Nationally Automated Colonoscopy Performance Feedback Increases Polyp Detection: The NED APRIQOT Randomized Controlled Trial.

BACKGROUND & AIMS: Postcolonoscopy colorectal cancer incidence and mortality rates are higher for endoscopists with low polyp detection rates. Using the UK's National Endoscopy Database (NED), which automatically captures real-time data, we assessed if providing feedback of case-mix-adjusted mean number of polyps (aMNP), as a key performance indicator, improved endoscopists' performance. Feedback was delivered via a theory-informed, evidence-based audit and feedback intervention. METHODS: This multicenter, prospective, NED Automated Performance Reports to Improve Quality Outcomes Trial randomized National Health Service endoscopy centers to intervention or control. Intervention-arm endoscopists were e-mailed tailored monthly reports automatically generated within NED, informed by qualitative interviews and behavior change theory. The primary outcome was endoscopists' aMNP during the 9-month intervention. RESULTS: From November 2020 to July 2021, 541 endoscopists across 36 centers (19 intervention; 17 control) performed 54,770 procedures during the intervention, and 15,960 procedures during the 3-month postintervention period. Comparing the intervention arm with the control arm, endoscopists during the intervention period: aMNP was nonsignificantly higher (7%; 95% CI, -1% to 14%; P = .08). The unadjusted MNP (10%; 95% CI, 1%-20%) and polyp detection rate (10%; 95% CI, 4%-16%) were significantly higher. Differences were not maintained in the postintervention period. In the intervention arm, endoscopists accessing NED Automated Performance Reports to Improve Quality Outcomes Trial webpages had a higher aMNP than those who did not (aMNP, 118 vs 102; P = .03). CONCLUSIONS: Although our automated feedback intervention did not increase aMNP significantly in the intervention period, MNP and polyp detection rate did improve significantly. Engaged endoscopists benefited most and improvements were not maintained postintervention; future work should address engagement in feedback and consider the effectiveness of continuous feedback. CLINICAL TRIALS REGISTRY:  www.isrctn.org ISRCTN11126923 .

Using a prognostic medical device for early identification of pressure ulcers: protocol for study design.

BACKGROUND: An objective, physiological measurement taken using a medical device may reduce the incidence of pressure ulcers through earlier detection of problems signs before visual signs appear. Research in this field is hampered by variations in clinical practice and patient-level confounders. AIM: The authors outline key considerations for designing a protocol for a study to assess the efficacy and safety of a prognostic medical device in reducing pressure ulcer incidence in a hospital, including comparators, randomisation, sample size, ethics and practical issues. METHOD: Key issues relating to methodology and ethics are considered alongside a theoretical protocol, which could support future researchers in wound care trials. RESULTS: A prospective, three-armed, multi-centre, stratified cluster-randomised controlled trial is proposed. The third arm is recommended as it is expected that patients will need to be moved for the medical device to be used and repositioning is a preventive strategy. A minimum of 16 200 patients in 33 wards would needed to be recruited to achieve statistical significance. Ethical considerations in terms of consent or assent need to be considered. CONCLUSION: The hypothetical study designed to evaluate the effectiveness of a diagnostic or prognostic medical device in reducing pressure ulcer incidence in secondary care, while accounting for biases, would require large sample sizes and involves risks of inter-operator and inter-device reliability, heterogeneity of users and the vague clinical interpretation of device results. Robust research in this field has the potential to influence or change policy and practice relating to the prevention of pressure ulcers in secondary care.

Minithoracotomy vs Conventional Sternotomy for Mitral Valve Repair: A Randomized Clinical Trial.

