American College of Chest Physicians consensus statement on the respiratory health effects of asbestos. Results of a Delphi study.

BACKGROUND The diagnosis of and criteria for the evaluation of asbestos-related disease impairment remains controversial after decades of research. Assessing agreement among experts who study pneumoconiosis, and diagnose and treat patients with asbestos-related respiratory conditions may be the first step in clarifying clinical and forensic/administrative issues associated with asbestos-related pulmonary conditions. METHODS We conducted a Delphi study, an iterative method of obtaining consensus among a group of experts. An expert panel was identified using an objective, nonbiased algorithm, based on the number of asbestos-related disease publications authored during the preceding 10-year period. Identified experts were invited to participate by accessing an Internet site. Each expert was presented statements developed by the authors regarding the diagnosis or treatment of asbestos-related disease; experts then ranked their degree of agreement or disagreement utilizing an 11-level modified Likert scale for each statement. Each expert was asked to justify their selection and to suggest references in support of their opinion. The Wilcoxon signed rank test and the interquartile range were used to define "consensus." The results of the collective Likert rankings, deidentified comments, and suggested references as well as the initial consensus results were then provided to the participating experts. Each panel member then ranked their extent of agreement with a modified statement for which consensus was not achieved. The process was repeated three times. RESULTS Consensus was achieved on all but 9 of 32 statements. CONCLUSIONS Consensus was not achieved for nine statements. These statements may be topics for future research.