Improved Correction of Misclassification Bias With Bootstrap Imputation
Bibliographic record
Abstract
OBJECTIVE: Diagnostic codes used in administrative database research can create bias due to misclassification. Quantitative bias analysis (QBA) can correct for this bias, requires only code sensitivity and specificity, but may return invalid results. Bootstrap imputation (BI) can also address misclassification bias but traditionally requires multivariate models to accurately estimate disease probability. This study compared misclassification bias correction using QBA and BI. STUDY DESIGN: Serum creatinine measures were used to determine severe renal failure status in 100,000 hospitalized patients. Prevalence of severe renal failure in 86 patient strata and its association with 43 covariates was determined and compared with results in which renal failure status was determined using diagnostic codes (sensitivity 71.3%, specificity 96.2%). Differences in results (misclassification bias) were then corrected with QBA or BI (using progressively more complex methods to estimate disease probability). RESULTS: In total, 7.4% of patients had severe renal failure. Imputing disease status with diagnostic codes exaggerated prevalence estimates [median relative change (range), 16.6% (0.8%-74.5%)] and its association with covariates [median (range) exponentiated absolute parameter estimate difference, 1.16 (1.01-2.04)]. QBA produced invalid results 9.3% of the time and increased bias in estimates of both disease prevalence and covariate associations. BI decreased misclassification bias with increasingly accurate disease probability estimates. CONCLUSIONS: QBA can produce invalid results and increase misclassification bias. BI avoids invalid results and can importantly decrease misclassification bias when accurate disease probability estimates are used.
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 | 0.127 | 0.398 |
| Meta-epidemiology (narrow) | 0.002 | 0.001 |
| Meta-epidemiology (broad) | 0.003 | 0.003 |
| Bibliometrics | 0.003 | 0.005 |
| Science and technology studies | 0.001 | 0.001 |
| Scholarly communication | 0.002 | 0.002 |
| Open science | 0.003 | 0.004 |
| Research integrity | 0.002 | 0.003 |
| Insufficient payload (model declined to judge) | 0.003 | 0.001 |
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".