Benzodiazepines (BZs) are important, effective drugs when used appropriately for anxiety, sleep and seizure disorders. However, they are also sometimes misused by physicians and patients, with negative consequences including habituation, falls and fractures, overdose, and unnecessary cost. Few data exist on the details of their misuse in large populations, or on the intended and unintended effects of regulatory interventions aimed at reducing inappropriate prescribing and abuse. One common regulatory strategy for reducing misuse is to establish a triplicate prescription program (TPP) in which one copy of the prescription is forwarded to a state government surveillance unit. In January, 1989, the state of New York began requiring TPPs for all BZs in order to reduce BZ abuse and diversion and reduce inappropriate prescribing without affecting legitimate use. Although previous studies suggest that the TPP reduced BZ use by about 50%, many compelling questions remain unanswered, including the relative impact of the TPP on appropriate versus inappropriate use, and on use and costs of physician, emergency room, and hospital services among large populations of BZ recipients. Using a strong quasi-experimental design, we propose to study these effects among 180,000 Medicaid-enrolled adults in New York, and an identically defined comparison group in New Jersey Medicaid before and after initiation of the regulation, using patient-level Medicaid claims databases currently in the possession of the investigators. The study will be conducted in three phases. Phase I will describe patterns of use of BZs and other psychoactive drugs among adults (> 19 years) in three one-year cohorts (1988, the year before the TPP; 1990, the second year of the TPP; and 1994, the most recent year of data), stratified by three Medicaid eligibility categories, namely, Aid to Families with Dependent Children (AFDC), Aid to the Permanently and Totally Disabled (APTD), and Old Age Assistance (OAA). Comparing equivalently defined NY and NJ cohorts, Phase II will evaluate the impact of the TPP on both problematic use (e.g., high dose, doctor shopping, see A.,D.4) and probably appropriate (e.g., low-dose, short-term use, for convulsive disorders, see A..,D.4.), and on the use of substitute medications, among individuals who received BZs at all in 1988, and among those who received BZs regularly (> 4 months in 1988). Using focus group discussions with samples of psychiatrists and primary care physicians in NY and NJ, Phase II will also explore the current impact of the TPP on clinical decision-making in the treatment of anxiety and sleep problems, and reasons for the observed quantitative findings. Finally, using the same cohorts of recipients and regular recipients in NY and NJ, Phase III will utilize time-series and survival analysis to determine whether the TPP was associated with changes in use of other health services, such as physician visits, emergency room and clinic visits, hospital admissions, or total health expenditures.
