Development Quality Criteria to Evaluate Nontherapeutic Studies of Incidence, Prevalence, or Risk Factors of Chronic Diseases
Pilot Study of New Checklists
Published in: Journal of Clinical Epidemiology, v. 64, no. 6, June 2011, p. 637-657
Posted on RAND.org on June 01, 2011
OBJECTIVE: To develop two checklists for the quality of observational studies of incidence or risk factors of diseases. STUDY DESIGN AND SETTING: Initial development of the checklists was based on a systematic literature review. The checklists were refined after pilot trials of validity and reliability were conducted by seven experts, who tested the checklists on 10 articles. RESULTS: The checklist for studies of incidence or prevalence of chronic disease had six criteria for external validity and five for internal validity. The checklist for risk factor studies had six criteria for external validity, 13 criteria for internal validity, and two aspects of causality. A Microsoft Access database produced automated standardized reports about external and internal validities. Pilot testing demonstrated face and content validities and discrimination of reporting vs. methodological qualities. Interrater agreement was poor. The experts suggested future reliability testing of the checklists in systematic reviews with preplanned protocols, a priori consensus about research-specific quality criteria, and training of the reviewers. CONCLUSION: We propose transparent and standardized quality assessment criteria of observational studies using the developed checklists. Future testing of the checklists in systematic reviews is necessary to develop reliable tools that can be used with confidence.
This report is part of the RAND Corporation External publication series. Many RAND studies are published in peer-reviewed scholarly journals, as chapters in commercial books, or as documents published by other organizations.
The RAND Corporation is a nonprofit institution that helps improve policy and decisionmaking through research and analysis. RAND's publications do not necessarily reflect the opinions of its research clients and sponsors.