Importance: The safety and effectiveness of mitral valve repair via thoracoscopically-guided minithoracotomy (minithoracotomy) compared with median sternotomy (sternotomy) in patients with degenerative mitral valve regurgitation is uncertain. Objective: To compare the safety and effectiveness of minithoracotomy vs sternotomy mitral valve repair in a randomized trial. Design, Setting, and Participants: A pragmatic, multicenter, superiority, randomized clinical trial in 10 tertiary care institutions in the UK. Participants were adults with degenerative mitral regurgitation undergoing mitral valve repair surgery. Interventions: Participants were randomized 1:1 with concealed allocation to receive either minithoracotomy or sternotomy mitral valve repair performed by an expert surgeon. Main Outcomes and Measures: The primary outcome was physical functioning and associated return to usual activities measured by change from baseline in the 36-Item Short Form Health Survey (SF-36) version 2 physical functioning scale 12 weeks after the index surgery, assessed by an independent researcher masked to the intervention. Secondary outcomes included recurrent mitral regurgitation grade, physical activity, and quality of life. The prespecified safety outcomes included death, repeat mitral valve surgery, or heart failure hospitalization up to 1 year. Results: Between November 2016 and January 2021, 330 participants were randomized (mean age, 67 years, 100 female [30%]); 166 were allocated to minithoracotomy and 164 allocated to sternotomy, of whom 309 underwent surgery and 294 reported the primary outcome. At 12 weeks, the mean between-group difference in the change in the SF-36 physical function T score was 0.68 (95% CI, -1.89 to 3.26). Valve repair rates (≈ 96%) were similar in both groups. Echocardiography demonstrated mitral regurgitation severity as none or mild for 92% of participants at 1 year with no difference between groups. The composite safety outcome occurred in 5.4% (9 of 166) of patients undergoing minithoracotomy and 6.1% (10 of 163) undergoing sternotomy at 1 year. Conclusions and relevance: Minithoracotomy is not superior to sternotomy in recovery of physical function at 12 weeks. Minithoracotomy achieves high rates and quality of valve repair and has similar safety outcomes at 1 year to sternotomy. The results provide evidence to inform shared decision-making and treatment guidelines. Trial Registration: isrctn.org Identifier: ISRCTN13930454.

Sample size considerations for the external validation of a multivariable prognostic model: a resampling study.

After developing a prognostic model, it is essential to evaluate the performance of the model in samples independent from those used to develop the model, which is often referred to as external validation. However, despite its importance, very little is known about the sample size requirements for conducting an external validation. Using a large real data set and resampling methods, we investigate the impact of sample size on the performance of six published prognostic models. Focussing on unbiased and precise estimation of performance measures (e.g. the c-index, D statistic and calibration), we provide guidance on sample size for investigators designing an external validation study. Our study suggests that externally validating a prognostic model requires a minimum of 100 events and ideally 200 (or more) events.

Quantifying the impact of different approaches for handling continuous predictors on the performance of a prognostic model.

Continuous predictors are routinely encountered when developing a prognostic model. Investigators, who are often non-statisticians, must decide how to handle continuous predictors in their models. Categorising continuous measurements into two or more categories has been widely discredited, yet is still frequently done because of its simplicity, investigator ignorance of the potential impact and of suitable alternatives, or to facilitate model uptake. We examine three broad approaches for handling continuous predictors on the performance of a prognostic model, including various methods of categorising predictors, modelling a linear relationship between the predictor and outcome and modelling a nonlinear relationship using fractional polynomials or restricted cubic splines. We compare the performance (measured by the c-index, calibration and net benefit) of prognostic models built using each approach, evaluating them using separate data from that used to build them. We show that categorising continuous predictors produces models with poor predictive performance and poor clinical usefulness. Categorising continuous predictors is unnecessary, biologically implausible and inefficient and should not be used in prognostic model development. © 2016 The Authors. Statistics in Medicine published by John Wiley & Sons Ltd.

Sensitivity analyses for partially observed recurrent event data.

