Derivation and Validation of a Novel Automated Algorithm for Staging Infant Blood Pressures
Bibliographic record
Abstract
Background: Infant blood pressure (BP) is assessed by visual evaluation of sex and age specific BP curves (i.e., “manual staging”). Manual staging is time consuming and error prone, but is the standard method to assess infant BP. We developed and evaluated accuracy of a novel computerized algorithm to stage BP category in infants. Methods: We retrospectively acquired BP data in electronic health records (EHRs) from infants hospitalized at a quaternary healthcare center between June 2018-August 2019. Infants <1 year old with paired systolic/diastolic BP were included. First or last admission BP was randomly selected for evaluation. An algorithm to estimate published BP curves was created by digitizing age and sex-based infant BP cutoff curves into a series of points, then deriving each curve's equation of best fit via regression. All BP’s were staged manually by two raters (normal, elevated or hypertensive [stage 1 or 2]). The algorithm was evaluated for agreement (using Cohen’s kappa and % agreement) with manual staging. Results: Of 2407 BP measurements, 1304 patients were staged as normal, 298 elevated BP, and 805 hypertensive by manual staging (inter-rater agreement for manual staging was kappa 0.82 [0.58-1.00]). By the algorithm, 1368 had normal BP, 323 elevated BP and 716 hypertensive. Agreement between manual and algorithm classification was 89.4% (kappa 0.83 [0.82-0.85]). Discrepancy between manual vs. algorithm BP staging was noted for 255 BPs. Five BPs (2% of errors) were incorrectly staged by the algorithm and correctly staged by manual staging; 63 BP’s (24.7% of errors) were incorrectly staged by manual staging and correctly staged by the algorithm. The remaining discrepancies (73.3% of errors) were BPs on/near the curve at the border of normal vs. abnormal BP. The algorithm has been made into a ShinyApp (developed by A. Khondker), available at https://sickkidsnephrology.shinyapps.io/InfantHypertension/ (Figure 1). Conclusion: The new infant BP staging algorithm has strong agreement with manual staging and may be used to stage BP in infants in the clinical setting and within EHRs.
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 | 0.000 | 0.000 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 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 | 0.000 |
| 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".