Context: Multiple medication use is associated with an increased incidence of adverse drug-drug interactions (DDIs), medication errors, noncompliance, hospitalization, and healthcare costs. Drugs acting systemically or gastrointestinally ("SG" drugs) are of particular concern because of their potential to interact. A better understanding is needed of the relationship between multiple medication use, particularly of SG drugs, and age, number of prescribers, and common drug regimens.
Objective: to determine the levels of multiple SG medication use in relation to age, number of prescribers, and common drug regimens in an outpatient population served by U.S. Veterans Integrated Service Network 15 (VISN 15).
Design, setting, and participants: cross-sectional analysis of the subset of 5,003 currently active patients from a stratified random sample of 7,000 potentially active outpatients (1,000 each from 7 sites comprising VISN 15) selected from the prescription database on a single day.
Main outcome measures: number of SG drugs/patient; number and frequency of SG drug entities and regimens.
Results: Most patients (97%) were dispensed at least one SG drug: 80% received > or =2 SG drugs, of which 42% received 2-4 SG drugs, 24%, 5-7 SG drugs, and 14%, > or =8 SG drugs. 394 different SG drugs were dispensed, only 88 of which occurred in 1% or more of patients. A significant increase (p < 0.0001) in level of multiple medication use occurred with increasing age and number of prescribers. Proportions of patients receiving 8 or more SG drugs approximately doubled with each additional prescriber, up to 4 or more prescribers. No drug regimen containing 2 or more drugs occurred in 1% or more of patients; 71% of patients were receiving a unique drug regimen (based on specific SG drugs without regard to dose or administration schedule).
Conclusions: The uniqueness of SG drug regimens suggests no single prescriber could have extensive clinical experience with even a small fraction of the drug regimens patients receive. These findings suggest that potential DDIs cannot be predicted based on occurrence of common drug regimens in a general patient population. A follow-up study (reported separately) investigated whether common drug regimens can be identified by selecting for a specific drug treatment (e.g., an antidepressant). The improved ability to predict DDIs is particularly relevant for psychiatric patients, who are at increased risk for DDIs because of greater frequency of multiple medication use. In addition, DDIs may present in this population in ways that mimic worsening of primary symptoms, which may lead to increased doses of the medication that is actually responsible for the problem, causing still more toxicity.