Polypharmacy

In the first phase of this study, a literature review was conducted within OVID for original articles published between January 1997 and May 2007, using the following search terms and phrases: ‘polypharmacy,’ ‘elderly,’ ‘geriatrics,’ ‘inappropriate medication,’ and ‘multiple medication use.’ English language articles available in local holding which described ‘polypharmacy’ or the issue of the simultaneous use of multiple medications in elderly patients were evaluated. Discrete definitions of polypharmacy were identified and recorded.
Throughout the literature, numerous articles used the term ‘polypharmacy’ and the phrase ‘inappropriate drug use’ interchangeably. The research reported that methods most often used for identifying inappropriate drug use involved the use of criteria, primarily the criteria developed, and more recently revised, by Beers and colleagues (Beers 1997 (link); Fick et al 2003 (link)). The investigators adapted a list of inappropriate drugs from two primary sources for use in this study (Fick et al 2003 (link); Bressler and Bahl 2003 (link)). The medications used in this research were based upon the Beers’ criteria but were limited to those identified as “high risk” (see Table 2). This list is labeled as ‘potentially inappropriate,’ because we recognize that use of one or more of these agents in an older adult could be justified by specific circumstances, for example, if safer alternatives had been exhausted.
Two different definitions of polypharmacy were applied to the database, which consisted of outpatient medical record data randomly collected by physician assistant students at the time of patient encounters during supervised clinical training from August 2006 to May 2007. Polypharmacy was defined as either “use of at least one potentially inappropriate drug” (see Table 2) or “the presence of six or more concurrent medications”. The database included patient demographics (ie, age, gender, ethnicity, educational level, and marital status), vital signs, diagnoses, prescription medications, health-related quality of life, and disease-specific markers when applicable (ie, blood pressure, urine microalbumin, hemoglobin A1c, creatinine clearance, liver transaminases, estimated left ventricular ejection fraction, and others). These outpatient data were collected from more than 500 outpatient clinical sites throughout the state of South Carolina representing the following disciplines: family medicine, general internal medicine, pediatrics, general surgery, gynecology and obstetrics, emergency medicine, and internal medicine subspecialties. At the time of analysis the database contained 10,455 discrete patient entries, and the 1270 entries involving patients 65-years or older were selected for investigation.