COSMIN review of the PANSS Marder factor solution and other factor models in people with schizophrenia.

The Positive and Negative Syndrome Scale (PANSS) is widely used to assess schizophrenia symptoms. Initially designed with three subscales, Marder et al.´s 5-factor-Model (M5M) first proposed in 1997 has been frequently used in treatment trials, but it has never been systematically reviewed for its measurement properties. We utilized the COnsensus-based Standards for the selection of health Measurement INstruments (COSMIN) guideline for systematic reviews and meta-analytical procedures to assess the psychometric properties of the M5M-PANSS. COSMIN comprises several steps: literature search, risk-of-bias assessments, assessing the updated criteria for good measurement properties, feasibility aspects and grading the quality of the evidence. We further assessed the goodness of fit of other PANSS factor models. We included 95 publications. The M5M-PANSS showed good construct validity, but "insufficient" structural validity. Evidence of other COSMIN domains is largely lacking. Among the multiple (73) factor solutions examined with confirmatory methods, several other 5-factor solutions had better model fit. According to COSMIN rules the M5M should not be recommended for use. Other five-factor models such as the one proposed by Wallwork et al. warrant further evaluation. Nevertheless, the factor composition of the M5M and these other models was relatively similar, so previously published results should not be disregarded.