BACKGROUND: In Dutch mental health care there is an ongoing debate about the benefits of rom and utility of benchmarking. Opinions vary regarding the reliability and validity of performance indicators.<br/> AIM: Investigation of the reliability of the main indicator of Foundation Benchmark Mental Health Care (sbg), Delta-T, the indicator of treatment outcome.<br/> METHOD: The reliability was established with two indices: the intraclass correlation coefficient (icc) for the agreement between repeated assessments of average treatment outcome and the consistency in rank order of mental health care providers over time.<br/> RESULTS: The reliability of Delta-T proved to be excellent.<br/> CONCLUSION: Reliability is a basic requirement, but only the first step in establishing the utility of Delta-T. Further investigation of its validity is ongoing, especially on how robust treatment outcome is for bias due to instrumentation, selection, and confounding by casemix composition. Ultimately, the usefulness of treatment outcome as indicator of quality of care needs to be demonstrated in practice.
BACKGROUND: In Dutch mental health care there is an ongoing debate about the benefits of rom and utility of benchmarking. Opinions vary regarding the reliability and validity of performance indicators.<br/> AIM: Investigation of the reliability of the main indicator of Foundation Benchmark Mental Health Care (sbg), Delta-T, the indicator of treatment outcome.<br/> METHOD: The reliability was established with two indices: the intraclass correlation coefficient (icc) for the agreement between repeated assessments of average treatment outcome and the consistency in rank order of mental health care providers over time.<br/> RESULTS: The reliability of Delta-T proved to be excellent.<br/> CONCLUSION: Reliability is a basic requirement, but only the first step in establishing the utility of Delta-T. Further investigation of its validity is ongoing, especially on how robust treatment outcome is for bias due to instrumentation, selection, and confounding by casemix composition. Ultimately, the usefulness of treatment outcome as indicator of quality of care needs to be demonstrated in practice.