Analytics for medical decision making: Applications to the management of treatment-resistant depression
Bibliographic record
Abstract
The objective of this thesis is the use and design of analytics methods (i.e., methods from operations research and management science, artificial intelligence, and statistics) for medical decision making. In particular, this work focuses on methods to assist physicians towards achieving remission in patients suffering from treatment-resistant depression, a severe form of the major depressive disorder. Following a literature review of medical decision making methods relevant for treating depression, this thesis proposes medical decision making methods for (1) finding the best initial treatment modification for incoming patients, (2) characterizing the current timing decisions between appointments and (3) recommending potential successful treatments. All of these tasks are addressed using observational longitudinal data from the Depressive and Suicide Disorders Program of the Douglas Mental Health University Institute in Montreal.In particular, the first method focuses on the task of using observational data to determine which of five treatment modification strategies is best at the initial visit. To do so, the proposed method balances the five strategy groups using an improved approach for causal inference. The chapter associated with this method is also used as a tutorial to causal inference for the operations research and management science community.The second method identifies the relevant variables among the patient's, physician's and clinic's characteristics for the timing decisions between appointments. This decision is of importance due to the trade-off between high-frequency appointments that lead to a waste of resources and low-frequency appointments that lead to the degradation of patients. Using imitation learning on data, this method infers these variables and their weights. This knowledge can then be used by the physicians to refine and standardize their practice with respect to this decision.The third method recommends potential successful treatments using similarities between past patients and treatments. This recommender system consists somewhat of an extension of the first method where causal inference is again used. However, the treatment drugs are now considered instead of the five treatment modification strategies. For this method, we assume that the sequence of treatments that have been administered to the patient does not affect the efficacy of the current treatment
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 | 0.009 | 0.034 |
| Meta-epidemiology (narrow) | 0.001 | 0.001 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.001 | 0.002 |
| Science and technology studies | 0.001 | 0.001 |
| Scholarly communication | 0.003 | 0.002 |
| Open science | 0.001 | 0.002 |
| Research integrity | 0.002 | 0.004 |
| Insufficient payload (model declined to judge) | 0.003 | 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".