A discrete choice experiment to identify the most efficient quality indicators for the supervision of psychiatric hospitals

Pieter van Dijk*, Ron Schellings, Brigitte A. B. Essers, Alfons G. Kessels, Ian Leistikow, Maurice P. Zeegers

*Corresponding author for this work

Research output: Contribution to journalArticleAcademicpeer-review


Background In the Netherlands, health care is regulated by the Health and Youth Care Inspectorate. Forty-six indicators are used to prioritize supervision of psychiatric hospitals. The objective of this study is to define a smaller set of weighted indicators which reflects a consensus among inspectors about which aspects are most important for risk assessment. Methods The set of 46 indicators, complemented with missing information, was reduced to six indicators by means of interviews, group discussions and ranking among the inspectors. These indicators were used as attributes in a discrete choice experiment (DCE) to define their weights. Results Twenty-six inspectors defined the top four indicators suitable for the risk assessment of psychiatric hospitals. These are: the policy on prevention of compulsory treatment; the policy on dysfunctional professionals; the quality of internal research after a serious incident; and the implementation of multidisciplinary guidelines on suicidal behaviour. These indicators share the same importance with regard to risk assessment. The screening of somatic symptoms and the policy on integrated care are important indicators too, but less relevant. Conclusion Through a DCE, we reduced the amount of information for risk assessment of psychiatric hospitals to six weighted indicators. Inspectors can use these indicators to prioritize their inspections.

Original languageEnglish
Article number192
Number of pages6
JournalBMC Health Services Research
Issue number1
Publication statusPublished - 12 Mar 2020


  • Risk assessment
  • Discrete choice experiment
  • Health care regulatory agencies
  • Indicators
  • Mental health care
  • Quality and safety
  • Risk-based supervision

Cite this