nabu.science
MethodologyAboutContact

 

  • Snapshot
  • Key Claims
  • Strengths & Limitations
  • Commentary
  • From the Authors
  • Author Note / Field Feedback
  • Dimension breakdown
  • Impact Profile
Free · About 5 minutes
Report #NBU-B1DFFA25·Framework v5.9·Pipeline v0.4·Evaluated 08 Oct 2026·OECD: 3.2 Clinical medicine·Subfield: Artificial Intelligence·Basis: Full text·Type: Machine learning or AI study

Clinically-applicable prediction of hospital stay and patient similarity retrieval in paediatric cardiology using machine learning.

Rigny et al.Rigny, Biggart, Zakka et al.·2026·Nature communications·10.1038/s41467-026-73021-3 ↗

Primary claim

a Random Forest model achieves the best length of stay prediction accuracy (0.88 ± 0.02), outperforming clinicians.

Tentative · Table 1 identifies a higher-accuracy model, while the small clinician comparison used BioClinical-BERT without a superiority test
In brief

Prospective NHS testing gives this paediatric cardiology pipeline practical value, but contaminated preprocessing and internally inconsistent retrieval results limit confidence.

ReliabilityCaveats4 of 4 checks completeRead the 2 entries ↓
Fragile — the claim's weakest requirement is that the effect exists at all. Experimental. Every point below is traceable to a quoted passage, but this forecast has not yet been validated against known replication outcomes.
Abstract

Paediatric cardiology presents challenges due to the rarity and complexity of conditions like congenital heart disease. Using retrospective electronic healthcare records from 1,522 Great Ormond Street Hospital cases, we benchmark machine learning models to predict length of stay and retrieve similar patient cases. BioClinical-BERT is used for embedding-based retrieval, while a Random Forest model achieves the best length of stay prediction accuracy (0.88  ± 0.02), outperforming clinicians. The Random Forest model shows mean precision and sensitivity of 0.77  ± 0.03 and 0.76  ± 0.04 during NHS silent deployment (1,052 admissions). K-means identifies three clinically distinct subgroups. Cosine similarity retrieval reveals diagnosis-driven top matches, while complications dominate broader sets. In a 25-case intensive care unit pilot, clinician-rated utility improves from 4.23  ± 2.42 to 4.41  ± 2.16 (scale 0-10). Our models surpass clinician performance in length of stay prediction and show promise for case retrieval, supporting data-driven decision-making in paediatric cardiology and beyond.

Key Claims

Primary finding · Tentative

a Random Forest model achieves the best length of stay prediction accuracy (0.88 ± 0.02), outperforming clinicians.

Fragile — the claim's weakest requirement is that the effect exists at all. Experimental. Every point below is traceable to a quoted passage, but this forecast has not yet been validated against known replication outcomes.

Table 1 identifies a higher-accuracy model, while the small clinician comparison used BioClinical-BERT without a superiority test

Abstract; Results, Table 1; Fig. 2
Quote verified
Secondary · Supported

The Random Forest model shows mean precision and sensitivity of 0.77 ± 0.03 and 0.76 ± 0.04 during NHS silent deployment (1,052 admissions).

prospective single-site monitoring supports the direction, with limited duration and a minor accuracy inconsistency

Abstract; Results, Silent Deployment Performance; Fig. 3
Quote verified
Secondary · Tentative

K-means identifies three clinically distinct subgroups.

three clusters were recovered, but clinical separation was modest and only one LoS contrast was significant

Abstract; Results, Patient Similarity Validation; Fig. 4; Table 4
Quote verified
Secondary · Tentative

Cosine similarity retrieval reveals diagnosis-driven top matches, while complications dominate broader sets.

Table 3 contradicts the stated retrieval pattern and contains invalid uncertainty values for bounded metrics

Abstract; Discussion, patient-similarity paragraph; Table 3
Quote verified
Secondary · Tentative

In a 25-case intensive care unit pilot, clinician-rated utility improves from 4.23 ± 2.42 to 4.41 ± 2.16 (scale 0-10).

the uncontrolled pilot cannot establish improvement from a 0.18-point change with standard deviations above two points

Abstract; Results, Pilot Study in Clinical Care; Table 5
Quote verified

Derived from the full evaluation — not a separate score.

