Background: Malignant pleural effusion (MPE) is common and diagnosis is often problematic. A cancer ratio (serum lactate dehydrogenases: pleural adenosine deaminase ratio) has been proposed for diagnosing MPE. However, the usefulness of this "cancer ratio" and the clinical-radiological criteria for diagnosing MPE has not been clearly determined to date. The aim of this study was to assess the performance of those parameters in the diagnosis of MPE.
Methods: We analyzed 240 patients including 120 with MPE and 120 with non-MPE (93 tuberculous and 27 parapneumonic). Patients were divided into two groups: MPE and non-MPE (eg, tuberculous and parapneumonic). We constructed two predictive models to assess the probability of MPE: (a) clinical-radiological data only and (b) a combination of clinical-radiological data, the cancer ratio, and the carcinoembryonic antigen (CEA). The performances of the predictive models were assessed using receiver operating characteristic (ROC) curves and by examining the calibration.
Results: The area under the ROC curves for model 1 and model 2 were excellent, 0.936 and 0.998, respectively. The overall diagnostic accuracies for model 1 and model 2 were 87.5% and 98.8%, respectively.
Conclusion: The results confirm that both models achieved a high diagnostic accuracy for MPE; however, model 2 was superior with the addition of its simplicity of use in daily practice. This model should be applied to determine which patients with a pleural effusion of unknown origin would not benefit from further invasive procedures.
Keywords: diagnosis; lactate dehydrogenases; logistic models; malignant; pleural effusion.
© 2018 Wiley Periodicals, Inc.