Quality assessment of clinical practice guidelines for mammography screening in women aged 40 49 years

Article type
Authors
Burda B1, Norris S1, Holmer H2, Ogden L1, Beth-Smith M3
1Dept. of Medical Informatics & Clinical Epidemiology, Oregon Health & Science University, Portland, Oregon, United States
2Medical Informatics and Clinical Epidemiology, Oregon Health & Science University, Portland, Oregon, United States
3Dept. of Medicine, Oregon Health & Science University, Portland, Oregon, United States
Abstract
Background: With the recent controversy surrounding mammography screening, the quality of clinical practice guidelines (CPGL) is critical. Objectives: To assess the quality of CPGL for mammography screening in women aged 40 49 years. Methods: We searched the National Guideline Clearinghouse and Medline for CPGL published from 2005 to 2010, including guidelines that provided recommendations on the frequency of mammography screening in asymptomatic women aged 40 49 years at average risk. Five independent reviewers rated the quality of the guidelines and underlying evidence reviews using AGREE and AMSTAR instruments, respectively. Results: Eleven guidelines fulfilled inclusion criteria. Ten were appraised; one provided insufficient data for assessment. All guidelines referenced an underlying review; two referenced the same review. Three reviews were rated good, one wasmoderate, fivewere poor quality. Most reviews did not provide a list of included or excluded studies, describe the methods used to combine the findings, or assess publication bias. Many failed to state conflicts of interest or inclusion of grey literature. On overall assessment of the guidelines, two were strongly recommended, one was recommended, and seven were not recommended or the reviewers were unsure. Most guidelines clearly presented their scope and recommendations, but the rigor of development, applicability, and stakeholder involvement varied considerably across CPGL. Eight guidelines recommended routine screening in the target population and two recommended against such screening. The latter two guidelines had good quality reviews and were strongly recommended. Conclusions: CPGL on mammography screening in the target population were of variable quality and the majority were not recommended by reviewers. The underlying evidence reviews were of variable quality and half were rated of poor quality. The two guidelines recommending against routine screening, were of good quality and strongly recommended by the reviewers.