Recurrent events involve the occurrences of the same type of event repeatedly over time and are commonly encountered in longitudinal studies. Examples include seizures in epileptic studies or occurrence of cancer tumors. In such studies, interest lies in the number of events that occur over a fixed period of time. One considerable challenge in analyzing such data arises when a large proportion of patients discontinues before the end of the study, for example, because of adverse events, leading to partially observed data. In this situation, data are often modeled using a negative binomial distribution with time-in-study as offset. Such an analysis assumes that data are missing at random (MAR). As we cannot test the adequacy of MAR, sensitivity analyses that assess the robustness of conclusions across a range of different assumptions need to be performed. Sophisticated sensitivity analyses for continuous data are being frequently performed. However, this is less the case for recurrent event or count data. We will present a flexible approach to perform clinically interpretable sensitivity analyses for recurrent event data. Our approach fits into the framework of reference-based imputations, where information from reference arms can be borrowed to impute post-discontinuation data. Different assumptions about the future behavior of dropouts dependent on reasons for dropout and received treatment can be made. The imputation model is based on a flexible model that allows for time-varying baseline intensities. We assess the performance in a simulation study and provide an illustration with a clinical trial in patients who suffer from bladder cancer.

Feasibility study of rehabilitation for cardiac patients aided by an artificial intelligence web-based programme: a randomised controlled trial (RECAP trial)-a study protocol.

INTRODUCTION: Cardiac rehabilitation (CR) delivered by rehabilitation specialists in a healthcare setting is effective in improving functional capacity and reducing readmission rates after cardiac surgery. It is also associated with a reduction in cardiac mortality and recurrent myocardial infarction. This trial assesses the feasibility of a home-based CR programme delivered using a mobile application (app). METHODS: The Rehabilitation through Exercise prescription for Cardiac patients using an Artificial intelligence web-based Programme (RECAP) randomised controlled feasibility trial is a single-centre prospective study, in which patients will be allocated on a 1:1 ratio to a home-based CR programme delivered using a mobile app with accelerometers or standard hospital-based rehabilitation classes. The home-based CR programme will employ artificial intelligence to prescribe exercise goals to the participants on a weekly basis. The trial will recruit 70 patients in total. The primary objectives are to evaluate participant recruitment and dropout rates, assess the feasibility of randomisation, determine acceptability to participants and staff, assess the rates of potential outcome measures and determine hospital resource allocation to inform the design of a larger randomised controlled trial for clinical efficacy and health economic evaluation. Secondary objectives include evaluation of health-related quality of life and 6 minute walk distance. ETHICS AND DISSEMINATION: RECAP trial received a favourable outcome from the Berkshire research ethics committee in September 2022 (IRAS 315483).Trial results will be made available through publication in peer-reviewed journals and presented at relevant scientific meetings. TRIAL REGISTRATION NUMBER: ISRCTN97352737.

Effects of a novel, brief psychological therapy (Managing Unusual Sensory Experiences) for hallucinations in first episode psychosis (MUSE FEP): Findings from an exploratory randomised controlled trial.

Hallucinations are a common feature of psychosis, yet access to effective psychological treatment is limited. The Managing Unusual Sensory Experiences for First-Episode-Psychosis (MUSE-FEP) trial aimed to establish the feasibility and acceptability of a brief, hallucination-specific, digitally provided treatment, delivered by a non-specialist workforce for people with psychosis. MUSE uses psychoeducation about the causal mechanisms of hallucinations and tailored interventions to help a person understand and manage their experiences. We undertook a two-site, single-blind (rater) Randomised Controlled Trial and recruited 82 participants who were allocated 1:1 to MUSE and treatment as usual (TAU) (n = 40) or TAU alone (n = 42). Participants completed assessments before and after treatment (2 months), and at follow up (3-4 months). Information on recruitment rates, adherence, and completion of outcome assessments was collected. Analyses focussed on feasibility outcomes and initial estimates of intervention effects to inform a future trial. The trial is registered with the ISRCTN registry 16793301. Criteria for the feasibility of trial methodology and intervention delivery were met. The trial exceeded the recruitment target, had high retention rates (87.8%) at end of treatment, and at follow up (86.6%), with good acceptability of treatment. There were 3 serious adverse events in the therapy group, and 5 in the TAU group. Improvements were evident in both groups at the end of treatment and follow up, with a particular benefit in perceived recovery in the MUSE group. We showed it was feasible to increase access to psychological intervention but a definitive trial requires further changes to the trial design or treatment.

