Related Experiment Video
Updated: Jan 17, 2026
Development of a Virtual Reality Assessment of Everyday Living Skills
Published on: April 23, 2014
The Mini-CAARMS: Development and Validation of a Short Version of the Comprehensive Assessment of AT Risk Mental
Alberto Stefana1,2, Dominic Oliver3,4,5,6, Andrés Estradé6
1Department of Clinical and Experimental Sciences, University of Brescia, Brescia, Italy.
Background And Hypothesis:
Accurate detection of individuals at risk for psychosis with established psychometric instruments such as the Comprehensive Assessment of At-Risk Mental States (CAARMS/2006) is crucial to implement effective preventive strategies. However, the CAARMS/2006's scalability is limited by its lengthy administration. This study developed and validated a shorter CAARMS/2006 version: the mini-CAARMS.
Study Design:
A total of 490 participants (mean age 23.8 ± 5 years) underwent the full CAARMS/2006 interview (60 items, all individually scored). The development and validation of the mini-CAARMS employed advanced statistical methods: regularized logistic regression with a LASSO penalty within a nested cross-validation framework to retain core items; Cohen's κ, Harrell's C index, sensitivity, specificity, balanced accuracy, positive (PPV) and negative (NPV) predictive values to assess performance; Monte Carlo simulations to assess the 2-year prognostic accuracy for predicting psychosis onset. A sensitivity analysis developed and validated an ultra-mini-CAARMS.
Study Results:
The mini-CAARMS retained only 23 items and demonstrated excellent accuracy compared to the full version: Cohen's κ 0.90; Harrell's C index 0.93, sensitivity 95.6%, specificity 100%, balanced accuracy 97.8%, PPV 100%, and NPV 86.3%. The prognostic sensitivity and specificity of the mini-CAARMS were 82.2% and 55.1%, respectively. Sensitivity analyses further developed and validated an ultra-mini-CAARMS (based on 12 items only), which achieved comparable performance, with κ 0.90, C index 0.91, sensitivity 94.4%, balanced accuracy 97.2%, PPV 100%, NPV 82.2%, and prognostic sensitivity and specificity 81.2% and 55.1%, respectively.
Conclusions:
The mini-CAARMS maintains high accuracy while substantially reducing administration time, and it could facilitate scalability of preventive psychiatry in low-resource settings.
