Search

Objective To develop a validated clinical prognostic model to determine the risk of atrial fibrillation after cardiac surgery as part of the PARADISE project (NIHR131227). Methods Prospective cohort study with linked electronic health records from a cohort of 5.6 million people in the United Kingdom Clinical Practice Research Datalink from 1998 to 2016. For model development, we considered a priori candidate predictors including demographics, medical history, medications, and clinical biomarkers. We evaluated associations between covariates and the AF incidence at the end of follow-up using logistic regression with the least absolute shrinkage and selection operator. The model was validated internally with the bootstrap method; subsequent performance was examined by discrimination quantified with the c-statistic and calibration assessed by calibration plots. The study follows TRIPOD guidelines. Results Between 1998 and 2016, 33,464 patients received cardiac surgery among the 5,601,803 eligible individuals. The final model included 13-predictors at baseline: age, year of index surgery, elevated CHA2DS2-VASc score, congestive heart failure, hypertension, acute coronary syndromes, mitral valve disease, ventricular tachycardia, valve surgery, receiving two combined procedures (e.g., valve replacement + coronary artery bypass grafting), or three combined procedures in the index procedure, statin use, and ethnicity other than white or black (statins and ethnicity were protective). This model had an optimism-corrected C-statistic of 0.68 both for the derivation and validation cohort. Calibration was good. Conclusions We developed a model to identify a group of individuals at high risk of AF and adverse outcomes who could benefit from long-term arrhythmia monitoring, risk factor management, rhythm control and/or thromboprophylaxis. Graphical abstract

In prediction model research, external validation is needed to examine an existing model's performance using data independent to that for model development. Current external validation studies often suffer from small sample sizes and consequently imprecise predictive performance estimates. To address this, we propose how to determine the minimum sample size needed for a new external validation study of a prediction model for a binary outcome. Our calculations aim to precisely estimate calibration (Observed/Expected and calibration slope), discrimination (C-statistic), and clinical utility (net benefit). For each measure, we propose closed-form and iterative solutions for calculating the minimum sample size required. These require specifying: (i) target SEs (confidence interval widths) for each estimate of interest, (ii) the anticipated outcome event proportion in the validation population, (iii) the prediction model's anticipated (mis)calibration and variance of linear predictor values in the validation population, and (iv) potential risk thresholds for clinical decision-making. The calculations can also be used to inform whether the sample size of an existing (already collected) dataset is adequate for external validation. We illustrate our proposal for external validation of a prediction model for mechanical heart valve failure with an expected outcome event proportion of 0.018. Calculations suggest at least 9835 participants (177 events) are required to precisely estimate the calibration and discrimination measures, with this number driven by the calibration slope criterion, which we anticipate will often be the case. Also, 6443 participants (116 events) are required to precisely estimate net benefit at a risk threshold of 8%. Software code is provided.

Objective Assess the prognostic value of pre-operative haemoglobin concentration (Hb) for identifying patients who develop severe post-operative anaemia or require blood transfusion following primary total hip or knee, or unicompartmental knee arthroplasty (THA, TKA, UKA). Background Pre-operative group and save (G&S), and post-operative Hb measurement may be unnecessary for many patients undergoing hip and knee arthroplasty provided individuals at greatest risk of severe post-operative anaemia can be identified. Methods and Materials Patients undergoing THA, TKA, or UKA between 2011 and 2018 were included. Outcomes were post-operative Hb below 70 and 80 g/L, and peri-operative blood transfusion. Logistic regression assessed the association between pre-operative Hb and each outcome. Decision curve analysis compared strategies for selecting patients for G&S and post-operative Hb measurement. Results 10 015 THA, TKA and UKA procedures were performed in 8582 patients. The incidence of blood transfusion (4.5%) decreased during the study. Using procedure specific Hb thresholds to select patients for pre-operative G&S and post-operative Hb testing had a greater net benefit than selecting all patients, no patients, or patients with pre-operative anaemia. Conclusions Pre-operative G&S and post-operative Hb measurement may not be indicated for UKA or TKA when adopting restrictive transfusion thresholds, provided clinicians accept a 0.1% risk of patients developing severe undiagnosed post-operative anaemia (Hb

