Article type
Year
Abstract
Introduction/Objective: Developing a reliable and easy to use checklist for quality in psychiatric studies.
Methods: About 20 general check-lists and rating scales were reviewed. It was found that several aspects, specific for psychiatry, were missing. The use of weighted scores on some items have not been validated. On the basis of these considerations, a preliminary checklist was constructed. This was tested by 25 Scandinavian research psychiatrists and psychologists. Using a Delphi technique, the checklist was modified in several steps. The check-list is adopted for use with the reference managing program Endnote and Excel, so that quality scoring can be performed simultaneously with the coding of effect measures and demographic data.
Results: The checklist has been found simple to use and inter-rater reliability has consistently been satisfactory (kappa > 0.7).
Discussion: The use of a more comprehensive check-list, in contrast with recent recommendations of only checking randomization procedure, blinding procedure and way of accounting for attrition, is valuable. This is so if the check-list is used not so much as a mean for excluding studies but more as a help in structuring reading and as a basis for different sub-group analysis's. It is possible that a general checklist, applicable to all areas of health care, is not an attainable goal.
Methods: About 20 general check-lists and rating scales were reviewed. It was found that several aspects, specific for psychiatry, were missing. The use of weighted scores on some items have not been validated. On the basis of these considerations, a preliminary checklist was constructed. This was tested by 25 Scandinavian research psychiatrists and psychologists. Using a Delphi technique, the checklist was modified in several steps. The check-list is adopted for use with the reference managing program Endnote and Excel, so that quality scoring can be performed simultaneously with the coding of effect measures and demographic data.
Results: The checklist has been found simple to use and inter-rater reliability has consistently been satisfactory (kappa > 0.7).
Discussion: The use of a more comprehensive check-list, in contrast with recent recommendations of only checking randomization procedure, blinding procedure and way of accounting for attrition, is valuable. This is so if the check-list is used not so much as a mean for excluding studies but more as a help in structuring reading and as a basis for different sub-group analysis's. It is possible that a general checklist, applicable to all areas of health care, is not an attainable goal.