A Case-finding Strategy for Moderate-to-Severe COPD in the United States
COPD (chronic obstructive pulmonary disease) was relatively uncommon a few generations ago, but it is now a major public health problem, affecting more than 20 million Americans and causing 5% of all deaths among adults in the U.S. Public education regarding the disease, research on its pathogenesis and treatment and the development of effective therapies have not kept pace with the growth of COPD. As a result, this condition stands out as a prime opportunity for substantial improvements in awareness, scientific knowledge, and patient care over a relatively short period of time.
One deficiency regarding COPD that may be straightforward to address is the problem of under-diagnosis. There are many individuals in the U.S., perhaps several million in number, who have not yet been diagnosed with COPD and for whom disease-specific treatments would be beneficial. Nevertheless, no national program exists for identifying these cases – not even a defined strategy for how to screen for the disease or test individuals at risk. To explore possibilities for how the diagnostic rate in COPD might be improved, the National Heart, Lung, and Blood Institute convened a meeting of investigators on June 9-10, 2008, in Bethesda, Maryland, to discuss case-finding strategies in COPD. Workshop participants reviewed existing data and approaches; considered the benefits, costs, and hazards of alternative strategies; and defined the general features of an algorithm that could be used to efficiently identify a substantial fraction of those individuals with COPD for whom diagnosis of the disease would be of clinical value. Workshop participants recommended that additional research be performed to refine and test the case-finding strategy that they developed. The workshop participants reached consensus on the following recommendations.
Recommendation 1. There is an urgent need to develop and test a strategy for COPD case-finding in the U.S. Although COPD prevalence is not known precisely, there is no doubt that a substantial number of Americans both have undiagnosed COPD and would benefit from proper diagnosis. In addition, there exists sufficient knowledge regarding COPD respiratory questionnaires, and pulmonary function testing, to allow a practical algorithm for COPD case-finding to be developed quickly and move rapidly into large scale testing in real world environments. Demonstrating that a particular algorithm is sensitive and efficient is the most urgent research need, since that knowledge is necessary for policy makers and insurers to evaluate the costs and benefits of implementing a case-finding strategy on a national level. Several studies may be required to measure the sensitivity and specificity of the algorithm as a whole in the various populations to which it might be applied, such as the public at large, those attending health fairs, and those receiving routine medical care from a primary care physician. In addition, these studies should fully evaluate some randomly selected individuals at each stage of the algorithm to guide future modifications that might improve performance or reduce the costs of testing.
Recommendation 2. Case-finding efforts in the U.S. should especially target those with moderate-to- severe COPD, specifically those with FEV1 < 60% predicted. Focusing on the identification of those with moderate-to-severe disease allows greater efficiency in a case-finding algorithm, since symptom questionnaires and simple spirometry without bronchodilator are both much better at identifying severe disease than at discriminating between the normal and mildly ill populations. Furthermore, focusing on moderate-to-severe disease does little to diminish the practical clinical value of case-finding, since the more severe patients are those who most clearly benefit from diagnosis and treatment. Taken together, these two considerations suggest that the benefit-to-cost ratio of COPD case-finding will be substantially greater for a strategy targeting moderate-to-severe cases than for a strategy attempting to find mild cases as well. It should be noted, however, that future improvements in COPD treatment, such as the discovery of a disease-modifying drug that is efficacious in mild disease, could undermine the logical basis of this recommendation and justify the development of an approach to case-finding in COPD that is completely different from the one proposed in this report.
The proposed breakpoint at FEV1 <60% is arbitrary and inconsistent with the GOLD categorization of COPD, which uses a threshold value of 50% FEV1 predicted to distinguish between mild and moderate COPD. Nevertheless, the Workshop participants expected that COPD patients with FEV1 between 50% and 60% predicted could be distinguished from the normal population by screening tools and that ongoing clinical trials are likely to demonstrate efficacy of treating this group. Thus, these subjects have been included in the recommended target group for screening.
Recommendation 3. COPD case-finding strategies should employ a three-stage algorithm consisting of 1) a risk factor/symptom questionnaire designed to eliminate those unlikely to have severe disease; 2) a simple measure of expiratory airflow capable of excluding those with normal or nearly normal pulmonary function; and 3) appropriate diagnostic evaluation, including physician appraisal, consideration of alternative diagnoses, and post-bronchodilator spirometry, performed with rigorous quality control. A promising option for the second stage is measurement of peak flow without bronchodilator. Several studies have indicated that individuals with peak flow measurements greater than 75% of the predicted value are very unlikely to have an FEV1 less than 50% predicted.
Recommendation 4. Research is needed to optimize and validate the case-finding strategy described in Recommendation 3. Although existing screening questionnaires, standards for peak flow measurements and interpretation, and guidelines for COPD diagnosis, all can be combined immediately to create a workable algorithm for case-finding, there is a need for optimization of the individual tools. Existing questionnaires have been developed to identify mild as well as more severe cases of COPD, and their validation studies have been performed in that context. While simply changing the scoring threshold of these questionnaires will likely allow them to be used to find more severe cases, better results may be possible with a completely new instrument developed for this specific purpose and involving differently worded or completely different questions.
There may also be a better alternative to the peak flow measurements in the second stage of this case-finding algorithm. Peak flow was an attractive choice to the Workshop participants, because it is simple, familiar to physicians, and potentially reimbursable by Medicare. By specifying peak flow, the group also made clear that they did not intend this initial test for airway obstruction to involve rigorous quality control procedures, since these would require more extensive training of personnel and increase the costs of implementation. Nevertheless, alternatives to peak flow should be considered. One possibility would be to measure FEV1 using a hand-held spirometer, foregoing the calibration and quality control procedures that have been established for diagnostic spirometry. Research would be needed to determine the optimal threshold for this or any alternative measure in a case-finding algorithm.
Opportunities for Additional Research. In the course of their discussions, the Workshop participants identified two patient groups who do not satisfy the classic definition of COPD (based largely on a measure of post-bronchodilator FEV1), but who may have variant forms of the disease. One group combines a history of cigarette smoking with a restrictive pattern of spirometric abnormalities. Another group shows x-ray evidence of emphysema but with normal spirometry. As studies of COPD increase in number and size and case-finding efforts intensify, it is likely that many more of these cases will be recognized. Studies are needed to identify risk factors for these conditions, to determine the prognostic implications of these finding, and to develop guidelines for clinical management of these patients.
- Fernando J. Martinez, M.D., Ann Arbor, MI
- Antonio Anzueto, M.D., San Antonio, TX
- William Bailey, M.D., Birmingham, AL
- R. Graham Barr, M.D., Dr.P.H., New York, NY
- Bart Celli, M.D., Boston, MA
- Dennis Doherty, M.D., Lexington, KY
- Paul Enright, M.D., Tucson, AZ
- Sam Giordano, MBA, RRT, FAARC, Irving, TX
- Lisa LaVange, Ph.D., Chapel Hill, NC
- David Mannino, M.D., Lexington, KY
- Dennis Niewoehner, M.D., Minneapolis, MN
- George O’Connor, M.D., Boston, MA
- Frank Sciurba, M.D., Pittsburgh, PA
- Stuart Stoloff, M.D., Carson City, NV
- Byron Thomashow, M.D., New York, NY
- John Walsh, Miami, FL
- Barbara Yawn, M.D., Rochester, MN
- Thomas Croxton, Ph.D., M.D., Division of Lung Diseases
- Antonello Punturieri, M.D., Ph.D., Division of Lung Diseases
- Robert Smith, Ph.D., Division of Lung Diseases
- Amy Pianalto, Office of Communications