ObjectiveWhile many studies have consistently found incomplete reporting of regression-based prediction model studies, evidence is lacking for machine learning-based prediction model studies. We aim to systematically review the adherence of Machine Learning (ML)-based prediction model studies to the Transparent Reporting of a multivariable prediction model for Individual Prognosis Or Diagnosis (TRIPOD) Statement.Study design and settingWe included articles reporting on development or external validation of a multivariable prediction model (either diagnostic or prognostic) developed using supervised ML for individualized predictions across all medical fields (PROSPERO, CRD42019161764). We searched PubMed from 1 January 2018 to 31 December 2019. Data extraction was performed using the 22-item checklist for reporting of prediction model studies (www.TRIPOD-statement.org). We measured the overall adherence per article and per TRIPOD item.ResultsOur search identified 24 814 articles, of which 152 articles were included: 94 (61.8%) prognostic and 58 (38.2%) diagnostic prediction model studies. Overall, articles adhered to a median of 38.7% (IQR 31.0-46.4) of TRIPOD items. No articles fully adhered to complete reporting of the abstract and very few reported the flow of participants (3.9%, 95% CI 1.8 to 8.3), appropriate title (4.6%, 95% CI 2.2 to 9.2), blinding of predictors (4.6%, 95% CI 2.2 to 9.2), model specification (5.2%, 95% CI 2.4 to 10.8), and model’s predictive performance (5.9%, 95% CI 3.1 to 10.9). There was often complete reporting of source of data (98.0%, 95% CI 94.4 to 99.3) and interpretation of the results (94.7%, 95% CI 90.0 to 97.3).ConclusionSimilar to prediction model studies developed using conventional regression-based techniques, the completeness of reporting is poor. Essential information to decide to use the model (i.e. model specification and its performance) is rarely reported. However, some items and sub-items of TRIPOD might be less suitable for ML-based prediction model studies and thus, TRIPOD requires extensions. Overall, there is an urgent need to improve the reporting quality and usability of research to avoid research waste.What is new?Key findings: Similar to prediction model studies developed using regression techniques, machine learning (ML)-based prediction model studies adhered poorly to the TRIPOD statement, the current standard reporting guideline.What this adds to what is known? In addition to efforts to improve the completeness of reporting in ML-based prediction model studies, an extension of TRIPOD for these type of studies is needed.What is the implication, what should change now? While TRIPOD-AI is under development, we urge authors to follow the recommendations of the TRIPOD statement to improve the completeness of reporting and reduce potential research waste of ML-based prediction model studies.

Contains fulltext : 251202.pdf (Publisher’s version ) (Open Access) A growing number of artificial intelligence (AI)-based clinical decision support systems are showing promising performance in preclinical, in silico, evaluation, but few have yet demonstrated real benefit to patient care. Early stage clinical evaluation is important to assess an AI system’s actual clinical performance at small scale, ensure its safety, evaluate the human factors surrounding its use, and pave the way to further large scale trials. However, the reporting of these early studies remains inadequate. The present statement provides a multistakeholder, consensus-based reporting guideline for the Developmental and Exploratory Clinical Investigations of DEcision support systems driven by Artificial Intelligence (DECIDE-AI). We conducted a two round, modified Delphi process to collect and analyse expert opinion on the reporting of early clinical evaluation of AI systems. Experts were recruited from 20 predefined stakeholder categories. The final composition and wording of the guideline was determined at a virtual consensus meeting. The checklist and the Explanation & Elaboration (E&E) sections were refined based on feedback from a qualitative evaluation process. 123 experts participated in the first round of Delphi, 138 in the second, 16 in the consensus meeting, and 16 in the qualitative evaluation. The DECIDE-AI reporting guideline comprises 17 AI specific reporting items (made of 28 subitems) and 10 generic reporting items, with an E&E paragraph provided for each. Through consultation and consensus with a range of stakeholders, we have developed a guideline comprising key items that should be reported in early stage clinical studies of AI-based decision support systems in healthcare. By providing an actionable checklist of minimal reporting items, the DECIDE-AI guideline will facilitate the appraisal of these studies and replicability of their findings.

