Minimally invasive thoracoscopically-guided right minithoracotomy versus conventional sternotomy for mitral valve repair: a multicentre randomised controlled trial (UK Mini Mitral)

Design: multi-centre, randomised controlled trial including an internal pilot. Setting: UK cardiac surgery units. Literature review: Cochrane, PubMed, ISRCTN, clinicaltrials.gov had no relevant trials. Population: Adults with degenerative mitral valve disease requiring isolated Mitral Valve repair (MVr), able to provide written informed consent, fit for surgery. Exclusions: concomitant surgery (except patent foramen ovale or atrial fibrillation), acute infective endocarditis, redo surgery, emergency or salvage surgery. Health Technology: MVr via thoracoscopically-guided right minithoracotomy. Comparator: MVr via median sternotomy. Measurement of costs and outcomes: Primary outcome: return to usual activity using change in SF-36v2 physical functioning scale (1) at 12 weeks. Primary economic outcome: cost/QALY gained after one year of treatment. Secondary outcomes include: cardiac function assessed echocardiographically (by central blinded assessor) post-operatively and at 52 weeks; operative results, peri-operative complications, length of hospital stay, NYHA class (5); resource use, mortality and mitral valve related events; physical function (SF-36v2(1)), level of physical activity, quality of sleep (accelerometer (6-10)), quality of life, (SF-36v2 & ED-5D-5L (1,2-4)). Patients will consent to follow-up beyond 12 months using hospital episode statistics (HES), informing clinical and economic models extrapolating outcomes to 4.5 years. Primary analysis of change in SF-36v2 at 12 weeks using intention to treat principles will use a general linear model accounting for intervention, baseline scores, valve pathology and concomitant surgery. Secondary analysis will be: continuous outcomes by general linear model, binary outcomes by generalised estimating equation, time to event outcomes by log-rank test and frailty modelling. Analysis of repeated measures will account for intra-patient correlation. Hierarchical modelling of patients, surgeons and centres, will provide an estimate of treatment effect. Base case economic evaluation will estimate the effect of surgery using bootstrapped, bivariate regression modelling of costs and QALYs adjusted as per the primary analysis. Multiple imputation will sensitise incomplete data. All outcomes will also be analysed using simple statistics. Sample size: Assuming alpha of 5% and 90% power, 382 patients are required to detect a minimally clinically important difference of 10 points in the SF-36v2(1) physical functioning scale at 12 weeks (assuming a SD of 30 (11)). Patients will be recruited from 4 sites to enable recruitment to target. Allowing for attrition, 400 patients will be randomised. Timetable: Prior to grant: 6 months set-up; protocol and study material development, ethics and governance, trial database. Data access contracts. 0-6 months: internal pilot, recruitment, surgery, assessments at 4 centres. Performance against stopping guidelines evaluated. 7-55 months: recruitment, surgery, trial assessments and follow-up using HES data, validated with data from the National Institute for Cardiovascular Research (NICOR) and medical records by an expert panel. 55-60 months: final expert panel, final data cleaning, analysis, report writing, dissemination, consultation on further collection of follow-up data. Expertise: a track record of working together on successful nationally-funded cardiac surgery trials; together we have key clinical, academic and trials expertise.