Patterns of Newly-Prescribed Benzodiazepines in Late Life

As the proportion of older adults in the United States continues to increase, concerns about use of psychotropic medications continue to be an important consideration for providers. Use of sedative-hypnotic medications in older adults, particularly benzodiazepines (BZDs), has been a topic of debate in the medical community for several decades. The topic was most recently revisited in the American Geriatric Society Beers Criteria update, which states that use of BZDs should be avoided in older adults due to increased risk of cognitive impairment, delirium, falls, fractures and motor vehicle accidents. The goal of the current study is to describe patterns of new prescriptions of BZDs in a defined population of community-dwelling older adults who are participants in the State of Pennsylvania Pharmaceutical Assistance Contract for the Elderly (PACE/PACENET) Behavioral Health Laboratory (BHL) Project.

PACE/PACENET enrollees who receive new prescriptions of psychotropic medications (i.e. anxiolytics, antidepressants, or antipsychotics) are referred to the BHL weekly. The BHL conducts a standardized structured phone interview of participants that incorporates demographic information, prescription patterns (dose, scheduled intervals, frequency, compliance, perceived indication and prescriber data), and measures of functioning and overall health-related quality of life. These include the Blessed Orientation-Memory-Concentration (BOMC), Mini International Neuropsychiatric Interview, Patient Health Questionnaire-9 (PHQ-9), Medical Outcomes Survey (SF-12), and the Generalized Anxiety Disorder-7 Scale (GAD-7). Current analysis identified participants with newly-prescribed BZDs and divided them into two groups: those prescribed standing doses versus on as-needed basis (PRN). Dose equivalents were calculated for those with standing prescriptions where the daily dose was known using lorazepam as the standard (1mg). To compare PRN against standing prescriptions, independent samples t-tests and chi-square tests were applied. For standing prescriptions, Pearson's correlations and Spearman's rho were calculated to examine associations with equivalent dosage conversions.

Comparison of data from participants who were prescribed standing doses of BZDs versus PRN revealed that 122 (39.5%) were prescribed a standing dose, 161 (52.1%) on a PRN basis, 16 subjects reported not taking the medication, and the remaining 10 subjects did not provide enough information to conclude their dosing schedule. The ages of the standing and PRN groups varied in a statistically significant manner [79.5 (SD 6.7) and 77.9 (SD 6.4) years respectively, p = 0.039], while the groups did not vary in other demographic variables (sex, race, marital status or financial status). Those prescribed BZDs on a standing basis were more cognitively impaired with a BOMC mean score of 5.1 (SD 3.8) versus 4.3 (SD 3.3) for the PRN group (p = 0.050). There were no significant differences between groups in PHQ-9 scores; or GAD-7 scores. Those prescribed BZDs on a PRN basis were more likely to self-report taking the medication for anxiety/worry/nervousness/tension than those prescribed a standing dose (PRN 62.5% versus standing 50.8%, p = 0.049). Primary care physicians accounted for 81.9% of the prescribers, psychiatrists for 1.0% and other specialists for 11.7%. Analysis of dose equivalents in the standing BZD group revealed a mean equivalent dose of 0.94mg per day with a standard deviation of 0.58mg. Men received slightly higher average doses than women (1.2mg (SD 0.6) mg/day, versus 0.9mg (SD 0.6), p = 0.011). There was no statistically significant difference in dosing based on race, marital status or use of other psychotropic medications. Pearson correlation calculations demonstrated statistically significant differences in age (-0.314, p=0.000) and GAD-7 scores (0.209, p=0.049), although Spearman's rho calculation only confirmed the age difference to be statistically significant (-0.282, p=0.002).

Despite continuing concerns about BZD prescriptions in older adults, they continue to be commonly prescribed in community-dwelling older adults. Primary care physicians account for the majority of prescribers in this sample and the mean dose for standing medications is less than 1mg of lorazepam equivalent. There were no clinical differences in those receiving a PRN versus standing dose suggesting that the prescribing practice may be driven more by differences in providers than differences in clinical presentation. Further research on the long term effects of initiating treatment with BNZ is needed to understand the risks.