Background: Current osteoporosis guidelines do not identify individuals with intellectual disabilities (ID) as at risk of fracture, potentially missing opportunities for prevention. We aimed to assess the incidence of fractures in people with ID over the life course. Methods: Descriptive analysis of open cohort study using anonymised electronic health records from the UK Clinical Practice Research Datalink, linked to the Hospital Episode Statistics database (Jan 1, 1998–Dec 31, 2017). All individuals with ID were matched on age and sex to five individuals without ID. We calculated the incidence rate (95% CI) per 10000 person-years (py) and incidence rate ratio (IRR, 95% CI) to compare fractures between individuals with and without ID (age 1–17 and ≥18 years) for any fracture, and in those aged 18–49 and ≥ 50 years for major osteoporotic fracture (vertebra, shoulder, wrist, hip), and for hip fracture. Findings: 43176 individuals with ID (15470 children aged 1–17 years; 27706 adults aged ≥ 18 years) were identified and included (40.4% females) along with 215733 matched control individuals. The median age at study entry was 24 (10th–90th centiles 3–54) years. Over a median (10th–90th centile) follow-up of 7.1 (0.9–17.6) and 6.5 (0.8–17.6) years, there were 5941 and 24363 incident fractures in the ID and non ID groups respectively. Incidence of any fracture was 143.5 (131.8–156.3) vs 120.7 (115.4–126.4)/10000 py (children), 174.2 (166.4–182.4)/10000 py vs 118.2 (115.3–121.2)/10000 py (adults) in females. In males it was 192.5 (182.4–203.2) vs 228.5 (223.0–234.1)/10000 py (children), 155.6 (149.3–162.1)/10000 py vs 128.4 (125.9–131.0)/10000 py (adults). IRR for major osteoporotic fracture was 1.81 (1.50–2.18) age 18–49 years, 1.69 (1.53–1.87) age ≥ 50 years in women. In men it was 1.56 (1.36–1.79) age 18–49 years, 2.45 (2.13–2.81) age ≥ 50 years. IRR for hip fracture was 7.79 (4.14–14.65) age 18–49 years, 2.28 (1.91–2.71) age ≥ 50 years in women. In men it was 6.04 (4.18–8.73) age 18–49 years, 3.91 (3.17–4.82) age ≥ 50 years. Comparable rates of major osteoporotic fracture and of hip fracture occurred approximately 15 and 20 years earlier respectively in women and 20 and 30 years earlier respectively in men with ID than without ID. Fracture distribution differed profoundly, hip fracture 9.9% vs 5.0% of any fracture in adults with ID vs without ID. Interpretation: The incidence, type, and distribution of fractures in people with intellectual disabilities suggest early onset osteoporosis. Prevention and management strategies are urgently required, particularly to reduce the incidence of hip fracture. Funding: National Institute for Health and Care Research.

In 2019 we published a pair of articles inStatistics in Medicinethat describe how to calculate the minimum sample size for developing a multivariable prediction model with a continuous outcome, or with a binary or time‐to‐event outcome. As for any sample size calculation, the approach requires the user to specify anticipated values for key parameters. In particular, for a prediction model with a binary outcome, the outcome proportion and a conservative estimate for the overall fit of the developed model as measured by the Cox‐SnellR2(proportion of variance explained) must be specified. This proposal raises the question of how to identify a plausible value forR2in advance of model development. Our articles suggest researchers should identifyR2from closely related models already published in their field. In this letter, we present details on how to deriveR2using the reportedCstatistic (AUROC) for such existing prediction models with a binary outcome. TheCstatistic is commonly reported, and so our approach allows researchers to obtainR2for subsequent sample size calculations for new models. Stata and R code is provided, and a small simulation study.

Objective: To systematically examine the design, reporting standards, risk of bias, and claims of studies comparing the performance of diagnostic deep learning algorithms for medical imaging with that of expert clinicians. Design: Systematic review. Data sources: Medline, Embase, Cochrane Central Register of Controlled Trials, and the World Health Organization trial registry from 2010 to June 2019. Eligibility criteria for selecting studies: Randomised trial registrations and non-randomised studies comparing the performance of a deep learning algorithm in medical imaging with a contemporary group of one or more expert clinicians. Medical imaging has seen a growing interest in deep learning research. The main distinguishing feature of convolutional neural networks (CNNs) in deep learning is that when CNNs are fed with raw data, they develop their own representations needed for pattern recognition. The algorithm learns for itself the features of an image that are important for classification rather than being told by humans which features to use. The selected studies aimed to use medical imaging for predicting absolute risk of existing disease or classification into diagnostic groups (eg, disease or non-disease). For example, raw chest radiographs tagged with a label such as pneumothorax or no pneumothorax and the CNN learning which pixel patterns suggest pneumothorax. Review methods: Adherence to reporting standards was assessed by using CONSORT (consolidated standards of reporting trials) for randomised studies and TRIPOD (transparent reporting of a multivariable prediction model for individual prognosis or diagnosis) for non-randomised studies. Risk of bias was assessed by using the Cochrane risk of bias tool for randomised studies and PROBAST (prediction model risk of bias assessment tool) for non-randomised studies. Results: Only 10 records were found for deep learning randomised clinical trials, two of which have been published (with low risk of bias, except for lack of blinding, and high adherence to reporting standards) and eight are ongoing. Of 81 non-randomised clinical trials identified, only nine were prospective and just six were tested in a real world clinical setting. The median number of experts in the comparator group was only four (interquartile range 2-9). Full access to all datasets and code was severely limited (unavailable in 95% and 93% of studies, respectively). The overall risk of bias was high in 58 of 81 studies and adherence to reporting standards was suboptimal ( Conclusions: Few prospective deep learning studies and randomised trials exist in medical imaging. Most non-randomised trials are not prospective, are at high risk of bias, and deviate from existing reporting standards. Data and code availability are lacking in most studies, and human comparator groups are often small. Future studies should diminish risk of bias, enhance real world clinical relevance, improve reporting and transparency, and appropriately temper conclusions.

