Table 3.
Validity, reliability, and pre-test probability of TBNA outcome for diagnosing mediastinal and hilar adenopathies.
| Pathology | Patients | Se | Sp | PPV | NPV*** | Accuracy |
|---|---|---|---|---|---|---|
| Non-small cell lung carcinoma* | 280 | 68.3% | 100% | 100% | 9.5% | 69.3% |
| Small cell lung carcinoma** | 74 | 93.2% | 100% | 100% | 16.7% | 93.2% |
| Extrapulmonary carcinoma* | 44 | 69.8% | 100% | 100% | 7.1% | 70.5% |
| Sarcoidosis* | 57 | 63.6% | 100% | 100% | 9.1% | 64.9% |
| Tuberculosis* | 19 | 72.2% | 100% | 100% | 16.7% | 73.7% |
| Lymphoma | 15 | 35.7% | 100% | 100% | 10.0% | 40.0% |
| Anthracotic and reactive | 91 | 52.7% | - | 100% | - | 52.7% |
| Total | 580 | 68.0% | 100% | 100% | 10.0% | 68.8% |
Se, sensitivity; Sp, specificity; PPV, positive predictive value; NPV, negative predictive value.
* p of sensitivity and accuracy < 0.05 compared with the lymphoma group.
** p of sensitivity and accuracy < 0.001 compared with all other pathologies.
*** No statistically significant differences between the various pathologies.