Inappropriate Scripts Persist for Elderly in Primary Care

Larry Hand

August 23, 2012

August 23, 2012 — Twenty percent of medicines for elderly people in primary care are inappropriately prescribed, despite recent efforts to improve the quality of medical prescriptions for seniors, according to a study published online August 22 in PLoS One.

Dedan Opondo, a PhD student from the Department of Medical Informatics, University of Amsterdam, the Netherlands, and colleagues conducted a systematic review of published literature on inappropriate medical prescription (IMP) for elderly people in primary care, with the last search conducted on March 8, 2012.

IMP can be defined as prescriptions that subject patients to significant risk for adverse drug reactions when evidence exists that another medication is equally or more effective, or a prescription of a medication that fails to achieve optimal effects. IMP can be characterized as under-, mis-, or overprescribing.

Of 946 articles screened, 19 were included in the final analysis, meeting the criteria of describing experiences of elderly people aged 65 years or older who received medical prescriptions in outpatient, primary care settings; in other words, studies met inclusion criteria if the setting was described as outpatient clinic, office practice, general practice, and/or primary health care clinic. Reporting was based on drug-age criteria. The analysis excludes institutionalized and community-dwelling seniors when the setting in which the medication was prescribed was not indicated.

The studies covered 11 countries, including the United States, 7 European countries (Italy, United Kingdom, the Netherlands, Germany, Ireland, Norway, and Portugal), Taiwan, Iran, and India. Most of the studies (15/19) used Beers criteria. Four studies used multiple criteria sets to assess appropriateness of medication use.

Overall, the median rate for IMPs came to 20.0% (absolute range, 2.9% - 38.5%; interquartile range, 16.8% - 25.4%), with US studies showing a median IMP rate of 19.6% (range, 4.5% - 33.3%) and European studies showing a median rate of 19.1% (range, 2.9% - 38.5%).

When the researchers grouped prescriptions by therapeutic classes, they found that the analgesic propoxyphene had the highest median IMP rate, at 4.52% (range, 0.10% - 23.30%), followed by the antihypertensive doxazosin (3.96%; range, 0.32% - 15.70%), the anticholinergic diphenhydramine (3.30%; range, 0.02% - 4.40%), the antidepressant amitriptyline (3.20%; range, 0.05% - 20.5%), the antiarrhythmic digoxin (3.10%; range, 0.01% - 21.1%), the anticholinergic clidinium (3.1%; range, 3.01% - 3.1%), and the sedative hypnotic diazepam (2.74%; range, 0.05% - 30.0%). The anticholinergic belladonna alkaloids had the lowest IMP rate (0.04%; range, 0.0% - 0.5%).

Diphenhydramine and amitriptyline were the most common high-risk drugs inappropriately prescribed, whereas propoxyphene and doxazosin were the most common low-risk drugs inappropriately prescribed.

"This review found that one in five (20.0%) prescriptions to elderly persons is inappropriate with marked variation of rates of IMP within individual therapeutic classes," the researchers write. They point out that of the 19 studies, 16 were published after 2000, and even though much attention has focused on IMP and elderly people in the past 11 years, the rate seems not to have gone down.

Factors that may have contributed to the wide range of variation in IMP (ie, 2.9% - 38.5%) include lack of a universal list of recommended medications among different countries, regional differences in cost and purchasing systems for specific medications, and variation in local drug procurement policies and structure of medication financing.

The researchers write that prescriptions for high-risk medications such as diazepam and nifedipine subject elderly people to "frequent and severe adverse drug events," and that interventions such as clinical decision support systems need to be used. "These systems can provide alerts during prescription based on medication prescription guidelines such as the Beers criteria," they write.

Limitations of their study include possible publication bias among the 19 studies, plus possible relevance among the other studies not included in this analysis. Strengths include the analysis within therapeutic classes.

"This analysis is useful to policy makers and clinicians when making choices between medications from the same therapeutic class. We also compared IMP in studies which utilised multiple instruments to measure quality of prescription," the researchers write.

They conclude, "Focused and systematic interventions are needed to improve the quality of medication prescription in this patient group."

The authors have disclosed no relevant financial relationships.

PLoS One. Published online August 22, 2012. Full text