Diagnostic Accuracy in Fit-for-Purpose PD-L1 Testing
Why this work is in the frame
A frame that forgets how it found something cannot be audited. These are the routes that admitted this work.
Bibliographic record
Abstract
PD-L1 testing by immunohistochemistry (IHC) has presented significant challenges not only for clinical laboratories, but also for external quality assurance (EQA) entities that provide proficiency testing (PT) for clinical laboratories. Canadian Immunohistochemistry Quality Control (CIQC) has used educational runs to explore approaches to sample design and analysis of results that would enhance patient safety. As PT for predictive biomarkers requires modeling at every level (design of the run, assessment of the run, and reporting of "pass" or "fail") based on "fit-for-purpose" principles, CIQC has applied those principles to PD-L1 PT runs. Each laboratory received unstained slides with TMA tissue cores from 104 randomly selected primary NSCLC and tonsil tissues to test with their current PD-L1 assay. Diagnostic sensitivity and specificity were calculated against designated gold standards based on the "3D" approach (drug-disease-diagnostic assay). Depending on the selection of fit-for-purpose gold standards and also on the selection of what was considered fit-for-purpose cut-off points, great variation in the performance (accuracy) of both companion/complementary diagnostic assays and laboratory developed tests was seen. "Fit-for-purpose" in PT for PD-L1 testing entails that the purpose(s) of each PT run is declared a priori, that the PT program has selected/designated purpose-specific gold standard results for the PT challenge, and that the PT materials for the PT run are designed and constructed to enable calculations of diagnostic accuracy.
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.
Full frame distilled prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. Learned from the 10,348 direct Codex labels and 10,348 direct Gemma labels. Candidate is the union of thresholded teacher heads; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels or direct frontier model labels.
Codex and Gemma teacher scores by category
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.000 | 0.002 |
| Meta-epidemiology (narrow) | 0.000 | 0.001 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.000 | 0.000 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.000 | 0.000 |
| Open science | 0.001 | 0.000 |
| Research integrity | 0.000 | 0.000 |
| Insufficient payload (model declined to judge) | 0.000 | 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 it