Feasibility and acceptability of autism adapted safety plans: an external pilot randomised controlled trial.

Background: Autistic people are a high-risk group for self-harm and suicide. There are no evidence-based suicide prevention interventions developed specifically for autistic people. We undertook a pilot feasibility randomised controlled trial of autism adapted safety plans (AASP) to reduce self-harm and suicide for autistic people. Methods: This study took place in the United Kingdom and followed a randomised, two-arm, controlled design. Autistic adults (n = 53, mean age = 39, gender = 49% female, 29% not male or female) were recruited via third sector organisations and self-referral between 11.8.21 and 19.10.22. Participants were randomised without stratification to usual care with or without AASP. The AASP was completed by the autistic adults together with someone trained to support them. Research staff who completed follow-up assessments were blind to participant allocation. Primary outcomes were feasibility and acceptability. Participants were assessed at baseline, 1 and 6 months. Primary data were analysed under the intention to treat principle. Study protocol is published. The trial is closed to new participants. This study is registered with the ISRCTN registry, ISRCTN70594445. Findings: 53 participants consented, 49 were randomised to either AASP with usual care (n = 25) or usual care (n = 24). 68% of participants in the AASP arm were satisfied with the AASP and 41% rated it as useable. Feedback on the AASP and research methods were positive with suggested adaptations to some outcome measures. Retention and completion of outcomes measures in both arms was excellent, as was fidelity of delivery of the AASP. Interpretation: Study progression criteria were met, suggesting that the parameters of a future definitive trial of clinical and cost effectiveness of AASP to reduce self-harm and suicide in autistic adults are achievable, with minor recommended adaptions to outcome measures and AASP. Future research should explore the use of AASP in routine clinical practice. Funding: This study is funded by the NIHR [Public Health Research Programme (NIHR129196)].

Prehabilitation in elective patients undergoing cardiac surgery: a randomised control trial (THE PrEPS TRIAL) - a study protocol.

INTRODUCTION: Prehabilitation prior to surgery has been shown to reduce postoperative complications, reduce length of hospital stay and improve quality of life after cancer and limb reconstruction surgery. However, there are minimal data on the impact of prehabilitation in patients undergoing cardiac surgery, despite the fact these patients are generally older and have more comorbidities and frailty. This trial will assess the feasibility and impact of a prehabilitation intervention consisting of exercise and inspiratory muscle training on preoperative functional exercise capacity in adult patients awaiting elective cardiac surgery, and determine any impact on clinical outcomes after surgery. METHODS AND ANALYSIS: PrEPS is a randomised controlled single-centre trial recruiting 180 participants undergoing elective cardiac surgery. Participants will be randomised in a 1:1 ratio to standard presurgical care or standard care plus a prehabilitation intervention. The primary outcome will be change in functional exercise capacity measured as change in the 6 min walk test distance from baseline. Secondary outcomes will evaluate the impact of prehabilitation on preoperative and postoperative outcomes including; respiratory function, health-related quality of life, anxiety and depression, frailty, and postoperative complications and resource use. This trial will evaluate if a prehabilitation intervention can improve preoperative physical function, inspiratory muscle function, frailty and quality of life prior to surgery in elective patients awaiting cardiac surgery, and impact postoperative outcomes. ETHICS AND DISSEMINATION: A favourable opinion was given by the Sheffield Research Ethics Committee in 2019. Trial findings will be disseminated to patients, clinicians, commissioning groups and through peer-reviewed publication. TRIAL REGISTRATION NUMBER: ISRCTN13860094.

Activation of telomerase by TA-65 enhances immunity and reduces inflammation post myocardial infarction.

