Screening tool helpful in detecting treatment needs in COPD
Click Here to Manage Email Alerts
Key takeaways:
- Sensitivity to detect symptoms of COPD and moderate/severe exacerbations was “good” with use of a questionnaire and peak expiratory flow.
- Negative predictive value was also “good” with use of this tool.
Use of CAPTURE plus peak expiratory flow resulted in “good” sensitivity and negative predictive value for detecting patients with COPD in need of treatment, according to study results.
“CAPTURE was designed to identify possible patients with COPD with an FEV1 [less than] 60% predicted in an unselected primary-care population,” Yun Li, MD, of the department of pulmonary and critical care medicine at the First Affiliated Hospital of Guangzhou Medical University in China, and colleagues wrote in American Journal of Respiratory and Critical Care Medicine.
Using data from the COMPASS study, Li and colleagues assessed 1,696 patients (mean age, 65 years; 90% men; postbronchodilator FEV1, 66.5% predicted) with COPD and 307 control individuals (mean age, 60.2 years; 65% men; FEV1/FVC ratio, 0.78) to find out if the CAPTURE tool can detect those in need of treatment due to symptoms of or risk for moderate/severe exacerbations.
After testing the CAPTURE tool in two different ways (just five questions and five questions plus peak expiratory flow [PEF]), researchers observed better sensitivity in detection with use of questions and PEF. Notably, specificity went down by 15% to 20% with use of questions plus PEF vs. CAPTURE questions alone.
For detecting a COPD Assessment Test score of 10 or more vs. below 10 with CAPTURE questions plus PEF, sensitivity was high at 68.5%, specificity was moderate at 64%, positive predictive value was moderate at 47.8% and negative predictive value was high at 80.8%.
When testing the use of questions plus PEF to identify a modified Medical Research Council score of 2 or more vs. less than 2, researchers observed high sensitivity (85.6%) and negative predictive value (95.6%), with moderate specificity (61%) and a low positive predictive value (29.9%).
A high negative predictive value (92.1%) was again observed in detecting one or more moderate exacerbations vs. none in the previous year with use of questions plus PEF, followed by sensitivity (63.5%), specificity (55.6%) and positive predictive value (15.6%).
Lastly, in testing use of questions plus PEF to identify at least one severe exacerbation vs. none, researchers found high sensitivity (70.2%) and negative predictive value (88.8%), moderate specificity (59.4%) and a low positive predictive value (30.3%).
“These high [negative predictive] values give the CAPTURE user confidence that the risk of missing a patient with COPD who needs treatment is low,” Li and colleagues wrote.
Overall, sensitivity of CAPTURE to find patients with COPD in need of treatment due to symptoms or risk for exacerbations was deemed “good” and the negative predictive value was deemed “good to very good” by researchers.
Researchers also evaluated the ability of CAPTURE to identify COPD of different Global Initiative for Chronic Obstructive Lung Disease (GOLD) stages. With the combination of questions and PEF, sensitivity increased as COPD became more severe. This measure was low for GOLD I (17.1%) but high for GOLD IV (95.9%).
Across each GOLD stage, the positive predictive value was 87% or higher using questions plus PEF, with the highest value found for GOLD II (98.1%). Use of CAPTURE questions and PEF also contributed to an elevated negative predictive value for more severe GOLD stages (GOLD III, 92.5%; GOLD IV, 98.7%); however, this was not the case for less severe GOLD stages (GOLD I, 44.9%; GOLD II, 43.6%).
Researchers noted high specificity (96.4%) and positive predictive value (98.8%) with use of questions plus PEF in the total cohort of patients with COPD. In contrast, they found moderate sensitivity (54.4%) and low negative predictive value (27.7%).
“It has been shown that individuals who do not meet the CAPTURE criteria for spirometric testing are at low risk of being denied treatment for symptoms or risk of exacerbations,” Li and colleagues wrote. “That is an important property for a case-finding tool.”
Although this study demonstrates how the CAPTURE tool may be useful in managing patients with COPD, the tool requires further study before it should be implemented, according to an accompanying editorial by Barbara P. Yawn, MD, MSc, MSPH, adjunct professor in the department of family and community health at University of Minnesota, and Susan Murray, ScD, professor of biostatistics at University of Michigan.
“[Li and colleagues] introduce the concept of using CAPTURE in resource-poor regions to prioritize those in need of immediate COPD pharmacotherapy (perhaps without diagnostic evaluation; eg, spirometry or lung function testing),” Yawn and Murray wrote. “However, these recommendations should be viewed cautiously because the present study cohort does not reflect a routine practice or resource-poor community population and has not measured the direct impact on patient outcomes that may result from following these recommendations.”