI | AOR | 95% CI | AOR | 95% CI | |
| Income quintile | ||||||||
| Lowest (ref.) | 1.00 | 1.00 | 1.00 | 1.00 | ||||
| Second | 1.02 | 0.99–1.04 | 0.99 | 0.96–1.02 | 1.02 | 1.00–1.04 | 0.98 | 0.95–1.01 |
| Third | 1.00 | 0.97–1.03 | 0.94 | 0.91–0.97 | 0.98 | 0.96–1.00 | 0.92 | 0.90–0.95 |
| Fourth | 0.99 | 0.96–1.01 | 0.93 | 0.90–0.96 | 0.97 | 0.95–0.99 | 0.91 | 0.88–0.93 |
| Fifth | 1.02 | 0.99–1.05 | 0.88 | 0.85–0.91 | 0.95 | 0.93–0.97 | 0.86 | 0.83–0.88 |
| Sex and income interaction | ||||||||
| Women × lowest (ref.) | – | – | – | 1.00 | ||||
| Women × second | – | – | – | 1.05 | 1.01–1.09 | |||
| Women × third | – | – | – | 1.10 | 1.06–1.14 | |||
| Women × fourth | – | – | – | 1.11 | 1.07–1.15 | |||
| Women × fifth | – | – | – | 1.22 | 1.17–1.27 | |||
| Relationship status | ||||||||
| Single (ref.) | 1.00 | 1.00 | 1.00 | 1.00 | ||||
| Marriage-like relationship | 1.00 | 0.98–1.02 | 0.92 | 0.89–0.94 | 0.97 | 0.96–0.98 | 0.96 | 0.95–0.98 |
| Ethnicity | ||||||||
| Other (ref.) | 1.00 | 1.00 | 1.00 | 1.00 | ||||
| Chinese | 0.75 | 0.73–0.78 | 1.00 | 0.96–1.03 | 0.85 | 0.83–0.87 | 0.99 | 0.95–1.03 |
| South Asian | 0.83 | 0.79–0.87 | 0.95 | 0.91–1.00 | 0.89 | 0.86–0.92 | 0.95 | 0.90–0.99 |
| Sex and ethnicity interaction | ||||||||
| Women × European and other (ref.) | – | – | – | 1.00 | ||||
| Women × Chinese | – | – | – | 0.76 | 0.73–0.80 | |||
| Women × South Asian | – | – | – | 0.88 | 0.82–0.94 | |||
| Neighbourhood urbanisation | ||||||||
| Metropolitan (ref.) | 1.00 | 1.00 | 1.00 | 1.00 | ||||
| Mixed urban/rural | 1.12 | 1.10–1.14 | 1.08 | 1.05–1.10 | 1.10 | 1.08–1.11 | 1.10 | 1.08–1.11 |
| Rural | 1.13 | 1.09–1.17 | 1.13 | 1.09–1.17 | 1.13 | 1.10–1.16 | 1.13 | 1.11–1.16 |
For women and for men, being sicker, receiving polytherapy and receiving care from five or more doctors all increased the adjusted odds of filling one or more potentially inappropriate prescription. Being older was associated with lower adjusted odds of receiving a potentially inappropriate prescription. Tests for interactions between sex and age found that the protective effect of age was slightly greater for women than for men. Interactions between sex and health status were not statistically significant.
As shown in Table 3, higher income was associated with lower odds of filling potentially inappropriate prescriptions for men but not for women. Being married reduced the odds that a man would receive a potentially inappropriate prescription (AOR = 0.92, 95% CI = 0.89–0.94), but did not have a significant effect on those odds for women. Ethnicity had statistically significant effects on the odds that a woman received a potentially inappropriate prescription, but no statistically significant effects for men. For women, having a Chinese surname reduced the odds of receiving a potentially inappropriate prescription by 25% (sex-stratified AOR = 0.75, 95% CI = 0.73–0.78), having a South Asian surname reduced those odds by 17% (sex-stratified AOR = 0.83, 95% CI = 0.79–0.87). Tests for interactions between sex and ethnicity confirmed that being female significantly modified the effects of ethnicity.