Myocardial infarction (MI) accelerates immune ageing characterised by lymphopenia, expansion of terminally differentiated CD8+ T-lymphocytes (CD8+ TEMRA) and inflammation. Pre-clinical data showed that TA-65, an oral telomerase activator, reduced immune ageing and inflammation after MI. We conducted a double blinded randomised controlled pilot trial evaluating the use of TA-65 to reduce immune cell ageing in patients following MI. Ninety MI patients aged over 65 years were randomised to either TA-65 (16 mg daily) or placebo for 12 months. Peripheral blood leucocytes were analysed by flow cytometry. The pre-defined primary endpoint was the proportion of CD8+ T-lymphocytes which were CD8+ TEMRA after 12 months. Secondary outcomes included high-sensitivity C-reactive protein (hsCRP) levels. Median age of participants was 71 years. Proportions of CD8+ TEMRA did not differ after 12 months between treatment groups. There was a significant increase in mean total lymphocyte count in the TA-65 group after 12 months (estimated treatment effect: + 285 cells/µl (95% CI: 117-452 cells/ µ l, p < 0.004), driven by significant increases from baseline in CD3+, CD4+, and CD8+ T-lymphocytes, B-lymphocytes and natural killer cells. No increase in lymphocyte populations was seen in the placebo group. At 12 months, hsCRP was 62% lower in the TA-65 group compared to placebo (1.1 vs. 2.9 mg/L). Patients in the TA-65 arm experienced significantly fewer adverse events (130 vs. 185, p = 0.002). TA-65 did not alter CD8+ TEMRA but increased all major lymphocyte subsets and reduced hsCRP in elderly patients with MI after 12 months.

Adapted suicide safety plans to address self-harm, suicidal ideation, and suicide behaviours in autistic adults: protocol for a pilot randomised controlled trial.

BACKGROUND: Suicide prevention is a national priority for the UK government. Autistic people are at greater risk of experiencing self-harm and suicidal thoughts and behaviours than the general population. Safety plans are widely used in suicide prevention but have not yet been designed with and for autistic people. We developed the first safety plan specifically targeting suicidality in autistic adults: the Autism Adapted Safety Plan (AASP). It consists of a prioritised list of hierarchical steps that can be used prior to or during a crisis to mitigate risk of self-harm and suicidal behaviour. This is a pilot study that aims to assess the feasibility and acceptability of the AASPs and the research processes, including the response rates, potential barriers and reach of AASPs, methods of recruitment, what comprises usual care, and economic evaluation methods/tools. METHODS: This is an external pilot randomised controlled trial of a suicide prevention tool aimed at mitigating the risk of self-harm and suicidal behaviour in autistic adults: AASPs. Participants will be assessed at baseline and followed up 1 month and 6 months later. Assessments include questions about self-harm, suicidality, service use, and their experience of the AASP/taking part in the study. Autistic adults who have a clinical autism diagnosis and self-reported history of self-harm, suicidal thoughts, or suicidal behaviours within the last 6 months will be invited to take part in the study. Informed consent will be obtained. Participants will be recruited via community and third sector services (including community settings, autism charities, and mental health charities). They may also "self-refer" into the study through social media recruitment and word of mouth. Ninety participants will be randomised to either develop an AASP or receive their usual care in a 1:1 ratio. DISCUSSION: The present study will provide an evaluation of the suitability of the processes that would be undertaken in a larger definitive study, including recruitment, randomisation, methods, questionnaires, outcome measures, treatment, and follow-up assessments. TRIAL REGISTRATION: ISRCTN70594445, Protocol v4: 8/2/22.

Long-term outcomes of mini-sternotomy versus conventional sternotomy for aortic valve replacement: a randomized controlled trial.