Strengths · Limitations

Strengths

#contribution

Prospective monitoring in a live NHS setting

The Random Forest pipeline processed 1,052 admissions over six months without exposing predictions to clinicians. This supplies prospective performance evidence rarely available in retrospective EHR studies.

↳ Results, Silent Deployment Performance; Fig. 3

#methodological rigour

Broad comparative model evaluation

Decision Tree, Random Forest, Gradient Boosting, LSTM, RoBERTa, and BioClinical-BERT variants were evaluated, with patient-level stratification and upsampling restricted to training folds.

↳ Methods, Model Comparison and Rule-based Approaches; Results, Table 1

#application pathway

Concrete clinical users and workflow

The study connects categorical LoS prediction to bed planning and tests similar-case presentation during ICU consultations, establishing a visible application pathway.

↳ Discussion, paragraph 2; Methods, Pilot Study in Clinical Care

Limitations

#methodological rigour

Retrieval results are internally inconsistent

Table 3 contains impossible standard deviations for several bounded metrics, and its values do not support the claim that diagnosis drives top retrieval while complications dominate broader sets. This materially limits confidence in the retrieval conclusions.

↳ Results, Table 3; Discussion, patient-similarity paragraph

#methodological rigour

Held-out data entered model preprocessing

Training and test datasets were combined for encoding and standardisation, allowing held-out information to influence preprocessing. The prospective deployment partly tests real-world performance but does not repair the retrospective comparison design.

↳ Methods, Model Comparison and Rule-based Approaches

#reporting

Headline comparative claims exceed evidence

Table 1 reports BioClinical-BERT accuracy of 0.89 versus 0.88 for Random Forest, while Fig. 2 compares clinicians with BioClinical-BERT on only 50 cases and reports no significance test for superiority.

↳ Abstract; Results, Table 1; Fig. 2 caption

Does this match your read? Add your assessment ↓

Nabu Commentary

The six-month silent deployment provides the paper's strongest evidence and makes the contribution more consequential than a retrospective benchmark alone. Methodological confidence is reduced because training and test data were combined for encoding and standardisation, and the downstream retrieval evaluation lacks component ablations or a retrieval baseline. Table 3 also conflicts with the narrative account of retrieval and reports invalid uncertainty values for several bounded metrics. The manuscript is well structured and engages relevant deployment literature, but claims of clinician superiority, improved pilot utility, and broad transferability require substantial qualification.

Evaluate another paper

Free · About 5 minutes · Research preview

Author Note / Field Feedback

Nabu’s assessment, alongside the field’s view.

Are you an author of this paper?

Dimension breakdown

Evidence Score

Out of 5
012345
2.8±0.2

Contribution
3.2

The integrated LoS, retrieval, silent-deployment, and pilot workflow constitutes a useful applied contribution in paediatric cardiology, although its modelling ingredients are established. The advance is limited by overstated comparative, clustering, retrieval, and utility claims.

Abstract; Results, Fig. 3, Table 3, Fig. 4 and Table 5

“During the silent mode deployment, the model achieved a mean accuracy of 0.776.”

Methodological Rigour
2.3

Patient-level stratification, multiple model families, training-fold-only upsampling, and prospective monitoring support the design. Methodological Rigour is constrained by joint train/test preprocessing, incomplete tuning detail, missing retrieval ablations, and unstable or inconsistent downstream evaluations.

Methods, Model Comparison and Rule-based Approaches; Results, Tables 1, 3 and 5

“both training and test datasets were combined for consistent encoding. The encoded features were then standardised.”

Reporting
3.1

The manuscript follows a recognisable progression from model development through deployment, retrieval, and pilot evaluation. Its abstract nevertheless misidentifies the best model, conflates the model used in the clinician comparison, and frames a small utility-score change as improvement.

Abstract; Results, Table 1 and Fig. 2; Table 5

“a Random Forest model achieves the best length of stay prediction accuracy (0.88 ± 0.02), outperforming clinicians.”

