Send to

Choose Destination
Lancet Psychiatry. 2016 Apr;3(4):358-66. doi: 10.1016/S2215-0366(15)00575-1. Epub 2016 Feb 25.

Predictors of treatment resistance in patients with schizophrenia: a population-based cohort study.

Author information

National Centre for Register-based Research, School of Business and Social Sciences, Aarhus University, Aarhus, Denmark. Electronic address:
Section of Biostatistics, Department of Public Health, Aarhus University, Aarhus, Denmark.
The Lundbeck Foundation Initiative for Integrative Psychiatric Research, iPSYCH, Aarhus University, Aarhus, Denmark; Research Unit, Mental Health Centre Copenhagen, Faculty of Health Sciences, University of Copenhagen, Gentofte, Denmark.
National Centre for Register-based Research, School of Business and Social Sciences, Aarhus University, Aarhus, Denmark.
Department of Psychosis Studies, Institute of Psychiatry, Psychology and Neuroscience, King's College London, London, UK.
National Centre for Register-based Research, School of Business and Social Sciences, Aarhus University, Aarhus, Denmark; Centre for Integrated Register-based Research, CIRRAU, Aarhus University, Aarhus, Denmark.

Erratum in



Identification of patients at high risk of treatment-resistant schizophrenia at the time of schizophrenia diagnosis would be of great clinical benefit in minimising the delay to clozapine treatment in patients unlikely to respond to non-clozapine antipsychotics. However, little is known about predictors of treatment resistance in this patient population. We used a treatment-based proxy for treatment-resistant schizophrenia to identify candidate predictors of treatment resistance at first hospital contact with a schizophrenia diagnosis.


In this population-based cohort study, we obtained Danish national registry data for all adult patients (≥18 years) with incident schizophrenia diagnosed between Jan 1, 1996, and Dec 31, 2006, and followed up until Dec 31, 2010. Our main proxy definition of treatment-resistant schizophrenia was the earliest instance of either clozapine initiation or hospital admission for schizophrenia after having had two periods of different antipsychotic monotherapy. We did multivariable Cox proportional hazards regression analysis to estimate the association between baseline candidate predictors and treatment resistance.


8624 patients fulfilled the criteria for inclusion. In multivariable complete-case analyses, 1703 (21%) of 8044 patients fulfilled the main proxy definition of treatment-resistant schizophrenia during a median follow-up of 9·1 years (IQR 6·3-11·9). Younger age (hazard ratio 0·96 [95% CI 0·95-0·97]), living in a less urban area (provincial 1·38 [1·23-1·56], rural 1·44 [1·25-1·65]), primary education level (0·88 [0·79-0·98]), more than 30 bed-days in psychiatric hospital in the year before first schizophrenia diagnosis (1·54 [1·35-1·75]), inpatient at first schizophrenia diagnosis (2·07 [1·87-2·29]), paranoid subtype (1·24 [1·13-1·37]), comorbid personality disorder (1·24 [1·11-1·39]), psychotropic drug use (antipsychotics 1·51 [1·35-1·69], antidepressants 1·15 [1·03-1·29], and benzodiazepines 1·22 [1·10-1·37]), and previous suicide attempt (1·21 [1·07-1·39]) were all significantly associated with treatment-resistant schizophrenia.


Our study identifies several candidate predictors that could potentially be included in future prediction models for treatment-resistant schizophrenia. Notably, established risk factors for schizophrenia did not predict treatment resistance, suggesting that treatment-resistant disease might be a distinct subtype of schizophrenia and not merely a more severe form.


European Community's Seventh Framework Programme.

Comment in

[Indexed for MEDLINE]

Supplemental Content

Full text links

Icon for Elsevier Science
Loading ...
Support Center