Clinical Evaluation Report¶

This is a lightweight Clinical Evaluation Report (CER) for the RCPCH Digital Growth Charts (dGC) Platform, drawn together from existing evidence. It assesses whether the Platform achieves its Intended Purpose safely and effectively, as required in support of the Essential Requirements under the UK Medical Devices Regulations 2002.

1. Scope¶

This CER covers the dGC Platform as defined in the Intended Purpose: the API server and associated charting components that calculate and present childhood growth parameters (centiles and SDS) against recognised growth references.

2. Device description and intended purpose¶

See the canonical Intended Purpose statement. In summary, the Platform is decision-support software that assists a trained healthcare professional by calculating and plotting a child's growth measurements against the clinically appropriate reference. It is not diagnostic and is one input among many in a clinical assessment.

3. Clinical background and state of the art¶

Plotting a child's measurements on a population growth reference is long-established, standard clinical practice in child health. The references used by the Platform are the recognised national and international standards:

• UK-WHO growth references (combining WHO standards for early childhood with UK data);
• UK90 reference;
• specialist Trisomy-21 (Down syndrome) and Turner syndrome references for use where those conditions have been diagnosed.

These references are constructed using the LMS method (Cole & Green), the internationally accepted statistical method for summarising the distribution of a measurement by age as three smooth curves (L, M, S) from which any centile or SDS can be derived. The state of the art is well-defined, stable and published in the peer-reviewed literature; see Growth references and Growth chart papers.

The Platform's contribution is to make these references available digitally and reproducibly via an API, removing the transcription and plotting errors associated with manual chart use, while keeping interpretation firmly with the clinician.

4. Evaluation method and evidence sources¶

The clinical evidence for the Platform comes from four complementary sources:

1. Validity of the underlying references — the references are themselves published, peer-reviewed clinical standards in routine national use; their clinical validity is established independently of this Platform.
2. Computational verification — evidence that the software returns the values the references define (the test harness).
3. Expert clinical oversight — governance of the choice and application of references by a standing expert clinical body.
4. Post-market experience — real-world use and feedback (see Post-Market Surveillance).

5. Clinical evidence¶

5.1 Computational fidelity — the static test harness¶

The single most important piece of evidence is the ~4000-child static test harness. Because the LMS calculation is deterministic, a fixed dataset of children with known expected outputs provides mathematically reproducible proof that the API returns the correct centile/SDS for each case. The harness runs in continuous integration, so any regression introduced by a code or dependency change is detected before release. This provides a level of objective, repeatable accuracy evidence that is difficult to achieve for many other classes of medical software (see Software Lifecycle).

5.2 Statistical and clinical validation¶

The Platform and its underlying rcpchgrowth library have been developed and reviewed with the involvement of leading subject-matter experts, including Professor Tim Cole (originator of the LMS method and a principal author of the UK growth references) and consultant paediatric input from Dr Simon Chapman (paediatric diabetes and endocrinology). This ongoing statistician and clinician involvement provides expert assurance that the references are implemented and applied correctly.

5.3 Clinical governance and oversight¶

Throughout the project, clinical oversight and the direction of safety evaluation were provided by the formal Growth Charts Project Board, a clinical expert group comprising senior RCPCH leadership and eminent clinicians in the field of growth (see the team). The Project Board decided all changes to the nature of the growth charts (for example, the application of gestational-age correction throughout the chart). This governance function is being formalised and strengthened: in 2026–27 it is being replaced by an official RCPCH Growth Charts Committee, providing a durable, accountable body for ongoing clinical oversight.

5.4 Open peer review¶

The entire codebase, including the calculation library, is open source. This enables independent replication and challenge of the calculations and testing by external specialists — a continuous, open analogue of academic peer review — and supports ongoing engagement with the international state of the art, including recent activity concerning WHO growth references in the underlying library.

5.5 Post-market experience¶

To date there have been no reports of inaccuracy or of methodological or mathematical error in the growth chart calculations. Feedback channels and the surveillance approach are described in the Post-Market Surveillance plan.

6. Benefit–risk assessment¶

The clinical benefits — accurate, reproducible, transcription-error-free calculation and presentation of growth parameters against validated national references, embedded directly in clinical systems — are clear and directly support safe child health monitoring.

The principal residual risks (an incorrect value being returned, or the service being unavailable) are documented in the Hazard Log and rated acceptable after mitigation, owing to the deterministic test harness, the first-party validated calculation library, the conspicuousness of gross errors against a child's trajectory, and the availability of fallback methods. Interpretation always rests with a trained clinician who treats the chart as one input among many.

The benefit–risk balance is therefore favourable.

7. Conclusion¶

On the available evidence, the dGC Platform achieves its intended purpose: it correctly and reproducibly computes recognised, peer-reviewed national growth references and presents them to support clinical assessment, with a favourable benefit–risk profile and robust, objective verification of computational accuracy.

8. Ongoing clinical evaluation¶

This CER is a living document. It is reviewed:

• at each management review;
• on establishment of the RCPCH Growth Charts Committee (2026–27) and thereafter under its governance;
• when a new growth reference is added or an existing calculation method materially changes;
• in response to any post-market signal indicating a possible inaccuracy.

The Git commit history of this file is its version record.