Article type
Year
Abstract
Introduction/Objective: The NHS R&D Programme in the UK is supporting the development of a valid and reliable method to critically appraise guidelines. As part of this national study a representative sample of disease specific guidelines was obtained. This paper describes the number, source and quality of clinical guidelines produced in the United Kingdom between January 1991 and June 1995 for the management of coronary heart disease, asthma, breast cancer, lung cancer and depression.
Methods: Guidelines were identified by a postal questionnaire survey between March and June 1995 and electronic searches. A disproportional stratified random sample of sixty guidelines (15 per disease group) was created from the database. Each guideline was assessed by 6 individuals from different professional backgrounds (120 reviewers in total) using a validated and reliable critical appraisal instrument (see other presentation).
Results: Four hundred and seventy two guidelines were identified in the survey (CHD = 205, asthma = 168, breast cancer =45, lung cancer = 9, depression = 45). Most of the guidelines had been produced locally. In the 60 guidelines sampled a comprehensive literature review was undertaken in 13 (22%) of guidelines, and this was considered adequate for only 6 (10%). National guidelines were more likely to be based on a comprehensive search. Methods used for formulating the recommendations (i.e. consensus methods) were adequate for 14 (23%) guidelines, explicit links between evidence and recommendations were lacking in most guidelines, but was more likely in guidelines produced after 1994. Logistic modelling will be used to derive weightings for each item and dimension in order to create summary dimensional scores. Using these results the determinates of "guideline quality" will be identified.
Discussion: There is a proliferation of guidelines in most parts of the UK. The majority of these are developed locally and are unlikely to be evidence based. National guidelines tend to have a more rigorous development process. To ensure that only valid guidelines are promulgated it will be more cost-effective to concentrate on creating a few evidence-based national guidelines that can act as templates, than advocating widespread guideline development. Local efforts should he targeted on guideline dissemination and implementation.
Methods: Guidelines were identified by a postal questionnaire survey between March and June 1995 and electronic searches. A disproportional stratified random sample of sixty guidelines (15 per disease group) was created from the database. Each guideline was assessed by 6 individuals from different professional backgrounds (120 reviewers in total) using a validated and reliable critical appraisal instrument (see other presentation).
Results: Four hundred and seventy two guidelines were identified in the survey (CHD = 205, asthma = 168, breast cancer =45, lung cancer = 9, depression = 45). Most of the guidelines had been produced locally. In the 60 guidelines sampled a comprehensive literature review was undertaken in 13 (22%) of guidelines, and this was considered adequate for only 6 (10%). National guidelines were more likely to be based on a comprehensive search. Methods used for formulating the recommendations (i.e. consensus methods) were adequate for 14 (23%) guidelines, explicit links between evidence and recommendations were lacking in most guidelines, but was more likely in guidelines produced after 1994. Logistic modelling will be used to derive weightings for each item and dimension in order to create summary dimensional scores. Using these results the determinates of "guideline quality" will be identified.
Discussion: There is a proliferation of guidelines in most parts of the UK. The majority of these are developed locally and are unlikely to be evidence based. National guidelines tend to have a more rigorous development process. To ensure that only valid guidelines are promulgated it will be more cost-effective to concentrate on creating a few evidence-based national guidelines that can act as templates, than advocating widespread guideline development. Local efforts should he targeted on guideline dissemination and implementation.