A semi-structured clinical interview for psychosis sub-groups (SCIPS): development and psychometric properties

Background Clinical sub-groups of schizophrenia, namely drug related, traumatic, anxiety and stress sensitivity sub-types, have been proposed for use in research, training and practice. They were developed on the basis of clinical observation but have not yet been used in research or clinical practice to any great extent. Aims To develop a semi-structured clinical interview for psychosis sub-groups (SCIPS) and determine the best diagnostic criteria with the highest inter-rater reliability, test-retest reliability and concurrent validity for sub-grouping patients with schizophrenia according to a newly developed classification scheme. Methods The SCIPS was developed based upon discussion with the clinician researchers who had developed and were using the sub-groups. Kappa coefficients were calculated between two independent diagnostic assessments with the SCIPS (for inter-rater reliability and test-retest reliability, n = 20) and between the SCIPS diagnosis and the sub-groupings as determined independently with highest achievable validity ( for concurrent validity, n = 21) for patients with schizophrenia. These inter-rater reliability and concurrent validity were compared among five different sets of diagnostic criteria to determine which was most reliable and valid. Results A set of diagnostic criteria with the highest inter-rater reliability and concurrent validity was determined. Kappa coefficients (95% confidence interval) for the inter-rater reliability and concurrent validity were 0.93 (0.66-1.20) and 0.73 (0.47-1.00), respectively, with these diagnostic criteria. Conclusions The SCIPS is a promising tool with which to sub-group patients with schizophrenia according to this recently developed classification scheme. The semi-structured interview achieves acceptable inter-rater and test-retest reliability and concurrent validity.