Automated Chronic Obstructive Pulmonary Disease Phenotyping and Control Assessment in Primary Care: Retrospective Multicenter Study Using the Seleida Model
Bibliographic record
Abstract
Background: Chronic obstructive pulmonary disease (COPD) remains a leading global health burden. In primary care, the inconsistent availability of spirometry and symptom scores limits the detection of patients with poor disease control. There is a pressing need for scalable, data-driven tools that leverage routinely collected clinical information to support timely, equitable, and guideline-concordant interventions. Objective: This study aims to validate the performance of Seleida-a fully automated, deterministic, and bijective model for COPD control assessment and phenotyping-using real-world primary care data and to evaluate its feasibility for integration into electronic health record (EHR)-based informatics systems. Methods: Seleida estimates the probability of poor control (Pr) using two objective EHR variables: (1) annual dispensations of short-acting bronchodilators-specifically short-acting β2-agonists (SABA), short-acting muscarinic antagonists (SAMA), or both, and (2) number of dispensed antibiotic courses for bronchitis or COPD exacerbations. Its bijective structure supports both forward risk estimation and reverse phenotype inference. In a retrospective cohort of 106 patients, agreement was assessed between 2 phenotyping systems (a 126-combination model and a streamlined 21-combination version) and with clinician-assigned classifications. Due to sample size limitations, a provisional risk threshold of Pr>.50 was adopted for internal stratification. Results: Seleida showed perfect agreement between phenotyping systems (Cohen κ=1.00; P<.001) and substantial concordance with clinician-assigned profiles (Cohen κ=0.70; P<.001). The model operates transparently, without machine learning, and can be embedded into EHR platforms or applied manually using a visual framework. It enables individualized risk estimation, phenotype-driven treatment planning, and population-level case identification-particularly in settings with limited access to traditional diagnostic tools. Conclusions: Seleida provides a reproducible and interpretable framework for COPD control monitoring using high-frequency prescribing data. Its transparent logic, low data burden, and interoperability enable integration across diverse digital infrastructures, including resource-limited settings. By supporting both individualized care and population-level risk stratification, Seleida bridges predictive analytics with real-world clinical decision-making. Ongoing multicenter validation will determine its generalizability, clinical impact, and cost-effectiveness at scale.
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 | 0.018 | 0.051 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.001 | 0.001 |
| Science and technology studies | 0.001 | 0.001 |
| Scholarly communication | 0.002 | 0.001 |
| Open science | 0.002 | 0.002 |
| Research integrity | 0.001 | 0.001 |
| Insufficient payload (model declined to judge) | 0.001 | 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".