Objective: To investigate whether placebo controls should be used in the evaluation of surgical interventions. Design: Systematic review. Data sources: We searched Medline, Embase, and the Cochrane Controlled Trials Register from their inception to November 2013. Study selection: Randomised clinical trials comparing any surgical intervention with placebo. Surgery was defined as any procedure that both changes the anatomy and requires a skin incision or use of endoscopic techniques. Data extraction: Three reviewers (KW, BJFD, IR) independently identified the relevant trials and extracted data on study details, outcomes, and harms from included studies. Results: In 39 out of 53 (74%) trials there was improvement in the placebo arm and in 27 (51%) trials the effect of placebo did not differ from that of surgery. In 26 (49%) trials, surgery was superior to placebo but the magnitude of the effect of the surgical intervention over that of the placebo was generally small. Serious adverse events were reported in the placebo arm in 18 trials (34%) and in the surgical arm in 22 trials (41.5%); in four trials authors did not specify in which arm the events occurred. However, in many studies adverse events were unrelated to the intervention or associated with the severity of the condition. The existing placebo controlled trials investigated only less invasive procedures that did not involve laparotomy, thoracotomy, craniotomy, or extensive tissue dissection. Conclusions: Placebo controlled trial is a powerful, feasible way of showing the efficacy of surgical procedures. The risks of adverse effects associated with the placebo are small. In half of the studies, the results provide evidence against continued use of the investigated surgical procedures. Without well designed placebo controlled trials of surgery, ineffective treatment may continue unchallenged.

© 2019, Springer-Verlag GmbH Germany, part of Springer Nature. Accessory mineral geochronometers such as apatite, baddeleyite, monazite, xenotime and zircon are increasingly being recognized for their ability to preserve diagnostic microstructural evidence of hypervelocity-impact processes. To date, little is known about the response of titanite to shock metamorphism, even though it is a widespread accessory phase and a U–Pb geochronometer. Here we report two new mechanical twin modes in titanite within shocked granitoid from the Chicxulub impact structure, Mexico. Titanite grains in the newly acquired core from the International Ocean Discovery Program Hole M0077A preserve multiple sets of polysynthetic twins, most commonly with composition planes (K 1 ) = ~ { 1 ¯ 11 } , and shear direction (η 1 ) = < 110 > , and less commonly with the mode K 1 = {130}, η 1 = ~ . In some grains, {130} deformation bands have formed concurrently with the deformation twins, indicating dislocation slip with Burgers vector b = < 341 > can be active during impact metamorphism. Titanite twins in the modes described here have not been reported from endogenically deformed rocks; we, therefore, propose this newly identified twin form as a result of shock deformation. Formation conditions of the twins have not been experimentally calibrated, and are here empirically constrained by the presence of planar deformation features in quartz (12 ± 5 and ~ 17 ± 5 GPa) and the absence of shock twins in zircon (< 20 GPa). While the lower threshold of titanite twin formation remains poorly constrained, identification of these twins highlight the utility of titanite as a shock indicator over the pressure range between 12 and 17 GPa. Given the challenges to find diagnostic indicators of shock metamorphism to identify both ancient and recent impact evidence on Earth, microstructural analysis of titanite is here demonstrated to provide a new tool for recognizing impact deformation in rocks where other impact evidence may be erased, altered, or did not manifest due to generally low (< 20 GPa) shock pressure.