Positioning
3.2

The Introduction and Discussion engage relevant patient-similarity, clinical-AI, and silent-deployment literature and acknowledge the single-centre and objective-function limitations. Claims of strong generalisability and transferability remain more expansive than the single-specialty evidence supports.

Introduction, final paragraphs; Discussion, paragraphs 3, 6 and 8

“the proposed methodology and model architecture are transferable and scalable to other hospital settings.”

Reliability

Caveats4 of 4 checks


Results checkNoted
  • a Random Forest model achieves the best length of stay prediction accuracy (0.88 ± 0.02), outperforming clinicians: "The Random Forest Model performed best, demonstrating an overall accuracy of 88% across both stay categories." (Discussion, model-performance paragraph) vs "The binary classification approach yielded promising results, achieving an overall accuracy of 88% for the Random Forest Classifier and 89% for the BioClinical-BERT model." (Discussion, model-performance paragraph). Comparing the reported accuracies gives BioClinical-BERT a one-percentage-point advantage over Random Forest, contradicting the designation of Random Forest as the accuracy leader. → Not supported by the paper's own data
  • a Random Forest model achieves the best length of stay prediction accuracy (0.88 ± 0.02), outperforming clinicians: "while a Random Forest model achieves the best length of stay prediction accuracy (0.88  ± 0.02), outperforming clinicians." (Abstract) vs "Performance metrics (accuracy, precision, sensitivity, F1) of the BioClinical-BERT model, fine-tuned on the MIMIC dataset and subsequently on the GOSH dataset" (Figure 2 caption). The Abstract attributes clinician outperformance to Random Forest, whereas the caption identifying the model in the clinician comparison specifies BioClinical-BERT. → Not supported by the paper's own data
  • Cosine similarity retrieval reveals diagnosis-driven top matches, while complications dominate broader sets: "Diagnosis strongly influences early retrieval (@1), while other clinical factors,including complications, play a more prominent role at larger retrieval sets (@3, @5, @10)." (Discussion, patient-retrieval interpretation) vs "Primary diagnosis0.23 ± 0.350.25 ± 0.600.31 ± 0.310.33 ± 0.30All diagnoses0.47 ± 0.350.51 ± 0.160.53 ± 0.140.65 ± 0.60Procedures0.87 ± 0.600.85 ± 0.600.84 ± 0.580.65 ± 0.61Proportion complications0.52 ± 0.310.50 ± 0.210.48 ± 0.160.47 ± 0.13" (Table 3, diagnosis, procedure and complication rows). At @1, procedure similarity is 0.87 versus 0.23 for primary diagnosis and 0.47 for all diagnoses, while complication similarity decreases from 0.52 at @1 to 0.47 at @10 rather than becoming dominant. → Not supported by the paper's own data
  • Cosine similarity retrieval reveals diagnosis-driven top matches, while complications dominate broader sets: "Primary diagnosis0.23 ± 0.350.25 ± 0.600.31 ± 0.310.33 ± 0.30All diagnoses0.47 ± 0.350.51 ± 0.160.53 ± 0.140.65 ± 0.60Procedures0.87 ± 0.600.85 ± 0.600.84 ± 0.580.65 ± 0.61" (Table 3). For observations bounded by 0 and 1, sample SD is at most sqrt(n·mean·(1−mean)/(n−1)); for n≥3 this bounds the flagged primary-diagnosis, all-diagnoses and first three procedure SDs by 0.530, 0.584, 0.412, 0.437 and 0.449 respectively, and n=2 permits still smaller maxima, excluding the reported 0.60, 0.60, 0.60, 0.60 and 0.58. → Effect on result unclear
  • a Random Forest model achieves the best length of stay prediction accuracy (0.88 ± 0.02), outperforming clinicians: "BioClinical BERT (FT-MIMIC-GOSH)0.77 ± 0.140.94 ± 0.020.82 ± 0.050.79 ± 0.090.94 ± 0.020.77 ± 0.140.92 ± 0.930.93  ± 0.020.89 ± 0.04" (Table 1, BioClinical BERT (FT-MIMIC-GOSH), long-stay precision). A sample precision mean of 0.92 on a bounded 0–1 measure permits SD no greater than sqrt(2×0.92×0.08)=0.384 for any sample size of at least two, so the printed SD of 0.93 is impossible. → Effect on result unclear
  • The Random Forest model shows mean precision and sensitivity of 0.77 ± 0.03 and 0.76 ± 0.04 during NHS silent deployment (1,052 admissions): "During the silent mode deployment, the model achieved a mean accuracy of 0.776 (±0.021)." (Results, Silent deployment performance) vs "It demonstrated consistent performance, maintaining a high overall accuracy of 0.766 ± 0.021." (Discussion, silent deployment). The two reported mean accuracies for the same deployment differ by 0.776−0.766=0.010, and the supplied text does not identify which is correct. → Effect on result unclear
Reporting and transparencyNoted

