The production of a major SBU report, covering an entire field of medicine, such as dementia or diabetes, is a substantial commitment. In principle, reports which are limited to assessment of a single new method, follow the same routines.
The SBU Board of Directors prioritises which topics are to be assessed. At the start of the project, the project group determines the scope of the enquiry, ie which issues are to be included and excluded.
The project group agrees the standards to be applied to the quality of evidence presented by the studies. Only the results of studies which are considered sufficiently rigorous will be included. SBU may occasionally also comment on studies of lesser quality, if the results contribute other valuable information.
A systematic search is conducted for all available research results on the topic, in computerised databases and in reference lists in specialist journals and scientific reports. In the assessment of a broad field, such as all methods for treating obesity or chronic pain, these literature searches yield a large number of articles to be assessed by the group.
The project group reviews research reports using specially designed appraisal forms and assesses the strength of the evidence of each individual study. During this process, research reports that do not meet the predetermined criteria for evidence strength are screened out.
The results of the selected studies are aggregated, tabulated and analysed as evidence. As with the literature search and selection, aggregation requires a systematic and rigorous method. The conclusions must be supported by scientific evidence. Not only are the medical effects of different interventions addressed, so too is current established practice in Sweden, as well as health economic, social and ethical aspects.
Before SBU publishes its conclusions, the text is reviewed by external experts as well as experts on the SBU Scientific Advisory Committee or the Alert Advisory Board. The SBU Board of Directors and Advisory Committee approve the conclusions of every report. Where possible, the strength of the evidence for each individual result is presented, ie the results are graded according to the international GRADE system, which discloses the strength of the scientific support underlying the outcome