Machine learning models to predict posttraumatic stress injuries in a sample of firefighters: A proof of concept
Bibliographic record
Abstract
Firefighters face significant physical and psychological challenges in their profession that may increase the risk for post-traumatic stress injuries (PTSI). Longitudinal monitoring of PTSI is considered a preventive strategy to manage mental health at work, but tools to predict the probability of PTSI remain limited. To address this issue, our proof of concept aimed to use machine learning models to predict PTSI in firefighters through longitudinal intensive assessment. The study recruited 274 Canadian firefighters that monitored their mental health and psychosocial risk and protective factors with assessments every 2 weeks over 12 weeks. Our analyses trained and tested 27 models, which were developed by combining four different algorithms (logistic regression, support vector classifier and extreme gradient boosting), number of data collection points before target and typology of features (mental health symptoms and psychosocial predictors). Overall, most of the models showed medium-high values of accuracy and specificity, while sensitivity and precision showed greater variability depending on the composition of the models. Model comparison showed that (a) support vector classifier and extreme gradient boosting performed better than logistic regression, (b) the more assessment points prior to the target week are used, the best prediction is obtained, and (c) that full feature set performed better than distress measures only. Our results suggest that combining ML and intensive longitudinal assessment may lead to the development of a potentially useful prevention method in the future.
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 | 0.001 | 0.000 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.001 | 0.000 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.000 | 0.000 |
| Open science | 0.000 | 0.000 |
| Research integrity | 0.000 | 0.000 |
| Insufficient payload (model declined to judge) | 0.000 | 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".