Aims - To compare the ability of the National Early Warning Score (NEWS) and the National Early Warning Score 2 (NEWS2) to identify patients at risk of in-hospital mortality and other adverse outcomes.Methods - We undertook a multi-centre retrospective observational study at five acute hospitals from two UK NHS Trusts. Data were obtained from completed adult admissions who were not fit enough to be discharged alive on the day of admission. Diagnostic coding and oxygen prescriptions were used to identify patients with type II respiratory failure (T2RF). The primary outcome was in-hospital mortality within 24 hours of a vital signs observation. Secondary outcomes included unanticipated intensive care unit admission or cardiac arrest within 24 hours of a vital signs observation. Discrimination was assessed using the c-statistic.Results - Among 251,266 adult admissions, 48,898 were identified to be at risk of T2RF by diagnostic coding. In this group, NEWS2 showed statistically significant lower discrimination (c-statistic, 95% CI) for identifying in-hospital mortality within 24 hours (0.860, 0.857-0.864) than NEWS (0.881, 0.878-0.884). For 1,394 admissions with documented T2RF, discrimination was similar for both systems: NEWS2 (0.841, 0.827-0.855), NEWS (0.862, 0.848-0.875). For all secondary endpoints, NEWS2 showed no improvements in discrimination.Conclusions - NEWS2 modifications to NEWS do not improve discrimination of adverse outcomes in patients with documented T2RF and decrease discrimination in patients at risk of T2RF. Further evaluation of the relationship between SpO2 values, oxygen therapy and risk should be investigated further before wide-scale adoption of NEWS2.

On average, secondary impact craters are expected to deepen and become more symmetric as impact velocity (vi) increases with downrange distance (L). We have used high-resolution topography (1–2 m/pixel) to characterize the morphometry of secondary craters as a function of L for several well-preserved primary craters on Mars. The secondaries in this study (N = 2644) span a range of diameters (25 m ≤D≤400 m) and estimated impact velocities (0.4 km/s ≤vi≤2 km/s). The range of diameter-normalized rim-to-floor depth (d/D) broadens and reaches a ceiling of d/D≈0.22 at L≈280 km (vi= 1–1.2 km/s), whereas average rim height shows little dependence on vi for the largest craters (h/D≈0.02, D > 60 m). Populations of secondaries that express the following morphometric asymmetries are confined to regions of differing radial extent: planform elongations (L< 110–160 km), taller downrange rims (L < 280 km), and cavities that are deeper uprange (L< 450–500 km). Populations of secondaries with lopsided ejecta were found to extend to at least L ∼ 700 km. Impact hydrocode simulations with iSALE-2D for strong, intact projectile and target materials predict a ceiling for d/D versus L whose trend is consistent with our measurements. This study illuminates the morphometric transition from subsonic to hypervelocity cratering and describes the initial state of secondary crater populations. This has applications to understanding the chronology of planetary surfaces and the long-term evolution of small crater populations.

We conducted a paleomagnetic study of the matrix of Allende CV3 chondritic meteorite, isolating the matrix’s primary remanent magnetization, measuring its magnetic fabric and estimating the ancient magnetic field intensity. A strong planar magnetic fabric was identified; the remanent magnetization of the matrix was aligned within this plane, suggesting a mechanism relating the magnetic fabric and remanence. The intensity of the matrix’s remanent magnetization was found to be consistent and low (~6 μT). The primary magnetic mineral was found to be pyrrhotite. Given the thermal history of Allende, we conclude that the remanent magnetization formed during or after an impact event. Recent mesoscale impact mode ling, where chondrules and matrix are resolved, has shown that low-velocity collisions can generate significant matrix temperatures, as pore-space compaction attenuates shock energy and dramatically increases the amount of heating. Non-porous chondrules are unaffected, and act as heat-sinks, so matrix temperature excursions are brief. We extend this work to model Allende, and show that a 1km/s planar impact generates bulk porosity, matrix porosity, and fabric in our target that match the observed values. Bimodal mixtures of a highly porous matrix and nominally zero-porosity chondrules, make chondrites uniquely capable of recording transient or unstable fields. Targets that have uniform porosity, e.g., terrestrial impact craters, will not record transient or unstable fields. Rather than a core dynamo, it is therefore possible that the origin of the magnetic field in Allende was the impact itself, or a nebula field recorded during transient impact heating.

[This corrects the article DOI: 10.1186/s41512-016-0001-y.].