Table 4 presents sex-stratified prevalence of exposure to leading types of drugs on the Beers list among community-dwelling British Columbians over age 65 in 2013. It also lists odd ratios of such exposures for women (compared with men) after adjusting for age, health status, polypharmacy, receipt of prescriptions from multiple doctors, income, marital status, ethnicity and neighbourhood urbanisation. (Results for the population residing in long-term care facilities are in the Supplementary data, Appendix, available in Age and Ageing online.) By far, the potentially inappropriate medications most frequently prescribed for older adult British Columbians were benzodiazepines and other non-benzodiazepine hypnotics (e.g. eszopiclone) for long-term use. A greater proportion of women (12.9%) than men (8.4%) were prescribed 90 or more days' worth of these medicines in 2013. Women had 55% greater adjusted odds of such long-term hypnotic use than men (AOR = 1.55, 95% CI = 1.52–1.58).
Table 4.
Prevalence and adjusted sex differences in odds of filling at least one potentially inappropriate prescription, by drug types with >1% prevalence of use, community-dwelling British Columbians aged 65 and older, 2013
| Drug class | Women |
Men |
Women + men |
Adjusted odds ratio (women) |
||||
|---|---|---|---|---|---|---|---|---|
| n | % | n | % | n | % | AOR | 95% CI | |
| Benzodiazepines and other hypnotics (>90 days) | 38,829 | 12.9 | 21,657 | 8.4 | 60,486 | 10.8 | 1.55 | 1.52–1.58 |
| Nifedipine | 11,967 | 4.0 | 9,339 | 3.6 | 21,306 | 3.8 | 1.01 | 0.98–1.04 |
| Tertiary tricyclic antidepressants | 12,021 | 4.0 | 4,835 | 1.9 | 16,856 | 3.0 | 2.18 | 2.11–2.26 |
| Long-duration sulfonylureas | 6,664 | 2.2 | 9,015 | 3.5 | 15,679 | 2.8 | 0.56 | 0.54–0.58 |
| Estrogens with or without progestins | 14,744 | 4.9 | 33 | 0.0 | 14,777 | 2.6 | – | – |
| Non-COX selective NSAIDs (>90 days) | 7,590 | 2.5 | 5,386 | 2.1 | 12,976 | 2.3 | 1.20 | 1.15–1.24 |
| Skeletal muscle relaxants | 7,457 | 2.5 | 5,380 | 2.1 | 12,837 | 2.3 | 1.19 | 1.14–1.23 |
| Indomethacin | 2,580 | 0.9 | 8,231 | 3.2 | 10,811 | 1.9 | 0.25 | 0.24–0.27 |
| Spironolactone | 5,333 | 1.8 | 5,151 | 2.0 | 10,484 | 1.9 | 0.86 | 0.83–0.90 |
| Antiarrhythmic drugs | 3,815 | 1.3 | 4,281 | 1.7 | 8,096 | 1.4 | 0.87 | 0.83–0.91 |
| Fast-acting insulin | 3,228 | 1.1 | 4,434 | 1.7 | 7,662 | 1.4 | 0.64 | 0.61–0.67 |
| First-generation antihistamines | 4,261 | 1.4 | 3,173 | 1.2 | 7,434 | 1.3 | 1.10 | 1.04–1.15 |
| Anti-infective | 5,649 | 1.9 | 1,323 | 0.5 | 6,972 | 1.2 | 3.87 | 3.63–4.12 |
| Alpha 1 blockers | 898 | 0.3 | 4,965 | 1.9 | 5,863 | 1.0 | – | – |
| Androgens | 164 | 0.1 | 4,252 | 1.6 | 4,416 | 0.8 | – | – |
AOR, adjusted odds ratio, adjusted for age, health status, polypharmacy, receipt of prescriptions from multiple doctors, income, marital status, ethnicity and neighbourhood urbanisation.
Prescriptions for nifedipine matching the Beers criteria for being potentially inappropriate were the next most frequently prescribed drug type for older adult British Columbians in 2013. There were no sex differences in the odds of receiving potentially inappropriate nifedipine prescriptions after adjusting for all other factors that influence such risks. The third most frequently prescribed Beers list drugs were tertiary (first generation) tricyclic antidepressants, prescriptions for which were filled by 4.0% of women and 1.9% of men. The adjusted odds that a woman would receive a potentially inappropriate prescription for tertiary tricyclic antidepressants were more than twice that of men (AOR = 2.18, 95% CI = 2.11–2.26).
