Modeling drug exposures and their time-varying effects : comparison of statistical analysis methods
Bibliographic record
Abstract
Assessing the effects of drug exposure on the occurrence of health events is a real challenge. As individuals' drug exposures are likely to vary over time, they carry associated risks that depend on the dose, duration, and timing of treatment. Thus, different study designs and statistical analysis methods are used to estimate the risk associated with drug exposure. However, few studies have systematically evaluated and compared the respective performances of cohort and nested case-control (NCC) designs to estimate the effect of fixed or time-varying drug exposure. In this thesis work, we used simulations to examine and compare the performance of estimates from NCC versus whole cohort analyses to assess associations between a fixed or time-varying exposure and the risk of a health event. For this comparison, we also used data from the E3N cohort to assess the association between the use of menopausal hormone therapy and breast cancer risk. The results of the simulation study we conducted showed that the estimates obtained from the analysis of the whole cohort were unbiased in all scenarios considered. However, the estimates from the NCC analyses were substantially biased, especially when only one control was matched to each case. This bias in the nested case-control estimates increased with the proportion of events. A significant improvement in the NCC estimates was observed after the use of a bias reduction method, suggesting that the observed biases could be the result of sparse data. However, we were not satisfied with this explanation as the biases persisted regardless of the number of events. We, therefore, pursued our investigations by looking at the handling of tied event times by evaluating different methods to take them into account in the NCC analysis. Our simulation study and application to the E3N cohort data showed that NCC analyses with Breslow or Efron approximations could lead to a significant bias when there were a large number of tied event times in the data. However, once the tied event times were properly accounted for using the exact method or an approach that allowed for a single case in each stratum, the NCC estimates were almost unbiased and close to those of the whole cohort analysis. We strongly recommend that particular attention be paid to tied events in CTN analyses, in particular how they are handled both when forming matched strata and in the analysis by conditional logistic regression.
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 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 | 0.002 | 0.004 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.002 | 0.000 |
| Bibliometrics | 0.000 | 0.000 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.000 | 0.000 |
| Open science | 0.000 | 0.001 |
| Research integrity | 0.000 | 0.001 |
| Insufficient payload (model declined to judge) | 0.002 | 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 teacher head, 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".