Breakout 1: How do we make the UK world leading for COVID-19 clinical trials? This session will start at 10:45 BST. Chairs: • Martin Landray, Professor of Medicine and Epidemiology, University of Oxford & Research Director, HDR UK Oxford • Richard Torbett, Chief Executive, ABPI Panellists: • Janet Valentine, Director, Clinical Practice Research Datalink • Mark Toms, Chief Scientific Officer UK, Novartis Pharmaceuticals UK Ltd • Janet Frost, Patient and public representative This session will start at 10:45 BST. Please use the Q&A function to ask questions to speakers. You are welcome to comment using the chat function, but we cannot guarantee this will be monitored.
Randomised Evaluation of COVID-19 Therapy: the RECOVERY trial Martin Landray University of Oxford, UK on behalf of the RECOVERY trial investigators www.recoverytrial.net
Randomisation & rational treatment of COVID-19 Huge therapeutic uncertainty - Many candidates - Many opinions (from many sources) - No reliable data (uncontrolled case series, retrospective association studies) Addressing the treatment challenge - Unlikely to be a single “big win” but moderate benefits plausible - For example, reducing hospital mortality by one-fifth would be important - In future, combining several effective drugs may produce big effects
Selection of treatments Prioritisation principles: - Potentially effective (based on prior pre-clinical & clinical data) - Major safety issues understood - Sufficient treatment available for large-scale recruitment - Potential to rapidly scale up as a clinical treatment (if shown to be effective) Three broad categories: - Anti-viral treatments (lopinavir-ritonavir, hydroxychloroquine) - Immunomodulatory treatments (corticosteroid, azithromycin, tocilizumab) - Treatments targeted at SARS-CoV-2 (convalescent plasma)
Eligibility • Hospitalised (regardless of age, gender, ethnicity, co-morbidity, location) • SARS-CoV-2 infection (clinically suspected or laboratory confirmed) • No medical history that might, in the opinion of the attending clinician, put the patient at significant risk if he/she were to participate in the trial Note: If one or more of the treatment arms is not available or believed, by the attending clinician, to be contraindicated (or definitely indicted), then the patient can be randomised between the remaining arms
Informed consent • Simple information sheet (2 pages) • Option for witnessed consent • if participant cannot read or sign for themselves • If infection control procedures do not allow ICF out of the ‘red zone’ • Option for legal representative • if patient lacks capacity
Randomisation • Simple data entry (online) • Patient details • Inclusion criteria • Key co-morbidities • Any treatments for which the patient is not suitable • Any treatments not available in hospital pharmacy • Randomly assigns participant between suitable and available treatments • Doctor prescribes allocated treatment on usual hospital drug chart
Follow-up • Simple on-line form completed by research nurses • Which treatments did the patient receive • COVID-19 test result • Discharge status & date • Use of ventilation • Linkage to national health data sources • Vital status, death certificate • Coded hospital episode statistics (diagnoses, procedures) • Intensive Care audit data, SARS-CoV-2 PCR laboratory results • Primary care, national outpatient prescribing data • Permission to follow-up via record linkage for up to 10 years
Study outcomes • Primary outcome: All-cause mortality at 28 days • Secondary outcomes: Time to discharge from hospital Use of mechanical ventilation/ECMO or death (among those not on ventilation or ECMO at baseline) • Subsidiary outcomes: Cause-specific mortality Use of renal dialysis or haemofiltration Serious cardiac arrhythmia (recorded in a subset) Use of ventilation (overall and by type) Duration of ventilation (overall and by type)
Recruitment progress Total recruitment: 11,000 at 176 hospitals
Characteristics at main randomisation Male sex 6786 (63%) Age 66 (SD 16) Days since symptom onset 9 (IQR 5-13) Severity of disease No oxygen required 2622 (24%) Supplemental oxygen only 6633 (62%) Ventilation/ECMO 1502 (14%) Prior disease Diabetes 2896 (27%) Cardiovascular disease 2903 (27%) Chronic lung disease 2359 (22%)
Preliminary result - Hydroxychloroquine • 4 June – DMC advised unblinding • 5 June – results made public • No evidence of benefit
Preliminary result - Hydroxychloroquine