OBJECTIVES: Aortic valve replacement (AVR) for severe symptomatic aortic stenosis is one of the most common cardiac surgical procedures with excellent long-term outcomes. Multiple previous studies have compared short-term outcomes of AVR with mini-sternotomy versus AVR with conventional sternotomy. We have previously reported the results of the randomized MAVRIC trial, which aimed to evaluate early postoperative morbidity among patients undergoing mini-sternotomy and conventional sternotomy AVR. We now report the long-term all-cause mortality, reoperation, MACE outcomes and echocardiographic data from this trial. METHODS: The prospective, randomized, single-centre, single-blind MAVRIC (manubrium-limited mini-sternotomy versus conventional sternotomy for aortic valve replacement) trial compared manubrium-limited mini-sternotomy and conventional median sternotomy for the treatment of patients with severe aortic stenosis. The previously reported primary outcome was the proportion of patients receiving red cell transfusion postoperatively and within 7 days of the index procedure. Currently reported exploratory analyses of a combined long-term all-cause mortality and reoperation were compared between groups via the log-rank test. Sensitivity analyses reviewed individual components of the combined end point. The primary analysis and long-term exploratory analyses were based on an intention-to-treat principle. RESULTS: Between March 2014 and June 2016, 270 patients were enrolled and randomized in a 1:1 fashion to undergo mini-sternotomy AVR (n = 135) or conventional median sternotomy AVR (n = 135). At the median follow-up of 6.1 years, the composite outcome of all-cause mortality and reoperation occurred in 18.5% (25/135) of patients in the conventional sternotomy group and in 17% (23/135) of patients in the mini-sternotomy group. The incidence of chronic kidney disease, cerebrovascular accident and myocardial infarction was not significantly different between 2 groups. Follow-up echocardiographic data suggested no difference in peak and mean gradients or incidence of aortic regurgitation between 2 approaches. CONCLUSIONS: This exploratory long-term analysis demonstrated that, in patients with severe aortic stenosis undergoing isolated AVR, there was no significant difference between manubrium-limited mini-sternotomy and conventional sternotomy with respect to all-cause mortality, rate of reoperation, MACE events and echocardiographic data at the median of 6.1-year follow-up.

Managing Unusual Sensory Experiences in People with First-Episode Psychosis (MUSE FEP): a study protocol for a single-blind parallel-group randomised controlled feasibility trial.

INTRODUCTION: Hallucinations (hearing or seeing things that others do not) are a common feature of psychosis, causing significant distress and disability. Existing treatments such as cognitive-behavioural therapy for psychosis (CBTp) have modest benefits, and there is a lack of CBTp-trained staff. Shorter, targeted treatments that focus on specific symptoms delivered by a non-specialist workforce could substantially increase access to treatment.Managing Unusual Sensory Experiences (MUSE) explains why people have hallucinations and helps the person to develop and use coping strategies to reduce distress. MUSE focuses only on hallucinations, and treatment is short (four to six, 1-hour sessions per week). It is a digital intervention, run on National Health Service (NHS) laptops, which provides information about hallucinations in an engaging way, using audio, video and animated content. Crucially, it is designed for use by non-specialist staff like community psychiatric nurses. METHODS AND ANALYSIS: The study is a two-arm feasibility randomised controlled trial comparing MUSE and treatment as usual (TAU) (n=40) to TAU alone (n=40), recruiting across two NHS Trusts, using 1:1 allocation and blind assessments before and after treatment (2 months) and at follow-up (3 months). Quantitative information on recruitment rates, adherence and completion of outcome assessments will be collected. Qualitative interviews will capture service users' experience of therapy and clinicians' experiences of the training and supervision in MUSE. Clinicians will also be asked about factors affecting uptake, adherence and facilitators/barriers to implementation. Analyses will focus on feasibility outcomes and provide initial estimates of intervention effects. Thematic analysis of the qualitative interviews will assess the acceptability of the training, intervention and trial procedures. ETHICS AND DISSEMINATION: The trial has received NHS Ethical and Health Research Authority approval. Findings will be disseminated directly to participants and services, as well as through peer-reviewed publications and conference presentations. TRIAL REGISTRATION NUMBER: ISRCTN16793301.

