Twenty-five years of the Positive and Negative Syndrome Scale: an overview of its use in randomized clinical trials

Twenty-five years have been passed since Kay and collaborators developed the Positive and Negative Syndrome Scale (PANSS) for a proper assessment of schizophrenia symptoms [1]. This scale has been largely used in Randomized Clinical Trials (RCTs) to evaluate the efficacy of new therapeutic strategies. Contextually, several modifications of PANSS sub-scales, scores and interview procedures have been proposed to improve the validity of this scale [2,3]. In spite of this effort, however, little research investigated how PANSS is applied in RCTs. In the attempt to address this topic, a PubMed advanced search was used to retrieve all RCTs research articles providing PANSS data that have been published since 1987. Descriptive and inferential statistics were then applied in order to evaluate (i) the completeness of PANSS data provided by the research articles, (ii) the PANSS structure mostly used in RCTs, (iii) the characteristics of the sample population generally enrolled in RCTs. Only research articles describing original results were included in the analysis. The PubMed search indicated that an increasing number of RCTs has adopted PANSS for the assessment of schizophrenia symptoms in the last 25 years, confirming the suitability of this psychometric instrument. However, the meta-analysis of 280 research articles highlighted that PANSS data and procedures are often incompletely described. Particularly, the inter-rater variability and the interview procedures are rarely specified (7.3% and 2.7% of research articles, respectively) and even sub-scale scores are not always provided. Finally, several research articles have recently described PANSS data as a difference or percentage vs. baseline, thus further complicating the interpretation of RCTs results. The three-subscale PANSS structure firstly proposed by Kay is still the most commonly used (76.2%) [1], but a growing number of RCTs has recently adopted one of the newly proposed five-subscale structures in analyzing PANSS data (20.3%) [3]. Only few studies (<1%) have rescaled PANSS scores using the 0-6 scaling, and the semi-structured interview remains the procedure most commonly used to evaluate patient symptoms. The demographic characteristics of the population enrolled in RCTs appear homogenous showing a higher prevalence of males (65.45%) than females (34.55%), and an age of 38.67±0.16 (mean±SEM) years. Patients were generally affected by schizophrenia and schizophreniform disorders and characterized by minimal/mild PANSS total average scores at baseline (86.21±0.07). The unimodal distribution of positive and negative subscale scores does not support the hypothesis that schizophrenia may include two different sub-populations of patients showing a prevalence of positive or negative symptoms, respectively. The present analysis stimulates a discussion on the approach currently used in the clinical research when assessing the efficacy of new therapeutic strategies. PANSS confirms its validity for the assessment of schizophrenia symptoms, but the information potentially provided by this scale is only partially reported in research articles. Furthermore, the attempts to improve the validity of this scale have been poorly adopted in RCTs. Comparative analyses are possibly needed to verify whether the clinical characteristics of individuals enrolled in RCTs may properly reflect those of the general population of patients affected by schizophrenia or schizophreniform disorders. References [1] Kay, S.R., Fiszbein, A,, Opler, L.A., 1987 The positive and negative syndrome scale (PANSS) for schizophrenia. Schizophr. Bull. 13, 261-276. [2] von Knorring, L., Lindström, E., 1995 Principal components and further possibilities with the PANSS. Acta Psychiatr. Scand Suppl.388, 5-10. [3] Wallwork, R.S., Fortgang, R., Hashimoto, R., Weinberger, D.R., Dickinson, D., 2012 Searching for a consensus five-factor model of the Positive and Negative Syndrome Scale for schizophrenia. Schizophr. Res. 137, 246-250.