Acknowledgements - UK Research & Innovation - National Institute for Health Research - Wellcome Trust - Bill & Melinda Gates Foundation - Department for International Development - Department of Health & Social Care - National Health Service in England, Wales, Scotland, and Northern Ireland - NIHR Clinical Research Network - NHS DigiTrials - NIHR Oxford Biomedical Research Centre - Medical Research Council Population Health Research Unit - Nuffield Department of Medicine - Nuffield Department of Population Health - The very many doctors, research nurses, pharmacists, and R&D managers at over 175 NHS hospitals - And, most importantly, the patients who are participating Full protocol and trial materials available at www.recoverytrial.net
Using electronic health records to improve the efficiency of clinical trials • Dr Janet Valentine, Director Clinical Practice Research Datalink HDRUK One Institute Conference, 16 th June 2020 •
Traditional methods of delivering clinical trials are unsustainable Difficult to find eligible patients 80% of trials are delayed ~70% data duplication between EHRs & 50% fail to achieve target recruitment CT system >50% of protocols require amendments Loss to follow up
Can we use EHR to improve efficiency of trial delivery? Inform protocol design and feasibility Select and locate eligible patients Targeted site selection Reduce screen failures Minimise loss to follow-up
What is CPRD? UK Government health data research service supporting observational & interventional public health and clinical studies by academics, industry and regulators worldwide Services based on > 30 years of collecting longitudinal primary care EHR across UK In house quality checks Daily data collection 23% UK population coverage
What information is in CPRD primary care data? Hospital Demographics attendance and lifestyle EHR GP Laboratory Consultation results
Advantages of near-real time primary care EHR searches • GP = gatekeepers to NHS → manage most chronic conditions • Longitudinal primary care data contains all patient’s medical history • Representative population across geographies, demographics, SES • Rich information → enables sophisticated pre-screening • Daily data updates → dynamic list of eligible patients
Precision and timeliness in the Primary Care record Event data Adherence Laboratory data Treatment data Receiving treatment Recent CVD event Factors likely to limit HbA1c 6.5% - (e.g. MI) within 8 adherence to study with a stable dose 10.0% (47 - 86 (≥12 weeks) of a weeks of treatment or procedures mmol/mol) within qualifying statin screening visit eg substance abuse or past 30 days dementia
Traditional approaches to locating suitable patients Print Local Pool of known Social media databases advertising clinicians
EHR-enabled patient location and recruitment across the UK 15+ million patients High quality GP reviews Central search of CPRD network of GP invites only patients contact CPRD primary 1900 GP eligible patients suitable patients study centre care database for practices for suitability eligible patients 2 weeks
Real time, modifiable EHR search based on clinical feedback Phase 2 clinical trial investigating treatment for Major Depressive Disorder. Complex EHR selection criteria • Diagnosis of MDD less than 18 months • SSRI/SNRI prescription within the past 12 months • Excluded if diagnosis in past year of psychotic disorder, bipolar disorder Modified search list in 24 After 2 days noted 50% Rejection rate hours by restricting patient rejection rate on reduced to 3% inclusion timeframe GP review Faster review, high quality patients
Impact on eligibility for Oxford COVID-19 Vaccine Phase II/II Trial 70+ cohort Proportion of original Criteria cohort (%) Age 70+ 100% Cancer record in last two years 92.0% Any history of ischaemic stroke 87.2% Any history of renal transplant 87.1% Any history of renal dialysis 87.0% Cohort based Any history of cardiac surgery 82.6% Any history of alcoholic liver disease 82.4% on selection Any history of Crohn’s disease 82.0% criteria Current anticoagulant use 75.5% Any history of an immune disorder 75.4% Any history of a bleeding disorder 75.3% Any history of anaphylaxis 75.1% Any history of angioedema 74.9% Any history of severe mental illness 74.0% Any history of dementia, Alzheimer’s or Parkinson’s disease. 69.3% Live within a 20 mile radius of Study Site Refining the Have had a flu vaccination within 24 months http://www.isrctn.com/ISRCTN9 search 0906759 Ethnicity
CPRD data-enabled post authorisation vaccine safety monitoring Linkage to Drug vaccine Patient safety secondary Near real time interactions outcomes care & surveillance death data
Recommend
More recommend
