Methods
The study protocol was exempted from review as set forth in the Code of Federal Regulations, 45 CFR 46.101(b) by the Bridgeport Hospital Institutional Review Board. Spirometric tests, conducted at Bridgeport Hospital between March 2010 and August 2011, were retrieved from the laboratory's electronic database. Spirometry was performed according American Thoracic and European Respiratory Society (ATS-ERS) guidelines and patients were instructed not to take bronchodilators starting on the night prior to study. Metacholine broncho-provocation tests were performed when requested by the referring physician. Spirometric tests that were not considered acceptable by the reading pulmonologist were not enrolled. Interpretations of pulmonary function tests (PFTs) were performed by American Board of Internal Medicine-certified pulmonologists who applied ATS-ERS standards and guidelines. Per these standards, obstruction is defined as reduced forced expired volume in 1 second: vital capacity (FEV1/VC) ratio below the 5th percentile of the predicted value. Asthma is defined as obstruction and increase in FEV1 and/or forced vital capacity (FVC) of ≥12% from baseline in response to bronchodilator, or a positive metacholine challenge test. A restrictive defect is defined as total lung capacity (TLC) below the 5th percentile of the predicted value. In the absence of measured lung volumes, a restrictive ventilatory defect is suggested by a reduced VC when FEV1/VC is increased (85–90%) and the flow–volume curve demonstrates a convex pattern. A "mixed pattern" is defined as both FEV1/VC ratio and TLC below the 5th percentiles of their predicted values. In our study, abnormal PFTs that could not be classified in any of the above categories using ATS-ERS standards and guidelines were categorized as inconclusive. Our lab does not measure VC routinely and the FVC was used for interpretation of function tests.
The following data were extracted from electronic medical records for all patients with body mass index (BMI) > 30 kg/m: demographics, co-morbidities, smoking history, pre-testing pulmonary diagnosis, interpretation of spirometry, medications prior to testing, and medications following testing (≥6 months later). Patients were excluded if they had no hospital medical records before and after the spirometry. We followed all the patients for a minimum of 6 months after spirometry. Logistic regression analysis was performed using Epi Info™ to identify variables associated with perseverant treatment of patients with bronchodilators in the absence of airflow obstruction on spirometry. Age, BMI, congestive heart failure (CHF), diabetes mellitus (DM), hypertension (HTN), sex and smoking were chosen as independent variables for modeling based on biological plausibility and/or if they demonstrated an association with inappropriate bronchodilator treatment in univariate analyses.
A P < 0.05 signified statistical significance.