From acute presentation to post-discharge outcomes: intentional Z-drug versus benzodiazepine overdose in a retrospective cohort
7 affiliations
- 1Department of Rehabilitation, the Affiliated Suqian Hospital of Xuzhou Medical University
- 2Department of Emergency Medicine, Specialty Center for Poisoning Diagnosis and Treatment, the Affiliated Suqian Hospital of Xuzhou Medical University
- 3Department of Emergency Medicine, Specialty Center for Poisoning Diagnosis and Treatment, Suqian Clinical Medical College of Yangzhou University
- 4Department of Emergency Medicine, Specialty Center for Poisoning Diagnosis and Treatment, Nanjing Drum Tower Hospital Group Suqian Hospital
- 5Suqian Scientific Research Institute of Nanjing University Medical School
- 6Department of Emergency Medicine, the Affiliated Brain Hospital of Nanjing Medical University
- 7Department of Neurology, the Affiliated Brain Hospital of Nanjing Medical University
Abstract
Background: Z-drugs are alternatives to benzodiazepines (BZDs) for insomnia, but overdose safety remains uncertain. We compared acute and post-discharge outcomes after intentional Z-drug and BZD overdose.
Methods: This retrospective cohort included 239 adults (110 Z-drug; 129 BZD) treated at two emergency departments without flumazenil (2018–2024). Models adjusted for age, sex, dose category, intent, and center.
Results: Compared with BZD overdose, Z-drug overdose was associated with more recorded physical restraint use (adjusted odds ratio [aOR], 6.974; 95% confidence interval [CI], 3.358–14.483), more recorded patient-behavior-related medical interruption events (PB-MIEs; aOR, 5.955; 95% CI, 2.732–12.980), and a longer emergency-department stay (adjusted median difference, 2.516 h; 95% CI, 2.056–2.845), but fewer intensive care unit (ICU) admissions (aOR, 0.092; 95% CI, 0.031–0.272). Among admitted patients, total stay was longer after Z-drug overdose (228 vs 146 h, P < 0.001). The hazard of a first documented anxiety and/or depressive diagnosis was higher after Z-drug overdose (adjusted hazard ratio [aHR], 2.918; 95% CI, 1.928–4.416). Recurrence proportions were similar (25.5% vs 23.3%; log-rank P = 0.426), but timing differed. The recurrence Cox model violated the proportional-hazards assumption; the adjusted HR of 1.266 is descriptive, not a constant relative risk. Among 58 recurrent cases, standard single-dose (SSD) multiples declined and documented intent shifted toward sleep-related dose escalation; this pattern may partly reflect assessment or documentation differences.
Conclusion: Z-drug overdose was associated with more recorded behavioral-management events and a higher hazard of first documented psychiatric diagnosis, whereas BZD overdose was associated with more severe supportive-care outcomes. Initial care should integrate mental-status and respiratory monitoring with behavioral and self-harm assessment. Follow-up should address insomnia, medication safety, and mental health. Recurrence timing and intent shifts were documented associations, not causal effects.
Related
- Post-traumatic neurogenic thoracic outlet syndrome: Clinical presentation and surgical outcomes–A retrospective cohort study
- Primary non-adherence to newly prescribed discharge medications and 30-day adverse outcomes by mental health status: a retrospective cohort study in two tertiary care hospitals in Montréal, Québec, Canada
- Comparative outcomes of timely versus delayed presentation of pediatric fractures in post-conflict Iraq: a multicenter retrospective cohort study
- Real-world Longitudinal Patterns of Benzodiazepine Use in Korean Patients with First-onset Panic Disorder: A Multicenter Retrospective Study
- Multidimensional tumor anatomy, intentional incomplete resection, and long-term outcomes in cervical dumbbell tumors: an exploratory cohort study
BibTeX
@article{liu2026from,
title = {{From acute presentation to post-discharge outcomes: intentional Z-drug versus benzodiazepine overdose in a retrospective cohort}},
author = {Qinqin Liu and Chao Wu and Liang Gu and Beibei Huang and Jie Lu and Yao Tan and Donglin Zhu and Ming Sun},
journal = {Frontiers in Psychiatry},
year = {2026},
month = nov,
doi = {10.3389/fpsyt.2026.1991713},
url = {https://www.frontiersin.org/articles/10.3389/fpsyt.2026.1991713/full},
note = {CC BY 4.0},
}
RIS
TY - JOUR
TI - From acute presentation to post-discharge outcomes: intentional Z-drug versus benzodiazepine overdose in a retrospective cohort
AU - Qinqin Liu
AU - Chao Wu
AU - Liang Gu
AU - Beibei Huang
AU - Jie Lu
AU - Yao Tan
AU - Donglin Zhu
AU - Ming Sun
PY - 2026
DA - 2026/11/09
JO - Frontiers in Psychiatry
DO - 10.3389/fpsyt.2026.1991713
UR - https://www.frontiersin.org/articles/10.3389/fpsyt.2026.1991713/full
AB - Background: Z-drugs are alternatives to benzodiazepines (BZDs) for insomnia, but overdose safety remains uncertain. We compared acute and post-discharge outcomes after intentional Z-drug and BZD overdose. Methods: This retrospective cohort included 239 adults (110 Z-drug; 129 BZD) treated at two emergency departments without flumazenil (2018–2024). Models adjusted for age, sex, dose category, intent, and center. Results: Compared with BZD overdose, Z-drug overdose was associated with more recorded physical restraint use (adjusted odds ratio [aOR], 6.974; 95% confidence interval [CI], 3.358–14.483), more recorded patient-behavior-related medical interruption events (PB-MIEs; aOR, 5.955; 95% CI, 2.732–12.980), and a longer emergency-department stay (adjusted median difference, 2.516 h; 95% CI, 2.056–2.845), but fewer intensive care unit (ICU) admissions (aOR, 0.092; 95% CI, 0.031–0.272). Among admitted patients, total stay was longer after Z-drug overdose (228 vs 146 h, P < 0.001). The hazard of a first documented anxiety and/or depressive diagnosis was higher after Z-drug overdose (adjusted hazard ratio [aHR], 2.918; 95% CI, 1.928–4.416). Recurrence proportions were similar (25.5% vs 23.3%; log-rank P = 0.426), but timing differed. The recurrence Cox model violated the proportional-hazards assumption; the adjusted HR of 1.266 is descriptive, not a constant relative risk. Among 58 recurrent cases, standard single-dose (SSD) multiples declined and documented intent shifted toward sleep-related dose escalation; this pattern may partly reflect assessment or documentation differences. Conclusion: Z-drug overdose was associated with more recorded behavioral-management events and a higher hazard of first documented psychiatric diagnosis, whereas BZD overdose was associated with more severe supportive-care outcomes. Initial care should integrate mental-status and respiratory monitoring with behavioral and self-harm assessment. Follow-up should address insomnia, medication safety, and mental health. Recurrence timing and intent shifts were documented associations, not causal effects.
SN - 1664-0640
ER -