Optimal dynamic treatment regime structural nested mean models: improving efficiency through diagnostics and re-weighting and application to adaptive individual dosing
Bibliographic record
Abstract
Dynamic treatment regimes are common in medicine, for example in the treatment of chronic diseases. As information about a patient is gathered over time, it is desirable to make use of this accumulating information to make treatment decisions that are specifically tailored to the individual patient, or to base decisions on dynamically evolving observations. Dynamic treatment regimes have been the topic of much recent work in the area of causal inference. In particular, semi-parametric methods for estimating a "best" or "optimal" treatment rule or strategy from observational data have been developed. One such method proposed by Robins is the optimal dynamic treatment regime structural nested mean model (ODTR-SNMM) and associated g-estimation procedure. Of significant concern when applying this methodology are the modelling assumptions involved. In this work, checking of modelling assumptions using residual and influence diagnostics as is typically done in a traditional regression setting is extended to the ODTR-SNMM. The methodology is evaluated on simulated data under different model specification settings. These ideas are also applied to real data from a breastfeeding cessation study. Subsequently, partially misspecified models, which give rise to consistent though inefficient estimation of the parameter of interest due to misspecification of a nuisance model, are considered. In addition to the possibility of addressing partial misspecification through the proposed diagnostic techniques, re-weighting is considered as a means of improving the efficiency of estimators under these modeling assumptions. A re-weighting approach based on sample influence is proposed and studied with simulations. Finally, the application of optimal dynamic treatment regimes estimation to adaptive dosing strategies for drugs with narrow therapeutic windows and highly variable dosing is considered. Using oral anticoagulation therapy as a motivating example, a simulation is designed using realistic pharmacokinetic (PK) and pharmacodynamic (PD) models to generate the data. A modelling approach for ODTR-SNMM with continuous dosing is proposed and applied to the PK/PD simulated data. The performance of various models under different settings is compared.
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 | 0.017 | 0.063 |
| Meta-epidemiology (narrow) | 0.001 | 0.001 |
| Meta-epidemiology (broad) | 0.003 | 0.002 |
| Bibliometrics | 0.001 | 0.001 |
| Science and technology studies | 0.000 | 0.001 |
| Scholarly communication | 0.001 | 0.002 |
| Open science | 0.003 | 0.002 |
| Research integrity | 0.002 | 0.003 |
| 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 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".