How accurate are physician assessments of the severity of chronic obstructive pulmonary disease (COPD)?
Using immediate, in-office spirometry results as the reference standard, seasoned physicians accurately identified COPD severity in approximately one in three patients, underestimating severity in 41% of patients and overestimating severity in 29% of patients. This mismatch seems to be important because the physicians participating in this study changed their treatment plans for 37% of patients after reviewing the spirometry results. A second issue in this study: Although most of the physicians in the study had a spirometer in their office, they (or their staff) were unable to get usable spirometry results in 25% of their patients. (Level of Evidence = 1c)
The study included 899 patients with COPD who were randomly selected from the practices of 83 primary care physicians (63% family physicians and 37% general internists). The physicians had been in practice an average of 22 years and most had in-office spirometry available before this study. At one visit, both the physician and the patient rated the patient's pulmonary disease severity at that time on a five-point scale, ranging from 1 (no clinical symptoms or disease impact/mild symptoms) to 5 (very severe). Following this assessment, the patient immediately underwent in-office spirometry, although only 75% were able to produce at least one high-quality result. Overall, there was poor correlation among physician assessment, patient assessment, and spirometry results. Physicians underestimated severity in 41% of patients and overestimated severity in 29% of patients, using the spirometry results as the reference standard. Correlation was not much better with the patients' own estimates, with physicians underestimating severity in 42% of patients and overestimating severity in 18% compared with patients' self-assessments. Overall, physician ratings were accurate for only 30% of patients. More importantly, the physicians in this study recommended treatment changes for 37% of patients after reviewing spirometry results.
Study design: Cross-sectional
Funding source: Industry
Setting: Outpatient (primary care)
Reference: MapelDWDalalAAJohnsonPBeckerLHunterAGA clinical study of COPD severity assessment by primary care physicians and their patients compared with spirometry. Am J Med2015; 128( 6): 629– 637.