Citation

Verdolini N, Pacchiarotti I, Köhler CA, Reinares M, Samalin L, Colom F, Tortorella A, Stubbs B, Carvalho AF, Vieta E, Murru A. J. Affect. Disord. 2019; ePub(ePub): ePub.

Affiliation

Bipolar Disorder Unit, Institute of Neuroscience, Hospital Clinic, University of Barcelona, IDIBAPS, CIBERSAM, 170 Villarroel st, 12-0, 08036, Barcelona, Catalonia, Spain; CIBERSAM, Centro Investigación Biomédica en Red Salud Mental, Barcelona, Spain.

Copyright

(Copyright © 2019, Elsevier Publishing)

DOI

10.1016/j.jad.2019.10.022

PMID

31668994

Abstract

The authors regret that an error occurred in reporting the number of individuals who presented violent criminal behaviors in the general population sample of the article by Webb and colleagues (Webb et al., 2014). As a consequence, statistical analyses have been re-run and results changed accordingly.

Since the study by Fazel and colleagues (Fazel et al., 2007) is a study of a selected subgroup of violent offenders, namely those with sexual offences, we decided not to include it in the new analyses as it might lead to a count of 0 for violent crimes in the control group and a following biased odds ratio.

The following changes in the text have been done.

1. In the section 2.5 Statistical Analysis, the sentence “When evidence of publication bias was observed, ES estimates were adjusted with the trim-and-fill procedure (Fazel et al., 2007)” has been eliminated.

2. The correct text for the section 3. Results is:

“The database search generated 1,031 hits, 46 articles were identified after searching the references of included articles. After duplicates removal, the title/abstracts of 773 references were screened for eligibility; of those, 682 were excluded. Full-texts of 91 references were scrutinized in detail for eligibility, with 80 excluded (see Table S1 in the supplementary material for reasons), and 11 references included for systematic review and meta-analysis (see flowchart in Fig. 1, characteristics in Table S2). Overall, data from 9,505,095 participants (56,023 patients with BD, 9,233,772 general population controls and 215,300 patients with any psychiatric disorder) were included. Participants presenting any VCB were 91,545. Studies have followed cross-sectional (k=4), case-control (k=2), or prospective (k=5) designs. The mean % of criteria met in the NOS scale across studies was 81.8 (SD=18.1) (Table S3). A half of the included studies had a poor methodological quality. Nine out of eleven of the included studies used representative samples and only four specified in the methodology a diagnostic assessment with a structured interview ...

Language: en