Post-hoc testing, pilot feedback denominators and the pilot rating scale need clarification; the deployment end date and procedure-coding label contain correctable reporting errors.

Ethics approval: declared

Reference IntegrityClean

61 of 61 checkable references verified

77 references in manuscript 16 have no canonical index record — counted, but not index-checkable 2 references confirmed by manual review

References not verified automatically are resolved manually within 24 hours.

Post-Publication RecordClean

No retraction or expression of concern found in Crossref (includes Retraction Watch).

Last checked: 17 Oct 2026

Sources: Crossref (incl. Retraction Watch) ✓

Impact Profile

Impact areaHealth · Congenital · Congenital birth defects
Type of advanceMethodological
Route to useNamed
AdvanceExtends
BeneficiariesLarge
CitationsToo early
Policy · patentNot yet traced
ReachedNot reachedNot assessable or unverified
ReadinessValidated
Foundational
Controlled
Piloted
Validated
Implementable
FoundationalImplementable

The length-of-stay prediction was confirmed on prospective real-world NHS data during a six-month silent deployment across 1,052 admissions, with performance consistent with the retrospective test set, plus outperformance of a clinician panel.

From the paper
“By using ML-based CDS, healthcare teams can more effectively allocate resources and interventions to high-risk patients.”

The paper names healthcare teams as the actors who would use this ML decision support to allocate resources and target high-risk patients, and clinicians receiving similar cases to support decision-making.

Pathway

Paediatric cardiology clinicians and hospital bed-management teams could use admission-day stay predictions and similar-case retrieval to plan beds, staffing and complication monitoring.

Barrier

Evidence comes from one specialist centre with silent, non-clinician-facing testing and only modest pilot utility scores, so prospective multi-site clinical evaluation is still needed.

Evaluated 13 Nov 2026 · GBD 2023 · glm_v1_3_3 · impact_profile_v1_3_3

Report an issue →

AI-generated, human-governed. Something look off? Contact us to request a review.

Keep reading

Related evaluations

Other papers Nabu has evaluated blind, starting with the same field.

  1. Early sepsis prediction using a hybrid LSTM-GAT model: a study on the PhysioNet 2019 dataset

    Khorram, Kouchaki · 2026 · Clinical medicine

    Primary claim

    The LSTM-GAT model achieved an area under the receiver operating characteristic curve (AUROC) of 0.853±0.005, an F1-score of 0.627±0.006 and a specificity of 0.872±0.007, outperforming baseline models.

    Tentative
    Evidence Score
    2.8±0.2
    Reliability
    Caveats
  2. High-fidelity entangling gates and nonlocal circuits with neutral atoms

    2026 · Physical sciences

    Primary claim

    Using a neutral-atom quantum processor, we realized entangling controlled- Z gates with a high-Rabi-frequency smooth-amplitude pulse, employing state-selective readout and qubit reuse for fast calibration, and achieved a fidelity of 99.854(4)%, which improved to 99.941(3)% upon loss postselection

    Supported
    Evidence Score
    4.1±0.1
    Reliability
    Caveats
  3. Retrofitting language models to operate over bytes

    2026 · Computer and information sciences

    Primary claim

    Our byteified models outperformed, on average, all earlier publicly available byte-level LLMs of comparable size.

    Supported
    Evidence Score
    3.9±0.2
    Reliability
    No concerns

How the rubric works →