Correcting hazard ratio estimates for outcome misclassification using multiple imputation with internal validation data
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
OBJECTIVE: Outcome misclassification may occur in observational studies using administrative databases. We evaluated a two-step multiple imputation approach based on complementary internal validation data obtained from two subsamples of study participants to reduce bias in hazard ratio (HR) estimates in Cox regressions. METHODS: We illustrated this approach using data from a surveyed sample of 6247 individuals in a study of statin-diabetes association in Quebec. We corrected diabetes status and onset assessed from health administrative data against self-reported diabetes and/or elevated fasting blood glucose (FBG) assessed in subsamples. The association between statin use and new onset diabetes was evaluated using administrative data and the corrected data. By simulation, we assessed the performance of this method varying the true HR, sensitivity, specificity, and the size of validation subsamples. RESULTS: The adjusted HR of new onset diabetes among statin users versus non-users was 1.61 (95% confidence interval: 1.09-2.38) using administrative data only, 1.49 (0.95-2.34) when diabetes status and onset were corrected based on self-report and undiagnosed diabetes (FBG ≥ 7 mmol/L), and 1.36 (0.92-2.01) when corrected for self-report and undiagnosed diabetes/impaired FBG (≥ 6 mmol/L). In simulations, the multiple imputation approach yielded less biased HR estimates and appropriate coverage for both non-differential and differential misclassification. Large variations in the corrected HR estimates were observed using validation subsamples with low participation proportion. The bias correction was sometimes outweighed by the uncertainty introduced by the unknown time of event occurrence. CONCLUSION: Multiple imputation is useful to correct for outcome misclassification in time-to-event analyses if complementary validation data are available from subsamples. Copyright © 2017 John Wiley & Sons, Ltd.
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 | 0.004 | 0.023 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.000 | 0.000 |
| Science and technology studies | 0.001 | 0.000 |
| Scholarly communication | 0.000 | 0.000 |
| Open science | 0.000 | 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