Women were at greater crude and adjusted odds of using several other categories of potentially inappropriate medications than men. These included non-selective NSAIDs, muscle relaxants, first-generation antihistamines and nitrofurantoin. Similarly, women were less likely than men to fill prescriptions from several other categories of potentially inappropriate medications: these included long-duration sulfonylureas, spironolactone, indomethacin, antiarrhythmic drugs and fast-acting insulin. Some Beers list drug categories were almost exclusively used by women (estrogens) and men (α-1 blockers and androgens) owing to sex-specific indications for their use.
Discussion
We found that 28% of older adult residents of British Columbia filled one or more potentially inappropriate prescription in 2013. The crude prevalence of receiving potentially inappropriate prescriptions was higher among women than men (31 versus 26%). Women were at 16–23% greater odds of exposure to potentially inappropriate prescription drugs than men, even after adjusting for sex differences in clinical and socioeconomic factors associated with the use of potentially inappropriate medications. Much of this sex difference stemmed from women's increased odds of receiving potentially inappropriate prescriptions for benzodiazepines and other hypnotics, for tertiary tricyclic antidepressants and for non-selective NSAIDs.
The prevalence of potentially inappropriate prescription drug among older British Columbians is within the range of prevalence estimates for community-dwelling older adults in other countries, most of which fall between 20 and 30% [2, 6]. Our finding that women are at increased odds of receiving a potentially inappropriate prescription is consistent with some prior research [10, 12, 13, 28]. However, this finding is not unanimously supported in the literature [8]. For example, Bradley et al. [9] document that men are at increased odds of receiving a potentially inappropriate prescription in the UK. Other studies have found that women's increased odds of receiving a potentially inappropriate prescription is moderated when patient age, health status or income are taken into account [7, 10, 11].
That some results concerning sex differences have been sensitive to adjustments for age, health status and income suggest that the risk of potentially inappropriate medication use is shaped by both biological and social forces. Biological influences include direct effects of sex differences in the prevalence of conditions for which medications may be inappropriately prescribed. There may also be indirect biological influences, such as if sex differences in health status result in different patterns of health services use, including the number of care providers, that, in turn, increase the risk of potentially inappropriate care—as has been demonstrated with risks of polypharmacy [27, 29, 30]. Sex differences in the risks of potentially inappropriate medication use may also result from social forces, including social dimensions of gender, such as how health professionals differentially diagnose and treat women and men who present with similar conditions [31–36]. Finally, sex differences in risk may result from the intersection between sex, gender and other socioeconomic influences on health and health care, including income and ethnicity, both of which can shape patient expectations and relationships between patients and providers [26].
For older residents of British Columbia, we found evidence of direct biological influences on sex differences in risk of potentially inappropriate medication use: age and health status both contributed to risks in ways that explain, in part, crude sex differences in the prevalence of exposure. Similarly, we found evidence of indirect biological influences on sex differences: polypharmacy and having multiple prescribers contributed to risks in ways that help explain crude sex differences in the prevalence of potentially inappropriate medication use. All of these associations are consistent with findings of several other studies [1–7].
We also found socioeconomic factors associated with the risk of receiving potentially inappropriate prescriptions. Higher income was associated with lower odds of receiving a potentially inappropriate prescription among men but not women. This may be a result of an intersection between wealth, sex and power in relationships between patients and healthcare providers and/or between patients and social supports [26, 27]. Ethnicity also influenced women's likelihood of being exposed to a potentially inappropriate prescription, but did not do so for men. This finding is also consistent with other findings, particularly those concerning ethnic variations in psychotropic drugs, the use of which may carry greater stigma for women of Asian ethnicity found to be at lower risk of exposure in our study [37, 38].
Study limitations
This study is not without limitations. There are several different criteria with which to assess appropriateness of prescribing; we selected the Beers criteria, because it is the most widely applied measure in the literature and has been used in multiple jurisdictions [2, 5]. A recent review of methods for measuring the prevalence of potentially inappropriate medication use found women at higher risk of exposure across methods even though prevalence rates differed [28]. Furthermore, the drugs accounting for much of the sex difference observed in our study are found in most (NSAIDs) or all (benzodiazepines, tertiary tricyclic antidepressants) of the major lists of potentially inappropriateness prescriptions for older populations [17, 39–42]. As such, it is unlikely that the nature of sex differences identified in this study would differ using alternative criteria.
