HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Full Text (PDF) Services Similar articles in this journal Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via Google Scholar Google Scholar Articles by Sales, R. K. Articles by Antonangelo, L. Search for Related Content PubMed Articles by Sales, R. K. Articles by Antonangelo, L.

A COST-MINIMIZING DIAGNOSIS MODELS FOR DISCRIMINATION BETWEEN NEOPLASTIC AND TUBERCULOUS PLEURAL EFFUSIONS UTILIZING ROUTINE CLINICAL AND LABORATORY VARIABLES

Pulmonary Division - Heart Institute (InCor), University of São Paulo Medical Sc, São Paulo, Brazil

PURPOSE: To identify clinical and laboratory parameters capable to differentiate between tuberculous and malignant pleural effusions with high efficiency and low-costs.

METHODS: Laboratory tests (glucose, protein, albumin, globulin, lactate dehydrogenase, cholesterol, apolipoprotein A, apolipoprotein B, adenosine deaminase (ADA), quantitative and oncotic cytology) were analysed from 403 cases of confirmed tuberculous (Tb=200) or malignant (Mal=203) pleural effusions. Clinical variables like age, pleural effusions side of incidence (right or left) , size (> or < 75% of the compromised hemithorax) and macroscopic appearance (hemorragic or not) were also analysed. Statistical analysis: Firstly, we used univariate tests to detect the variables that significantly differentiated the groups (Tb and Mal). After, we submit these selected variables to logistic regression. With the ß coefficients associated to the variables that composed the best models we purposed algorithms capable to do these diagnosis with the best efficiencies and lower costs. After, we tested the classificatory power of the models in 64 pleural exsudates whose diagnosis were unknown.

RESULTS: For pleural tuberculosis, the best model included: ADA, globulin and negative oncotic cytology given a 99.4% sensitivity;96.1% specifity;95.7% PPV and 99.5% of NPV. For neoplastic pleurisy, the best model was composed by age, hemmorrhagic aspect, macrophage percentual and positive or suspicious cytology that given a 96.3% sensitivity;91.4% specifity;91.1% PPV and 96.6 of NPV. When these models were applied to the 64 exsudates, the sensitivity, specificity, PPV and NPV were respectively: For tuberculosis diagnosis: 100%, 91.1%, 99.9% and 100%.For malignancy: 82.1%, 100%, 100% and 87.8%.

CONCLUSION: These simplified models presented good efficiency in diagnosing pleural tuberculosis and malignancy and do this at the expense of routine and low-cost variables.

CLINICAL IMPLICATIONS: The possibility of using these models in the clinical practice without to be necessary doing pleural biopsies.
Scoring system to predict tuberculosis and malignancy

Tuberculosis - Score > 8.5 points Characteristics Coefficient Score ADA (>46.5) 6.288 4.0 Negative Oncotic Cytology 6.456 4.5 Globulin (>2.05 mg/dl) 1.897 1.5

Malignancy - Score > 8.5 points Characteristics Coeficient Score Age (>45.5 years) 2.511 6.0 Hemorrhagic Aspect 1.048 2.5 Macrophages Percentual 1.482 1.5 Positive or Suspicious Oncotic Cytology 5.589 5.0

DISCLOSURE: Roberta Sales, None.