Long-term benzodiazepine use: factors of importance and the development of individual use patterns over time--a 13-year follow-up in a Swedish community.

Using data from a research registry of prescriptions, we studied benzodiazepine use in a Swedish community with a general population of 20,000. A sample of benzodiazepine users in 1976 (n = 561) aged 15 years and older was identified and followed for 13 years with respect to continued benzodiazepine use. A strong tendency towards continued use was observed. A majority of the cohort, 65%, continued benzodiazepine use during the first follow-up year, and 55% used benzodiazepines during the second. One-quarter of the sample continued using benzodiazepines during all years of the 13-year follow-up. One of the aims was to analyze factors predicting long-term benzodiazepine use. The multivariate analyses, using Cox regression analysis, showed that frequent/daily use and age were important factors. Gender and type of generic benzodiazepine were of little importance. Further, patients who were prescribed benzodiazepines by doctors working at hospitals and those who obtained prescriptions from both primary and hospital care physicians continued to use benzodiazepines to a greater extent than those patients who obtained prescriptions only from private practitioners or health center doctors. Another aim was to analyze to what extent long-term users were using these drugs on an infrequent, occasional, frequent, or daily basis and to what extent this use changed over time. Of those with benzodiazepine use persisting for eight or more years (n = 119), between one-half and two-thirds were frequent or daily users in each of those years. Because repeated measurements for the same individuals were analyzed, the generalized estimating equations (GEE) method was chosen for the multivariate analyses. Among long-term users, age, a combined use of tranquilizers and hypnotics, and prescriptions from more than one of the prescriber categories studied (i.e. doctors working at health centers, hospital doctors, and private doctors) were significant factors in frequent or daily use. The study also showed that frequent/daily use increased among the identified long-term users during the time period analyzed.