Reliability and validity of proxy reports of impulsivity and aggression: An evidence-based assessment approach to psychological autopsy methods.

Abstract

Introduction: Psychological autopsy methods often include measures of impulsivity and aggression. The aim is to assess their reliability and validity in a Spanish sample.

Methods: Cross-sectional web-based survey was fulfilled by 184 proband and proxy pairs. Data was collected on sociodemographic characteristics, impulsivity through Barratt Impulsiveness Scale (BIS-11), aggression through Buss-Perry Aggression Questionnaire (BPAQ), and history of suicide ideation. Proxies filled out BIS-11, BPAQ and suicide ideation with the responses they would expect from the probands. Reliability was assessed using intraclass correlation coefficients (ICC) between proband and proxies. Logistic regression analysis was performed to assess the predictive validity of proxy reports in predicting probands' suicide ideation.

Results: Bivariate analysis showed differences in BPAQ (Median 68 vs. 62; p=0.001), but not in BIS-11 (p>.050). BIS-11 showed good concordance (ICC=0.754; CI 95% 0.671-0.816) and BPAQ acceptable (ICC=0.592; CI 95% 0.442-0.699). In the probands regression model BPAQ predicted suicide ideation (OR 1.038; CI 95% 1.016-1.061) but not BIS-11 (OR 0.991; CI 95% 0.958-1.025). In the proxy-report model BPAQ also predicted probands' suicide ideation (OR 1.036; CI 95% 1.014-1.058) but not BIS-11 (OR 0.973; CI 95% 0.942-1.004).

Conclusion: Used as proxy-reported assessment tools, BIS-11 showed better reliability than the BPAQ. However, both showed validity in Spanish population and could be included in psychological autopsy protocols.