How representative are clinical study patients with asthma or COPD for a larger "real life" population of patients with obstructive lung disease?

Evidence-based medicine is a corner stone in treatment decision making and large randomised, clinical trials are usually designed in order to provide highly significant results. This study was conducted in order to find out to what extend a "real life" patient population with obstructive lung disease could fit into criteria commonly used in clinical research trials. As a secondary aim of the study, we wanted to compare the OLD population recruited from GP's and specialist outpatient clinics, respectively. Eight-hundred and seventy prospective OLD patients were included. Criteria's for selecting asthma patients to a clinical trial were, absence of co-morbidity, FEV 50-85% of predicted, present or historical reversibility 12% last year, non-smoke or if ex-smoke a smoke burden less then 10 pack years. Only 5.4% of the study asthma patients met with these criteria. Additional criteria as being symptomatic and regular use of inhaled corticosteroids reduced the numbers of eligible asthma patients to 3.3% representing 1.3% of the entire population. The same procedure was applied for the COPD patients, requesting a FEV1 <70% of predicted normal, significant smoke history (>15 pack years) and absence of atopy. This selected 17% of the COPD population, representing 7% of the entire population. We conclude that "evidence based" treatment decisions for OLD are based on studies which include a very small and highly selected fraction of this patient population. It is questionable whether such data can extrapolated to a larger, "real life" population of patients with obstructive lung disease. Moreover, we found surprisingly minor differences between the Specialist and GP populations.