Preventing cardiotoxicity in patients with breast cancer and lymphoma: protocol for a multicentre randomised controlled trial (PROACT).

INTRODUCTION: Anthracyclines are included in chemotherapy regimens to treat several different types of cancer and are extremely effective. However, it is recognised that a significant side effect is cardiotoxicity; anthracyclines can cause irreversible damage to cardiac cells and ultimately impaired cardiac function and heart failure, which may only be evident years after exposure. The PROACT trial will establish the effectiveness of the ACE inhibitor enalapril maleate (enalapril) in preventing cardiotoxicity in patients with breast cancer and non-Hodgkin's lymphoma (NHL) receiving anthracycline-based chemotherapy. METHODS AND ANALYSIS: PROACT is a prospective, randomised, open-label, blinded end-point, superiority trial which will recruit adult patients being treated for breast cancer and NHL at NHS hospitals throughout England. The trial aims to recruit 106 participants, who will be randomised to standard care (high-dose anthracycline-based chemotherapy) plus enalapril (intervention) or standard care alone (control). Patients randomised to the intervention arm will receive enalapril (starting at 2.5 mg two times per day and titrating up to a maximum dose of 10 mg two times per day), commencing treatment at least 2 days prior to starting chemotherapy and finishing 3 weeks after their last anthracycline dose. The primary outcome is the presence or absence of cardiac troponin T release at any time during anthracycline treatment, and 1 month after the last dose of anthracycline. Secondary outcomes will focus on cardiac function measured using echocardiogram assessment, adherence to enalapril and side effects. ETHICS AND DISSEMINATION: A favourable opinion was given following research ethics committee review by West Midlands-Edgbaston REC, Ref: 17/WM/0248. Trial findings will be disseminated through engagement with patients, the oncology and cardiology communities, NHS management and commissioning groups and through peer-reviewed publication. TRIAL REGISTRATION NUMBER: NCT03265574.

A robust imputation method for missing responses and covariates in sample selection models.

Sample selection arises when the outcome of interest is partially observed in a study. Although sophisticated statistical methods in the parametric and non-parametric framework have been proposed to solve this problem, it is yet unclear how to deal with selectively missing covariate data using simple multiple imputation techniques, especially in the absence of exclusion restrictions and deviation from normality. Motivated by the 2003-2004 NHANES data, where previous authors have studied the effect of socio-economic status on blood pressure with missing data on income variable, we proposed the use of a robust imputation technique based on the selection-t sample selection model. The imputation method, which is developed within the frequentist framework, is compared with competing alternatives in a simulation study. The results indicate that the robust alternative is not susceptible to the absence of exclusion restrictions - a property inherited from the parent selection-t model - and performs better than models based on the normal assumption even when the data is generated from the normal distribution. Applications to missing outcome and covariate data further corroborate the robustness properties of the proposed method. We implemented the proposed approach within the MICE environment in R Statistical Software.

Adequate sample size for developing prediction models is not simply related to events per variable.

OBJECTIVES: The choice of an adequate sample size for a Cox regression analysis is generally based on the rule of thumb derived from simulation studies of a minimum of 10 events per variable (EPV). One simulation study suggested scenarios in which the 10 EPV rule can be relaxed. The effect of a range of binary predictors with varying prevalence, reflecting clinical practice, has not yet been fully investigated. STUDY DESIGN AND SETTING: We conducted an extended resampling study using a large general-practice data set, comprising over 2 million anonymized patient records, to examine the EPV requirements for prediction models with low-prevalence binary predictors developed using Cox regression. The performance of the models was then evaluated using an independent external validation data set. We investigated both fully specified models and models derived using variable selection. RESULTS: Our results indicated that an EPV rule of thumb should be data driven and that EPV ≥ 20 ​ generally eliminates bias in regression coefficients when many low-prevalence predictors are included in a Cox model. CONCLUSION: Higher EPV is needed when low-prevalence predictors are present in a model to eliminate bias in regression coefficients and improve predictive accuracy.