Evaluation of the psychometric properties of the Brazilian version of the Short Portable Mental Status Questionnaire (SPMSQ) by Pfeiffer

Abstract

Objective To evaluate the psychometric properties, including validity, reliability, sensitivity, and specificity of the SPMSQ-BR. Method The analysis was conducted on 93 older adults, assessed with the Brazilian version of the SPMSQ, Mini-Mental State Examination (MMSE), Clock Drawing Test (CDT), and Verbal Fluency Test (VFT). Internal consistency was estimated using Cronbach's alpha coefficient. In order to assess intra-rater stability and inter-rater reproducibility, the intraclass correlation coefficient (ICC) with a 95% confidence interval was employed. For concurrent and convergent validity, Spearman's correlation coefficient was utilized. Accuracy in establishing the cutoff point for the SPMSQ-BR to screen sensitivity and specificity was assessed through ROC curve analysis. Result The SPMSQ-BR demonstrated excellent internal consistency (α=0.803), high intra-rater and inter-rater reliability (ICC=0.977 and ICC=0.973), respectively. In terms of concurrent validity, it showed a strong negative linear correlation with the MMSE (-0.799, p<0.001), while moderate correlations were found with the CDT (-0.584, p<0.01) and VFT (-0.569, p<0.01). The SPMSQ-BR proved capable of discriminating cognitive impairment with a cutoff point of 3 errors, following the original scale. The area under the curve (AUC) considering education level was 0.905 (sensitivity=81.3%, specificity=76.9%), and without considering education level, the AUC was 0.927 (sensitivity=87.5%, specificity=80.8%), indicating good sensitivity and specificity in both cases. Conclusion The study demonstrated that the Brazilian version of the SPMSQ, established as a Brief Cognitive Capacity Scale, proved to be stable, reproducible, valid, and reliable for assessing the cognitive capacity of older adults. Furthermore, it showed sensitivity in identifying cognitive impairment.