PURPOSE: To study the use of cut off values of fluid ADA level in differential diagnosis of tuberculous pleurisy in country with low incidence of tuberculosis.
METHODS: 121 consecutive patients with pleural effusion were prospectively studied The study included 82 males ,39 females with a mean age of 59.3±17 years. Exudative effusions were classified: malignant pleural effusions, 39 (39.8%) , TB pleural effusions, 39 (39.8%), parapneumonic pleural effusions, 20 (20.4%) .The photometric method was used
RESULTS: In tuberculous group the mean of ADA level was 122.4 ± 53.3UI/L, malignant effusion it was 59.7±31.7 UI/L (p= 0.0001) parapneumonic effusions 96 ± 48UI/L. Fluid ADA in 24 / 39 of malignant effusions and 18 / 20 parapneumonic effusions was above diagnostic cut-off 40 UI/L; in four of 39 tuberculous pleurisy ADA was under 40 UI/L. Mean ADA level was greater in patients < 35 years old 141.9±46.6 UI/L. No patients of age < 35year had ADA level under 40 UI/L. The sensitivity of ADA in TB effusions at cut-off > 40 U/L was 89.7%, specificity 28.8%, PPV 54.5%. The cut-off value at 95 UI/L had the greatest YOUDEN index(0.488), best Lahr+ (3.4), the highest specificity ,PPV 79.6%, and 69.2% respectively,sensibility 69.2%.
CONCLUSIONS: In countries with low incidence of TB the diagnostic cut off value of ADA level have to be different from countries with high incidence. The best cut off value of pleural fluid ADA level in diagnosis of TB was 95UI/L.In many cases ADA levels in non-tuberculous exudative pleural effusions exceeded the cut off values of tubercular disease (40 UI/L). Adenosine deaminase level of less than 40 U/L in patients < 35 years excludes the tubercular etiology of exudative pleural effusion.
CLINICAL IMPLICATIONS: In diagnosis of pleural effusions the cut off values of ADA pleural effusion have to consider according the age of patients and incidence of tuberculosis.
DISCLOSURE: The following authors have nothing to disclose: Perlat Kapisyzi, Dhimitraq Argjiri, Anila Aliko,, Jeta Beli,, Ylli Vakeflliu,, Roland Kore, Elenka Shehu, Holta Tafa, Regina Hasa, Richard Light
No Product/Research Disclosure Information