The linked administrative data that we use contained information needed to adjust for the explicit dose, duration and medical diagnoses that designate inappropriate use of Beers drugs; the lack of such details has been a criticism of prior studies using the Beers criteria [5]. Nevertheless, we were unable to review the full clinical data (including lab values) that would be attainable through a chart audit or analysis of electronic medical records. As such, we may have over-adjusted for some of the Beers criteria.
Our measure of potentially inappropriate prescribing is based on prescription dispensations. While dispensation of prescribed drugs is not equivalent to consumption of the medicines, it is likely that most patients who invest the time and out-of-pocket costs necessary to have prescriptions filled do so with intent to consume them. Moreover, as some prescriptions will be written but not filled by patients, this measure is arguably an understatement of the extent of potentially inappropriate prescribing in British Columbia.
Conclusion
There are significant sex differences in older adults' risk of receiving a potentially inappropriate prescription in British Columbia, even after adjusting for clinical and socioeconomic factors that might influence sex difference. Sociodemographic disparities in access to potentially beneficial care might, in some cases, be justifiable on the grounds of differences in patient preferences for specific treatment options, including patient beliefs about the role of medications in their treatment. However, it would be difficult to justify such differences in risk of exposure to potentially inappropriate prescriptions based on patient beliefs or preferences, because no group should be exposed to a higher level of risk when lower risk alternatives exist.
Findings of this study—including sex differences in the effects of income, ethnicity and marriage—suggest that the elevated risks that women face are a result of complex intersections between biological and social constructs. Appropriate responses will therefore need to be both nuanced and fundamental. There is the obvious, fundamental need to invest in the dissemination of information and tools to assist with de-prescribing of potentially inappropriate medications. Such tools need to be targeted to and appropriate for both prescribers and patients. There is also a more nuanced need to study and invest in processes to address how gender—on its own and interacting with age, wealth and ethnicity—affect the norms of and relationships between prescribers and patients.
Key points.
The odds of receiving potentially inappropriate prescriptions is higher among women, even after adjusting for confounding.
Women receive inappropriate prescriptions for benzodiazepines, tricyclic antidepressants and NSAIDs more frequently than men.
Approaches to address inappropriate prescribing must include changes in norms and expectations of both prescribers and patients.
Authors' contributions
S.G.M. is responsible for study concept and design, acquisition of data, interpretation of results and preparation of manuscript. D.W. and B.P. assisted with study design, analysis of data, interpretation of results and editing of manuscript for important intellectual content. K.S., E.J.G., C.R. and B.M. assisted with study design, interpretation of results and editing of manuscript for important intellectual content.
Supplementary data
Supplementary data mentioned in the text are available to subscribers in Age and Ageing online.
Conflicts of interest
None declared.
Funding
This work was supported by the Canadian Institutes of Health Research (grant number MOP 119360). The funding agency had no role in product design, methods, data collection, analysis or preparation of the paper. All opinions and conclusions drawn are those of the authors and do not reflect the opinions or policies of the Data Stewards.
Supplementary Material
References
- 1.Aparasu RR, Mort JR. Inappropriate prescribing for the elderly: beers criteria-based review. Ann Pharmacother 2000; 34: 338–46. [DOI] [PubMed] [Google Scholar]
