Bayesian accuracy estimates for diagnostic tests to detect tuberculosis in captive sun bears (Helarctos malayanus) and Asiatic black bears (Ursus thibetanus) in Cambodia and Vietnam
Bibliographic record
Abstract
Effective control of tuberculosis (TB) depends on early diagnosis of disease, yet available tests are unable to perfectly detect infected individuals. In novel hosts diagnostic testing methods for TB are extrapolated from other species, with unknown accuracy. The primary challenge to evaluating the accuracy of TB tests is the lack of a perfect reference test. Here we use a Bayesian latent class analysis approach to evaluate five tests available for ante-mortem detection of pulmonary TB in captive sun bears and Asiatic black bears in Southeast Asia. Using retrospective results from screening of 344 bears at three rescue centres, we estimate accuracy parameters for thoracic radiography, a serological assay (DPP VetTB), and three microbiological tests (microscopy, PCR (Xpert MTB/RIF, Xpert MTB/RIF Ultra), mycobacterial culture) performed on bronchoalveolar lavage samples. While confirming the high specificities (≥ 0.99) of the three microbiological tests, our model demonstrated their sub-optimal sensitivities (<0.7). Thoracic radiography was the only diagnostic method with sensitivity (0.95, 95% BCI: 0.76, 0.998) and specificity (0.95, 95% BCI: 0.91, 0.98) estimated above 0.9. We recommend caution when interpreting DPP VetTB results, with the increased sensitivity resulting from treatment of weakly visible reactions as positive accompanied by a drop in specificity, and we illustrate how the diagnostic value of weak DPP VetTB reactions is particularly reduced if disease prevalence and/or clinical suspicion is low. Conversely, the reduced utility of negative microbiological tests on bronchoalveolar lavage fluid samples when prevalence and/or clinical suspicion is high is demonstrated. Taken together our results suggest multiple tests should be applied and accompanied by consideration of the testing context, to minimise the consequences of misclassification of disease status of bears at risk of TB in sanctuary settings.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.025 | 0.050 |
| Meta-epidemiology (narrow) | 0.001 | 0.001 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.002 | 0.000 |
| Science and technology studies | 0.001 | 0.002 |
| Scholarly communication | 0.002 | 0.001 |
| Open science | 0.001 | 0.001 |
| Research integrity | 0.001 | 0.001 |
| Insufficient payload (model declined to judge) | 0.001 | 0.000 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".