The validity of the German version of the Health of the Nation Outcome Scales (HoNOS‐D): a clinician‐rating for the differential assessment of the severity of mental disorders

Abstract

The Health of the Nation Outcome Scales (HoNOS) is an internationally-established clinician-rating instrument for the differential assessment of the severity of patients with mental disorders. The aim of this study was to examine the validity of the German version of the HoNOS (HoNOS-D). Evaluation of validity, including factor validity, convergent and discriminant validity and sensitivity to change, was conducted on a large, virtually representative, clinical sample of patients with mental disorders in inpatient psychotherapy (Study 1, N = 3169). Additional assessment of criterion-based validity was completed using another clinical sample of patients with mental disorders (Study 2, N = 55). Although factor validity of the HoNOS-D and its total score could not be confirmed as expected, the majority of single items of the HoNOS-D proved to be valid in terms of convergent validity, criterion-based validity and sensitivity to change. Hence, single items, rather than the total score of the HoNOS-D, can be recommended for obtaining a picture of the impairment of patients with mental disorders in the clinical setting.

The Concurrent Validity and Sensitivity of Change of the German Version of the Health of the Nation Outcome Scales in a Psychiatric Inpatient Setting

“…In our sample, HoNOS sensitivity to change assessed by rating patients at admission and discharge (table 3) shows limitations in comparison with the CGI. Comparable to Audin et al [32] and Andreas et al [16], items 7 and 10 showed medium values, and the majority of the items and the HoNOS total score proved to be moderately sensitive to change, although our sample differs from their sample with regard to 37% of the participants having a schizophrenia diagnosis. Item 11 (‘residential problems’) has to be understood in the context of the average duration of treatment (30 days), changes in this timeframe not being likely.…”

“…It was suggested that the scale could be subdivided into 4 subscales (‘behavioral problems’, ‘impairment’, ‘symptomatic problems’ and ‘social problems’). However, there is not really much evidence that subscale scores are reliable [14,16,25,26]. Therefore, we only used the HoNOS-D main score and single items for further calculations.…”

“…Psychometric properties were found to be satisfactory for most of the items, but the authors of a review conclude that validity in psychiatric routine care warrants closer examination [14]. Firstly, we tried to replicate the findings from evaluations of the validity of HoNOS and the German version of HoNOS (HoNOS-D) [11,15,16] by examining its association with the CGI as an established routine outcome measure. Secondly, we tested the concurrent validity concerning psychopathology using the Association for Methodology and Documentation in Psychiatry (AMDP) rating system [17,18,19,20,21].…”

Is the Health of the Nation Outcome Scales appropriate for the assessment of symptom severity in patients with substance-related disorders?

“…First, further specific external criteria for the validation of single items were not included in this study (e.g., Brief Psychiatric Rating Scale [Overall & Gorham, 1976] or Addiction Severity Index [Mäkela, 2004;McLellan et al, 1992]). However, in another study, we examined the validity of the German version of the HoNOS in inpatients with a widespread of mental disorders (Andreas et al, 2010). Evaluation of validity, including factor validity, convergent and discriminant validity, and sensitivity to change, was conducted on a large, virtually representative, clinical sample of patients with mental disorders in inpatient psychotherapy (Study 1, N = 3,169).…”

“…In line with international developments and due to the lack of suitable national clinician-rating instruments for the differential assessment of severity, the HoNOS was translated into German with authorization from the test authors. The translation process and first empirical results on the practicability, reliability, and validity have been previously described elsewhere (Andreas et al, 2007;Andreas et al, 2010).…”