Probabilistic Quantitative Bias Analysis for Misclassification and Uncontrolled Confounding: A Methodological Tutorial Using Real-World Data
Bibliographic record
Abstract
BACKGROUND: Observational studies and clinical trials are vulnerable to multiple sources of bias that can distort study findings. This methodological tutorial demonstrates the application of probabilistic quantitative bias analysis (QBA) to adjust for exposure misclassification, outcome misclassification, and uncontrolled confounding, using the association between obesity (exposure) and hypertension (outcome) as an exemplar. METHODS: Publicly available data from the National Health and Nutrition Examination Survey were temporally split into an analysis dataset to estimate the association of interest (2021-2023, n = 5165) and a validation dataset to derive bias parameters (2003-2016, n = 33,103). Misclassification was introduced by using self-reported measures to define obesity and hypertension rather than measured objective assessments (height, weight, and blood pressure). Uncontrolled confounding was introduced by omitting adjustment for socioeconomic status. Bias parameters were estimated and applied in a Monte Carlo-based probabilistic QBA to obtain bias-adjusted measures of association by age group and sex. RESULTS: After correcting for nondifferential misclassification of obesity, bias-adjusted estimates were consistently larger than conventional estimates across all age and sex subgroups, indicating that self-reported height and weight attenuated the obesity-hypertension association. Accounting for differential misclassification of hypertension generally yielded smaller bias-adjusted estimates, suggesting that self-reported hypertension overestimated the association, except among males 40-59 years of age, where it was attenuated. Adjustment for uncontrolled confounding by socioeconomic status resulted in larger bias-adjusted estimates, particularly in older adults, indicating further attenuation due to residual confounding. CONCLUSIONS: QBA offers a practical approach to assess and account for bias, improving the interpretation and reliability of results in cardiovascular studies.
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 | 0.158 | 0.322 |
| Meta-epidemiology (narrow) | 0.003 | 0.002 |
| Meta-epidemiology (broad) | 0.004 | 0.008 |
| Bibliometrics | 0.004 | 0.005 |
| Science and technology studies | 0.001 | 0.006 |
| Scholarly communication | 0.004 | 0.006 |
| Open science | 0.007 | 0.005 |
| Research integrity | 0.006 | 0.013 |
| Insufficient payload (model declined to judge) | 0.006 | 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".