Systematic prospective data mining in everyday practice - the NHS approach for proton beam therapy

Abstract

Purpose or Objective The rationale for proton beam therapy (PBT) use combines reduced toxicity and potential dose escalation, as supported by dosimetric studies comparing PBT with photon radiotherapy. However, questions remain of whether, and by how much, this leads to clinical benefit for patients. Whilst awaiting results from appropriately powered randomised trials, and to supplement those, alternative complementary research methodologies can strengthen the evidence base. This report aims to detail the approach taken by the first United Kingdom (UK) National Health Service (NHS)-funded high-energy PBT centre at The Christie NHS Foundation Trust in Manchester to evaluate PBT by utilising each patient’s detailed data, ensuring that ‘every patient counts’. Material and Methods The challenges with PBT data generation, collection and analysis are described, as well as the different approaches taken by jurisdictions internationally to address these challenges, including the European Particle Therapy Network (EPTN) and the Pediatric Proton/Photon Consortium Registry (PPCR). The rationale for the strong emphasis on outcomes data collection by the NHS with the establishment of the Proton Clinical Outcomes Unit (PCOU) which collects, curates and analyses PBT outcomes data prospectively for Christie-treated patients and retrospectively for UK patients treated overseas, is outlined in detail. The PCOU’s unique features in regard to its digital solution, its integration of patient-reported outcomes and its novel relationship with clinical trials and evaluative commissioning are highlighted. Results The Christie electronic patient record was developed as a versatile platform for ‘outcomes tracking’ as well as clinical care. A number of electronic eForms were developed for comprehensive data capture, using data tree and conditional logic approaches, where answers prompt further questions specific to tumour site and patient pathway time point. Clinicians are guided towards tailored questions, improving clinical assessment quality and completeness. Input data generates individualised outputs, e.g. letters, summaries, surveillance plans. Periodic audits of recorded data quality identify errors and outliers and provide user feedback. Patient reported outcomes are collected and outcomes data are correlated with dosimetric DICOM data to support radiation dose modelling. The outcomes dataset and associated outcomes analysis supports long-term follow-up and evaluation of treatment approaches for tumour control and toxicity. It ensures safety, enables comparison internationally, assists patient selection criteria and supports hypothesis generation for future clinical trials. Conclusion The NHS outcomes tracking through the PCOU is a novel approach at the international level. It can be used as an exemplar for others, not only in PBT, but potentially scaled for other areas of practice beyond radiation oncology.