Automated Monitoring of Clinical Practice Guideline Adherence Using FHIR and OMOP: A Multi-Center Study in Intensive Care Units
Bibliographic record
Abstract
Abstract Background Clinical practice guidelines are important tools for clinical decision support, but monitoring guideline adherence manually is highly resource-intensive. Therefore, we developed an automated system for evaluating guideline adherence based on computer-interpretable representations of guidelines. We implemented the system across multiple university hospitals and assessed its validity and performance by comparing its guideline adherence evaluations to those conducted by medical professionals. Methods We selected six representative clinical guideline recommendations from across 41 intensive care guidelines and translated these text-based recommendations into a computer-interpretable, Fast Healthcare Interoperability Resources (FHIR)-based format using an iterative consensus approach. Clinical data from five university hospitals were transformed into the Observational Medical Outcomes Partnership (OMOP) common data model. A decision support system was developed to interpret FHIR-encoded recommendations and apply them to OMOP-formatted patient data. We evaluated the system retrospectively on intensive care data covering 3.5 years and validated its performance by comparing system-generated decisions with human decisions in three hospitals. We created and iteratively refined a user interface for individual and ward-level adherence visualization. Findings We expert-reviewed more than 18,000 patient days to assess the applicability of and adherence to the recommendations. The system demonstrated 97.0% accuracy in identifying guideline applicability and adherence, with significantly higher accuracy than human reviewers (accuracy 86.6%, p<0.001, McNemar’s Test). The automated system processed more than 2000 patient days per second for a total of 2,200,000 patient days across 82,000 intensive care episodes, compared to humans’ two patient days per minute. Interpretation We demonstrate that an automated adherence monitoring system outperforms human reviewers in both accuracy and time efficiency. Using FHIR-encoded recommendations enables flexibility and scalability across hospitals with different data infrastructures. Future work should focus on integrating unstructured patient data and expanding the range of encoded recommendations. Funding Federal Ministry of Education and Research of Germany.
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 | 0.006 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 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) | 0.000 | 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".