- 2.Guaraldo L, Cano FG, Damasceno GS, Rozenfeld S. Inappropriate medication use among the elderly: a systematic review of administrative databases. BMC Geriatr 2011; 11: 79. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Hill-Taylor B, Sketris I, Hayden J, Byrne S, O'Sullivan D, Christie R. Application of the STOPP/START criteria: a systematic review of the prevalence of potentially inappropriate prescribing in older adults, and evidence of clinical, humanistic and economic impact. J Clin Pharm Ther 2013; 38: 360–72. [DOI] [PubMed] [Google Scholar]
- 4.Liu GG, Christensen DB. The continuing challenge of inappropriate prescribing in the elderly: an update of the evidence. J Am Pharm Assoc (Wash) 2002; 42: 847–57. [DOI] [PubMed] [Google Scholar]
- 5.Gallagher P, Barry P, O'Mahony D. Inappropriate prescribing in the elderly. J Clin Pharm Ther 2007; 32: 113–21. [DOI] [PubMed] [Google Scholar]
- 6.Davidoff AJ, Miller GE, Sarpong EM, Yang E, Brandt N, Fick DM. Prevalence of potentially inappropriate medication use in older adults using the 2012 Beers criteria. J Am Geriatr Soc 2015; 63: 486–500. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Cahir C, Fahey T, Teeling M, Teljeur C, Feely J, Bennett K. Potentially inappropriate prescribing and cost outcomes for older people: a national population study. Br J Clin Pharmacol 2010; 69: 543–52. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Santos AP, da Silva DT, dos Santos Junior GA et al. . Evaluation of the heterogeneity of studies estimating the association between risk factors and the use of potentially inappropriate drug therapy for the elderly: a systematic review with meta-analysis. Eur J Clin Pharmacol 2015; 71: 1037–50. [DOI] [PubMed] [Google Scholar]
- 9.Bradley MC, Fahey T, Cahir C et al. . Potentially inappropriate prescribing and cost outcomes for older people: a cross-sectional study using the Northern Ireland Enhanced Prescribing Database. Eur J Clin Pharmacol 2012; 68: 1425–33. [DOI] [PubMed] [Google Scholar]
- 10.Lane CJ, Bronskill SE, Sykora K et al. . Potentially inappropriate prescribing in Ontario community-dwelling older adults and nursing home residents. J Am Geriatr Soc 2004; 52: 861–6. [DOI] [PubMed] [Google Scholar]
- 11.Mort JR, Aparasu RR. Prescribing potentially inappropriate psychotropic medications to the ambulatory elderly. Arch Intern Med 2000; 160: 2825–31. [DOI] [PubMed] [Google Scholar]
- 12.Bierman AS, Pugh MJV, Dhalla I et al. . Sex differences in inappropriate prescribing among elderly veterans. Am J Geriatr Pharmacother 2007; 5: 147–61. [DOI] [PubMed] [Google Scholar]
- 13.Johnell K, Weitoft GR, Fastbom J. Sex differences in inappropriate drug use: a register-based study of over 600,000 older people. Ann Pharmacother 2009; 43: 1233. [DOI] [PubMed] [Google Scholar]
- 14.O'Brady S, Gagnon M-A, Cassels A. Reforming private drug coverage in Canada: inefficient drug benefit design and the barriers to change in unionized settings. Health Policy 2015; 119: 224–31. [DOI] [PubMed] [Google Scholar]
- 15.Renwick MJ, Smolina K, Gladstone EJ, Weymann D, Morgan SG. Postmarket policy considerations for biosimilar oncology drugs. Lancet Oncol 2016; 17: e31–8. [DOI] [PubMed] [Google Scholar]
- 16.Tannenbaum C, Diaby V, Singh D, Perreault S, Luc M, Vasiliadis H-M. Sedative-hypnotic medicines and falls in community-dwelling older adults: a cost-effectiveness (decision-tree) analysis from a US Medicare perspective. Drugs Aging 2015; 32: 305–14. [DOI] [PubMed] [Google Scholar]
- 17.American Geriatrics Society Beers Criteria Update Expert Panel. American Geriatrics Society updated Beers Criteria for potentially inappropriate medication use in older adults. J Am Geriatr Soc 2012; 60: 616–31. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Hanley GE, Morgan S, Reid RJ. Explaining prescription drug use and expenditures using the adjusted clinical groups case-mix system in the population of British Columbia, Canada. Med Care 2010; 48: 402–8. [DOI] [PubMed] [Google Scholar]
- 19.Weiner J, Abrams C. The Johns Hopkins ACG System Technical Reference Guide. Baltimore, MD: The Perl Foundation (2010) Perl 2009; 5: 762–765. [Google Scholar]
- 20.World Health Organization Collaborating Centre for Drug Statistics Methodology. Anatomical Therapeutic Chemical Code Classification index with Defined Daily Doses. http://www.whocc.no/atcddd/ (4 February 2014, date last accessed).
