Validation of a registry-derived risk algorithm based on treatment protocol as a proxy for disease risk in childhood acute lymphoblastic leukemia
Bibliographic record
Abstract
BACKGROUND: Administrative databases and cancer registries are frequently used to conduct population-based research, but often lack clinical data necessary for risk stratification. Our objective was to determine the criterion validity of a risk-stratification algorithm based on treatment characteristics available from a pediatric cancer registry as a proxy for disease risk, by comparing it to traditional biology-based risk classifications. METHODS: We identified all children with acute lymphoblastic leukemia diagnosed at a single institution between January 2000 and June 2011, and linked them to a population-based cancer registry. Several risk algorithms were then constructed using disease risk variables collected through chart review by a pediatric oncologist, and compared to a risk algorithm based on treatment protocol name and age, available from the registry. RESULTS: Of 596 patients identified, 579 (97.1%) met inclusion criteria and were successfully linked. The registry-based algorithm showed almost perfect agreement with a biology-based algorithm based on age, initial white blood cell count, immunophenotype and cytogenetics (kappa=0.85, 95th confidence interval 0.81-0.90). Discrepant cases were often due to the presence of unusual high risk features not captured by standard disease-risk variables but reflected in clinicians' choices of higher intensity treatment protocols. CONCLUSIONS: Protocol name represents a valid proxy of disease risk, allowing for risk stratification while conducting comparative effectiveness research using cancer registries and health services data. Future studies should examine the validity of treatment-based risk algorithms in other malignancies and using other treatment characteristics commonly found in health services data, such as the receipt of specific chemotherapeutic agents.
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 | 0.089 | 0.200 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.004 | 0.003 |
| Science and technology studies | 0.001 | 0.001 |
| Scholarly communication | 0.002 | 0.002 |
| Open science | 0.002 | 0.002 |
| 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".