The MEDLINE database was searched to identify studies published from 1992 to 1997 that reported the discriminative properties of diagnostic tests used to evaluate patients with suspected pulmonary conditions. The search command included the following terms: diagnostic tests; diagnostic tests, routine; predictive value of tests; and thoracic radiography. The search was restricted to six subspecialty journals that had a major interest in pulmonary conditions and nine general internal medicine journals. Buy proventil add comment The six subspecialty journals were American journal of Respiratory and Critical Care Medicine, Annals of Thoracic Surgery, Chest, European Respiratory Journal, Journal of Thoracic and Cardiovascular Surgery, and Thorax. The nine general internal medicine journals were American Journal of Medicine, Annals of Internal Medicine, Archives of Internal Medicine, British Medical Journal, Canadian Medical Association Journal, Journal of the American Medical Association, Lancet, Mayo Clinic Proceedings, and New England Jou rnal of Medicine. The MEDLINE abstracts of the identified articles were reviewed to determine eligibility for inclusion into the study. Articles were included for further review if their abstracts indicated that the discriminative properties of a diagnostic test were examined in humans, the diagnostic test was intended for the detection of an existing condition, and the target order was relevant to pulmonary medicine. Studies that examined the value of a test for determining the risk of a future event (eg,failure of weaning from mechanical ventilation or risk for the occurrence of acute lung injury) or for screening asymptomatic individuals to detect disease were excluded. Articles that appeared to be eligible for study inclusion were retrieved and read in their entirety. Upon this closer examination, articles were selected on the basis of the inclusion and exclusion criteria listed above. Methodologic and Data Analysis Standards
The methodologic standards evaluated in the study were established at the outset of the study and were compiled from established resources for assessing the quality of investigational design for studies in diagnostic test research. Additional standards assessed the descriptors used to quantitate the diagnostic accuracies of the tests, the methods selected for comparing two or more evaluated tests, and the degree to which the evaluated tests’ clinical value in patient care was discussed.