Long-term use of Benzodiazepines and Z-drugs: A register-based cohort study in Taiwan.
Benzodiazepines (BZDs) and Z-drugs are commonly prescribed for anxiety and insomnia, but their long-term use remains a public health concern. Evidence from Asia remains limited. This study investigated trajectories and predictors of long-term use in Taiwan by using a population-based cohort.
We conducted a register-based cohort study by using a nationally representative sample from Taiwan's National Health Insurance Research Database (NHIRD), covering a 20-year period from January 1, 2000, to December 31, 2019 after a 1-year washout. Incident users were identified and followed to evaluate long-term use patterns on the basis of two discontinuation thresholds: 90 days and 365 days. Cox regression models were used to estimate adjusted hazard ratios (aHRs) for long-term use by demographic, physical, and psychiatric factors.
Among 896,163 incident users, 3.8% progressed to long-term use under the 90-day definition, increasing to 14.7% under the 365-day definition. Under the 90-day definition, clonazepam and hypnotic BZDs showed the highest persistence (10.3% and 8.4%). More than 70% of initiations occurred in nonpsychiatric specialties. Long-term use was associated with male sex, older age, and higher Charlson Comorbidity Index; strong predictors included cancer, pneumonia, moderate-to-severe renal disease, alcohol- or drug-induced mental disorders, and personality disorders.
In Taiwan, most BZD and Z-drug use is short-term, and their usage patterns are affected by both clinical and system-level factors. These findings highlight the need for risk-based monitoring, cross-specialty prescribing oversight, and further research into treatment substitution strategies.
We conducted a register-based cohort study by using a nationally representative sample from Taiwan's National Health Insurance Research Database (NHIRD), covering a 20-year period from January 1, 2000, to December 31, 2019 after a 1-year washout. Incident users were identified and followed to evaluate long-term use patterns on the basis of two discontinuation thresholds: 90 days and 365 days. Cox regression models were used to estimate adjusted hazard ratios (aHRs) for long-term use by demographic, physical, and psychiatric factors.
Among 896,163 incident users, 3.8% progressed to long-term use under the 90-day definition, increasing to 14.7% under the 365-day definition. Under the 90-day definition, clonazepam and hypnotic BZDs showed the highest persistence (10.3% and 8.4%). More than 70% of initiations occurred in nonpsychiatric specialties. Long-term use was associated with male sex, older age, and higher Charlson Comorbidity Index; strong predictors included cancer, pneumonia, moderate-to-severe renal disease, alcohol- or drug-induced mental disorders, and personality disorders.
In Taiwan, most BZD and Z-drug use is short-term, and their usage patterns are affected by both clinical and system-level factors. These findings highlight the need for risk-based monitoring, cross-specialty prescribing oversight, and further research into treatment substitution strategies.