Article type
Year
Abstract
Background: Generic health-related quality of life measures (HRQL) are claimed to be applicable across disease areas. It is important to have robust methods of reviewing evidence that HRQL measures are valid and responsive in individual conditions.
Objectives: To review the methods of systematically reviewing the literature on the validity and responsiveness of the SF-36, SF-12, SF-6D and EQ-5D in five mental health conditions.
Methods: Systematic reviews were undertaken in schizophrenia, depression, anxiety, bipolar disorder and personality disorders. Ten databases were searched up to August 2009. Studies were appraised and data extracted. A narrative synthesis was performed on construct validity including known groups validity (KGV) (ability to detect differences in HRQL scores between groups), convergent validity (CV) (strength of association between generic HRQL and other related measures, e.g. symptoms or function) and responsiveness (R) (i.e. changes in scores in responders/non-responders to treatment and correlation with changes in related measures).
Results: There were several problems in the review process including determining eligibility at the abstract level and lack of relevant quality assessment criteria. The greatest problem was determining how to synthesize the data from included studies, particularly in relation to the robustness of methods used to assess KGV, CV and R. For example, schizophrenia studies demonstrated crude KGV by detecting a difference between scores of the general population and individuals with schizophrenia. The methods used to assess CV by measuring the correlation between HRQL and clinical measures might not truly indicate its validity or otherwise.
Conclusion: There are no guidelines on how to perform such reviews. However, such reviews are essential to determine if generic HRQL measures are valid for use in cost-effectiveness analyses for all diseases/conditions. Determining the methods that are appropriate to demonstrate KGV, CV and R is essential to developing the methods of systematic reviewing psychometric properties of HRQL measures.
Objectives: To review the methods of systematically reviewing the literature on the validity and responsiveness of the SF-36, SF-12, SF-6D and EQ-5D in five mental health conditions.
Methods: Systematic reviews were undertaken in schizophrenia, depression, anxiety, bipolar disorder and personality disorders. Ten databases were searched up to August 2009. Studies were appraised and data extracted. A narrative synthesis was performed on construct validity including known groups validity (KGV) (ability to detect differences in HRQL scores between groups), convergent validity (CV) (strength of association between generic HRQL and other related measures, e.g. symptoms or function) and responsiveness (R) (i.e. changes in scores in responders/non-responders to treatment and correlation with changes in related measures).
Results: There were several problems in the review process including determining eligibility at the abstract level and lack of relevant quality assessment criteria. The greatest problem was determining how to synthesize the data from included studies, particularly in relation to the robustness of methods used to assess KGV, CV and R. For example, schizophrenia studies demonstrated crude KGV by detecting a difference between scores of the general population and individuals with schizophrenia. The methods used to assess CV by measuring the correlation between HRQL and clinical measures might not truly indicate its validity or otherwise.
Conclusion: There are no guidelines on how to perform such reviews. However, such reviews are essential to determine if generic HRQL measures are valid for use in cost-effectiveness analyses for all diseases/conditions. Determining the methods that are appropriate to demonstrate KGV, CV and R is essential to developing the methods of systematic reviewing psychometric properties of HRQL measures.