- 21.Hanley G, Morgan S. On the validity of area-based income measures to proxy household income. BMC Health Serv Res 2008; 8: 79. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Shah BR, Chiu M, Amin S, Ramani M, Sadry S, Tu JV. Surname lists to identify South Asian and Chinese ethnicity from secondary data in Ontario, Canada: a validation study. BMC Med Res Methodol 2010; 10: 42. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.British Columbia. The Diversity of Visible Minorities and Ethnic Origins in BC. Victoria: Ministry of Attorney General and Minister Responsible for Multiculturalism, 2008. [Google Scholar]
- 24.Andersen RM. Revisiting the behavioral model and access to medical care: does it matter. J Health Soc Behav 1995; 36: 1–10. [PubMed] [Google Scholar]
- 25.Phillips KA, Morrison KR, Andersen R, Aday LA. Understanding the context of healthcare utilization: assessing environmental and provider-related variables in the behavioral model of utilization. Health Serv Res 1998; 33(3 Pt 1): 571–96. [PMC free article] [PubMed] [Google Scholar]
- 26.Johnson JL, Repta R, Kalyan S. Implications of sex and gender for health research: from concepts to study design. In: Oliffe JL, Greaves L, eds. Designing and Conducting Gender, Sex, and Health Research. Thousand Oaks: Sage Publications, 2011. [Google Scholar]
- 27.Krieger N. Gender, sexes, and health: what are the connections—and why does it matter? Int J Epidemiol 2003; 32: 652–7. [DOI] [PubMed] [Google Scholar]
- 28.Morin L, Fastbom J, Laroche M-L, Johnell K. Potentially inappropriate drug use in older people: a nationwide comparison of different explicit criteria for population-based estimates. Br J Clin Pharmacol 2015; 80: 315–24. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Bjerrum L, Søgaard J, Hallas J, Kragstrup J. Polypharmacy: correlations with sex, age and drug regimen A prescription database study. Eur J Clin Pharmacol 1998; 54: 197–202. [DOI] [PubMed] [Google Scholar]
- 30.Moen J, Antonov K, Larsson CA et al. . Factors associated with multiple medication use in different age groups. Ann Pharmacother 2009; 43: 1978–85. [DOI] [PubMed] [Google Scholar]
- 31.Verbrugge LM, Steiner RP. Physician treatment of men and women patients: sex bias or appropriate care? Med Care 1981; 19: 609–32. [DOI] [PubMed] [Google Scholar]
- 32.Williams D, Bennett K, Feely J. Evidence for an age and gender bias in the secondary prevention of ischaemic heart disease in primary care. Br J Clin Pharmacol 2003; 55: 604–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Hohmann AA. Gender bias in psychotropic drug prescribing in primary care. Med Care 1989; 27: 478–90. [DOI] [PubMed] [Google Scholar]
- 34.Cleeland CS, Gonin R, Hatfield AK et al. . Pain and its treatment in outpatients with metastatic cancer. N Engl J Med 1994; 330: 592–6. [DOI] [PubMed] [Google Scholar]
- 35.Weisse C, Sorum P, Sanders K, Syat B. Do gender and race affect decisions about pain management? J Gen Intern Med 2001; 16: 211–7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36.Taggart LAP, McCammon SL, Allred LJ, Horner RD, May HJ. Effect of patient and physician gender on prescriptions for psychotropic drugs. J Womens Health 1993; 2: 353–7. [Google Scholar]
- 37.Morgan SG, Hanley G, Cunningham C, Quan H. Ethnic differences in the use of prescription drugs: a cross-sectional analysis of linked survey and administrative data. Open Med 2011; 5: e87–93. [PMC free article] [PubMed] [Google Scholar]
- 38.Puyat JH, Hanley GE, Cunningham CM et al. . Ethnic disparities in antipsychotic drug use in British Columbia: a cross-sectional retrospective study. Psychiatr Serv 2011; 62: 1–9. [DOI] [PubMed] [Google Scholar]
- 39.O'Mahony D, O'Sullivan D, Byrne S, O'Connor MN, Ryan C, Gallagher P. STOPP/START criteria for potentially inappropriate prescribing in older people: version 2. Age Ageing 2015; 44: 213–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 40.McLeod PJ, Huang AR, Tamblyn RM, Gayton DC. Defining inappropriate practices in prescribing for elderly people: a national consensus panel. Can Med Assoc J 1997; 156: 385–91. [PMC free article] [PubMed] [Google Scholar]
- 41.Laroche ML, Charmes JP, Merle L. Potentially inappropriate medications in the elderly: a French consensus panel list. Eur J Clin Pharmacol 2007; 63: 725–31. [DOI] [PubMed] [Google Scholar]
- 42.Holt S, Schmiedl S, Thurmann PA. Potentially inappropriate medications in the elderly: the PRISCUS list. Dtsch Arztebl Int 2010; 107: 543–51. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.