Yield and Efficiency of Mental Disorder Screening at Intake to Prison: A Comparison of DIA-X Short- and Long-Screening-Protocols in Compensation Prisoners

Background: Inmates are several times more likely to suffer from mental disorders than the general population.In order to take appropriate curative or preventive measures, a precise psychiatric diagnosis at detention start would therefore be imperative, but is frequently not carried out for reasons of time. The computer-aided expert system DIA-X enables a rapid and reliable diagnosis of psychiatric disorders. DIA-X is available as a short screening questionnaire with a processing time of a few minutes and as a standardized interview, which takes ~1 h to complete. Objective: The aim of this study was to assess the efficiency and accuracy of the DIA-X short screening questionnaire. Methods: One hundred detainees were recruited randomly from compensation prisoners, who were imprisoned because they were unwilling or unable to pay a fine for committing a criminal offence, from the penal institution Berlin-Plötzensee in 2017. Both the short screening questionnaire and the standardized interview from the DIA-X expert system were used for diagnosing mental disorders. Based on the results of the standardized interview from four study populations of compensation prisoners from 1999, 2004, 2010, and 2017, the sensitivity, specificity and the predictive values of the screening form were inferred. Results: More than half of the compensation prisoners suffered from mental and behavioral disorders caused by the abuse of alcohol or psychoactive substances. Phobic anxiety disorders were detected in one out of ten compensation prisoners and two out of ten compensation prisoners suffered from major depressive disorders. The DIA-X screening questionnaire was able to detect all mental illnesses with a sensitivity of 100%. However, specificities were low for nicotine dependency, drug and alcohol abuse. High specificities and high predictive values were obtained for psychoses and anxiety disorders. Conclusions: As the main test quality criteria of the DIA-X screening forms were so low, we cannot recommend the application of the DIA-X screening form for obtaining a valid diagnosis. Therefore, we explicitly recommend using the long form DIA-X for the detection of the most serious cases of mental illness. Then, these prisoners could receive either therapy or special social training.