Symptom validity in expert assessment of mental disorders - Prevalence rate of negative-response bias and validation of a German translation of the Structured lnterview of Reported Symptoms (SIRS)

Project No. FF-FR 0223

Status:

completed 08/2016

Aims:

  • Focussing in the first sub-project on the retrospective base rate estimation of malingering in psychological expert evaluations and reviewing of the used criteria.
  • Comparing the expert rating with scientific recommended indicators. The value of available symptom validity tests (SVT) in Germany, especially for mental disorders, does not correspond to available English state of the art instruments.
  • Validating a German Version of the SIRS and testing its practicability in the second sub-project.
  • Evaluation of the classification rating of the SIRS and comparing with the value of conventional

Activities/Methods:

A German forward-backward translation of the SIRS was done and evaluated in test persons with mental disorders (n = 100) using a between-subjects simulation design. Additionally, conventional symptom validity tests (SIMS, WMT) were assessed to evaluate their converging and discriminating aspects. Furthermore, a retrospective analysis of psychological expert evaluations from 2002 to 2013 (approx. n = 1000) were performed. The base rate of possible malingering and its indicators have been assessed and reviewed.

Results:

In sub-project 1, the retrospective analysis of 1175 forensic psychological evaluations has shown a grown effort over the course of 16 years, together with the increasing length and complexity of the evaluations. With the use of increasingly complex methods, inconsistencies are reported more frequently and a higher rate of feigning and malingering is noted. A uniform multi-methods approach guarantees that there does not develop any bias in the assessment of this aspect across various evaluators. In agreement with existing estimations in the literature, the analyzed pool of evaluations shows many individual inconsistencies (40.9%). However, these should not be interpreted as malingering; further assessment and interpretation of these inconsistencies is necessary. A resulting rate of an overall rate of 15.8% of malingering matches recent reviews that call into question estimations which had yielded substantially higher rates. This represents the first estimation of a base rate of malingering in evaluation contexts for workers’ compensation claim samples which goes beyond individual indicators or test scores. Used as an indicator of feigning (in 79.6% of the cases), symptom validity tests (SVT) serve as an important decision-making tool for psychological evaluators. For the overall assessment, however, other possible inconsistencies are taken into account. Among the common SVTs, negative response bias is especially frequent in screening measures (47.2%). Yet overall, they seldom reach the level of deliberate, purposeful manipulations (8.2%).

For sub-project 2, we were able to validate a German version of the SIRS-2 and thus, to assess for the first time in German-speaking contexts a SVT that was designed as a standardized fully structured interview and allows for a specific analysis of response styles. Despite the particularities of the evaluated population of workers‘ compensation claim samples (high rates of somatic co-morbidities), the measures yielded in the performed simulation design large effect sizes (daverage = 2.36) for the discrimination between genuine and feigned response styles, high inter-rater reliability (raverage = 0.99), and stable classifications in the retest (medium concordance of 96.4%). For application in practice, sensitivity values of up to 93.2% and specificity values of up to 97.9% can be assumed. With this, this measure is superior to tests classified as specific and used up till now (WMT). Notably, the number of false positives of the analyzed sample was minimal (1%). Due to its time-consuming and complex application, it is recommendable to precede this test with a screening measure (SFSS). A further step will be its publication with a test publisher in order to make the measure available to practitioners and for testing it with real-life populations. This project adds to recent developments and enhances the diagnostic standards for a central aspect of forensic psychological evaluation.

Last Update:

21-Jun-2017

Project

Financed by:
  • Deutsche Gesetzliche Unfallversicherung e. V. (DGUV)
Research institution(s):
  • BG Kliniken Halle
Branche(s):

-cross sectoral-

Type of hazard:

-various

Catchwords:

rehabilitation

Description, key words:

Symptom validity, expert assessment of mental disorders, Prevalence rate of negative-response bias, Structured Interview of Reported Symptoms (SIRS)