Article Text
Abstract
Introduction Clinical trials in chronic obstructive pulmonary disease (COPD) have spirometric criteria for enrolment, and published screening failure rates against these criteria have been reported as low as 7%1. Primary care databases provide a source of patients with COPD to recruit into clinical trials. However, UK COPD lists are heterogenous, with only 73% having confirmed COPD when retested2. This may result in a higher than expected screening failure rate. We reviewed this using our experience of recruiting to a COPD trial through primary care databases.
Methods Local GP surgeries were requested to contact all patients on their COPD database after applying exclusion criteria for age and co-morbidities. Patients were sent an invitation letter with a reply slip to return directly to the study team and those interested in participating were contacted for screening spirometry and medical history. Surgeries were reimbursed up to £500 administrative costs. Patients fulfilled diagnostic inclusion criteria if the FEV1/FVC ratio was <0.7 and FEV1<80% predicted with a smoking history and without other obstructive lung disease.
Results Approximately 1400 letters were sent from 29 GP surgeries between February and July 2012; 283 replies were received by the study team (20%). Of these, 180 were not screened (125 [69%] declined to participate, 9 [5%] had diseases other than COPD, 9 [5%] were enrolled in other research, and 37 [21%] were unsuitable for other reasons). 8 appointments were pending with screening data available for 95 patients (Table 1). 40 patients (42%) did not satisfy the spirometric inclusion criteria; 27 did not show airflow obstruction, 10 had mild COPD, and 3 had asthma. Approximately 25 contact letters and up to £260 were therefore required per eligible patient identified.
Conclusion A high proportion of patients on primary care databases fail to meet spirometric criteria for COPD trials and the screening failure rate via this recruitment pathway is much higher than previously reported. A large number of initial contacts are required for each patient identified. COPD patients are increasingly managed in primary care and these findings therefore have implications for planning future studies.
Albert, NEJM, 2011; 365(8):p689–98.
Jones, Respir Res, 2008; 9:p62.