EVALUATING DIAGNOSTIC TESTS IN THE PLEURAL SPACE : Differentiating Transudates from Exudates as a Model - 09/09/11
Résumé |
The use of diagnostic tests to detect or confirm the presence of disease is a central function of clinical medicine and accounts for a large portion of health care expenditures in patient care. The increasing number, cost, and sophistication of diagnostic tests18 and their impact on patient care warrant rigorous assessment and validation of new tests before they are adopted into routine clinical practice.50
Unfortunately, relatively little attention has been directed toward applying a standard methodology to the assessment of new tests despite the fundamental importance of diagnostic testing to clinical medicine. Although methodologic standards for new test evaluation exist,* multiple studies indicate that those standards have been incompletely applied by investigators who study and report the accuracy of diagnostic tests.2, 4, 5, 6, 12, 20, 25, 36, 50, 57 In one recent critical appraisal, it was noted that most diagnostic tests are inadequately evaluated in studies that promote their use.50 Incomplete conformance to methodologies in diagnostic test-assessment research stands in marked contrast to the rigorous standards expected for the staged evaluation of new pharmaceutical agents or the reporting of randomized, controlled trials of new therapies.
Diagnostic test-assessment methodologies are intended to limit the introduction of potential biases into experimental design that escape the investigator's attention and thereby promote inaccurate assessment of test efficacy. In the absence of such methodologies, promising initial reports may introduce into clinical medicine diagnostic tests that subsequently prove disappointing.46, 47, 49 The dexamethasone suppression test for depression46 and serum carcinoembryonic antigen determinations for colon cancer19 represent two tests that were greeted with early enthusiasm and widespread application before subsequent investigations identified critical biases in the initial assessment studies. Later studies failed to support their clinical use. The premature adoption of diagnostic tests that later prove of little value both wastes health care resources and potentially harms patients by incorrectly labeling their state of health and misdirecting diagnostic and therapeutic care.18, 47
This article reviews major standards for assessing diagnostic tests to assist readers of the medical literature and investigators who are designing assessment studies. Standards are discussed in relationship to pleural-fluid tests that discriminate between exudative and transudative pleural effusions (hereafter termed exudate-transudate studies) because of the familiarity of those diagnostic tests to pulmonary physicians and their importance to clinical practice. Throughout the discussion, the important questions that clinicians should ask when considering the acceptability of a new diagnostic test will be kept in focus: (1) Are the studies supporting the use of the test valid? (2) What is the test's efficacy and diagnostic accuracy? (3) How should the test be applied to the patient at hand?
Le texte complet de cet article est disponible en PDF.Plan
| Address reprint requests to John E. Heffner, MD, FCCP, Department of Medicine, St. Joseph's Hospital and Medical Center, 350 West Thomas Road, Phoenix, AZ 85013 |
Vol 19 - N° 2
P. 277-293 - juin 1998 Retour au numéro
Article précédent
- CT AND MR IN PLEURAL DISEASE
- Theresa C. McLoud
- DIAGNOSTIC THORACOSCOPY
- Christian Boutin, Philippe Astoul
Bienvenue sur EM-consulte, la référence des professionnels de santé.
L’accès au texte intégral de cet article nécessite un abonnement.
Déjà abonné à cette revue ?