TY - JOUR T1 - Formal consensus and consultation: a qualitative method for development of a guideline for dementia. JF - Quality in Health Care JO - Qual Health Care SP - 192 LP - 199 DO - 10.1136/qshc.7.4.192 VL - 7 IS - 4 AU - H Trickey AU - I Harvey AU - G Wilcock AU - D Sharp Y1 - 1998/12/01 UR - http://qualitysafety.bmj.com/content/7/4/192.abstract N2 - OBJECTIVES: To use a formal consensus method--the nominal group technique--for developing a clinical practice guideline to improve the diagnosis, needs assessment, and management of dementia at the primary/secondary care interface. To assess the usefulness of this method in an area where there is a dearth of clinical trials thus precluding an evidence-based approach. DESIGN: A qualitative and iterative method was used. The guideline was developed by an expert panel using the nominal group technique, incorporating the deliberations of a Primary Care Professionals' Task Group, recent reviews in the scientific literature, information from semistructured interviews with local specialists, and a local questionnaire survey. SETTING: Mainly Bristol but incorporating other parts of the United Kingdom especially the south and west. SUBJECTS: Expert panel group (including users and non-statutory organisations), local general practitioners (GPs), professionals carrying out the health check for people over 75, local geriatricians and psychogeriatricians, and primary care professionals with an interest in dementia. MAIN MEASURES: Items important in dementia from the nominal group in response to three questions on diagnosis, needs assessment, and management, ranked into a series of top 10 items for each area and allocated a score according to relative importance to each member; summary outputs from the questionnaire survey, semistructured interviews, and primary care professionals' task group. RESULTS: A clinical practice guideline covering diagnosis, needs assessment, and management of dementia in primary care was produced in the form of an algorithm and a management sheet. CONCLUSIONS: A formal consensus method can be used as an acceptable alternative to the evidence-based approach when developing guidelines in situations in which evidence is scarce, in which the guideline is intended as an aid in linking different stages of care, and in which practical and political considerations--such as the links between agencies--are important. ER -