Therefore, the findings of the study are substantially inconclusive: spi-rometric testing did not seem to reinforce a clinical suspicion obtained with conventional evaluation. This would be in agreement with Thiadens et al, who also found that most cases of asthma and COPD can be identified only with history and objective examination. However, a number of possible negative biases in our study—some systemic, others strictly relevant to the study—must be underlined.
Among systemic biases, objective difficulties in GPs’ daily professional activities (mainly lack of time) caused a reduced enrollment by approximately one half of the number planned; as a matter of fact, the spirometry procedure is highly cooperation dependent and requires time for the instruction of patients and for correct performance (at least three maneuvers were recommended). In contrast to the structure of general practice in other countries (the Netherlands or United Kingdom), in Italy GPs work in most cases individually and without technical or nursing support, which can make the difference when patients require deeper investigation and have no prompt access to specialist care. Another important issue is that as yet no fee is granted for spirometry by the national health service in Italy. comments
Among biases specific for the study, the randomization procedure proved to be complicated and was not applied or was inappropriately applied in a large portion of patients. The quality control of the spirometry relied on the automatic control of the instrument, and neither a regular check nor a direct technical support by the reference centers was envisaged, if not requested by the GP. The underuse of bronchodilator tests certainly influenced the diagnostic accuracy by GPs. Incorrect labeling of patients with respiratory complaints of chronic bronchitis as COPD despite a FEV1/FVC ratio > 0.7 probably contributed to diagnostic misclassification; the inclusion of stage 0 in Global Initiative for Chronic Obstructive Lung Disease guidelines may have contributed to such a classification error.