Bibliographic record
Abstract
The Intensive Care Unit (ICU) is a resource-intensive, costly environment. Data gathered from patients during their stay in the ICU has traditionally been used for clinical purposes, but can have a significant impact on healthcare capacity planning and patient flow. There is a need to study how metrics collected in Canadian ICUs, such as the Multiple Organ Dysfunction Syndrome (MODS) score and the Nine Equivalents of Nursing Manpower Use Score (NEMS) can be used to improve capacity planning decisions. Using discrete-event simulation, statistical, survival and machine learning models, I have built long- and short-term capacity planning models to help hospital administrators better manage patient flows in the ICU. This dissertation consists of three essays that explore the use of these metrics in ICU capacity planning. In the first essay, I study the incorporation of the nursing manpower score NEMS into a discrete-event simulation model to estimate optimal long-term capacity levels of critical care beds in both Level 3 (ICU) and Level 2 (step-down) units. Using data from London Health Sciences Centre (LHSC) University Hospital, I demonstrate the benefits of simulating patients’ daily NEMS changes as triggers for transfer to a step-down unit. This essay also examines ways in which transfer to a step-down unit may improve patient length of stay (LOS), flow and costs. In the second essay, I demonstrate that the ICU LOS literature shows the predominance of multiple linear regression models for individual patients’ ICU LOS and outcome predictions (e.g., death, discharge, long stay). Using data from LHSC’s two ICUs, I compare the performances of well known statistical models with contemporary supervised machine learning models in predicting such outcomes. I show that there is no dominant model in terms of individual patients’ LOS predictions, but that outcome prediction (death, discharge, long stay) performance can be improved by using supervised machine learning techniques. In the third essay, I build on the use of NEMS to simulate realistic ICU LOS for long term capacity planning, and on the use of NEMS and MODS to predict individual ICU LOS in order to improve short-term capacity planning. First, I fit a parametric survival model called the Accelerated Failure Time (Weibull AFT) model with LHSC’s UH data. Then I analyze the model’s hazard rates, event time ratios and LOS, both at the time of the patient’s arrival in the ii ICU and after 3 days’ stay. Finally, I generate daily patient survival probabilities and pool them to predict future expected ICU occupancy rates. Using survival probability pooling for short term capacity planning is a novel use of the ATF model, and may be used to accurately predict ICU occupancy.
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 | 0.005 | 0.019 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.001 | 0.001 |
| Science and technology studies | 0.002 | 0.005 |
| Scholarly communication | 0.005 | 0.005 |
| Open science | 0.001 | 0.002 |
| Research integrity | 0.003 | 0.009 |
| Insufficient payload (model declined to judge) | 0.008 | 0.003 |
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".