Asteroids and comets 10–100 m in size that collide with Earth disrupt dramatically in the atmosphere with an explosive transfer of energy, caused by extreme air drag. Such airbursts produce a strong blastwave that radiates from the meteoroid's trajectory and can cause damage on the surface. An established technique for predicting airburst blastwave damage is to treat the airburst as a static source of energy and to extrapolate empirical results of nuclear explosion tests using an energy-based scaling approach. Here we compare this approach to two more complex models using the iSALE shock physics code. We consider a moving-source airburst model where the meteoroid's energy is partitioned as two-thirds internal energy and one-third kinetic energy at the burst altitude, and a model in which energy is deposited into the atmosphere along the meteoroid's trajectory based on the pancake model of meteoroid disruption. To justify use of the pancake model, we show that it provides a good fit to the inferred energy release of the 2013 Chelyabinsk fireball. Predicted overpressures from all three models are broadly consistent at radial distances from ground zero that exceed three times the burst height. At smaller radial distances, the moving-source model predicts overpressures two times greater than the static-source model, whereas the cylindrical line-source model based on the pancake model predicts overpressures two times lower than the static-source model. Given other uncertainties associated with airblast damage predictions, the static-source approach provides an adequate approximation of the azimuthally averaged airblast for probabilistic hazard assessment.

Large impact structures have complex morphologies, with zones of structural uplift that can be expressed topographically as central peaks and/or peak rings internal to the crater rim. The formation of these structures requires transient strength reduction in the target material and one of the proposed mechanisms to explain this behavior is acoustic fluidization. Here, samples of shock-metamorphosed quartz-bearing lithologies at the West Clearwater Lake impact structure, Canada, are used to estimate the maximum recorded shock pressures in three dimensions across the crater. These measurements demonstrate that the currently-observed distribution of shock metamorphism is strongly controlled by the formation of the structural uplift. The distribution of peak shock pressures, together with apparent crater morphology and geological observations, is compared with numerical impact simulations to constrain parameters used in the block-model implementation of acoustic fluidization. The numerical simulations produce craters that are consistent with morphological and geological observations. The results show that the regeneration of acoustic energy must be an important feature of acoustic fluidization in crater collapse, and should be included in future implementations. Based on the comparison between observational data and impact simulations we conclude that the West Clearwater Lake structure had an original rim (final crater) diameter of 35–40 km and has since experienced up to ~2 km of differential erosion.

Impact bombardment during the first billion years after the formation of the Moon produced at least several tens of basins. The Gravity Recovery and Interior Laboratory (GRAIL) mission mapped the gravity field of these impact structures at significantly higher spatial resolution than previous missions, allowing for detailed subsurface and morphological analyses to be made across the entire globe. GRAIL-derived crustal thickness maps were used to define the regions of crustal thinning observed in centers of lunar impact basins, which represents a less unambiguous measure of a basin size than those based on topographic features. The formation of lunar impact basins was modeled numerically by using the iSALE-2D hydrocode, with a large range of impact and target conditions typical for the first billion years of lunar evolution. In the investigated range of impactor and target conditions, the target temperature had the dominant effect on the basin subsurface morphology. Model results were also used to update current impact scaling relationships applicable to the lunar setting (based on assumed target temperature). Our new temperature-dependent impact-scaling relationships provide estimates of impact conditions and transient crater diameters for the majority of impact basins mapped by GRAIL. As the formation of lunar impact basins is associated with the first ~700 Myr of the solar system evolution when the impact flux was considerably larger than the present day, our revised impact scaling relationships can aid further analyses and understanding of the extent of impact bombardment on the Moon and terrestrial planets in the early solar system.

Early identification of ovarian cancer is an unresolved challenge and the predictive value of single symptoms is limited. We evaluated the performance of QCancer(®) (Ovarian) prediction model for predicting the risk of ovarian cancer in a UK cohort of general practice patients. A total of 1.1 million patients registered with a general practice surgery between 1 January 2000 and 30 June 2008, aged 30-84 years with 735 ovarian cancer cases, were included in the analysis. Ovarian cancer was defined as incident diagnosis of ovarian cancer during the 2 years after study entry. The results from this independent and external validation of QCancer(®) (Ovarian) prediction model demonstrated good performance on a large cohort of general practice patients. QCancer(®) (Ovarian) had very good discrimination with an area under the receiver operating characteristic curve of 0.86 and explained 59.9% of the variation. QCancer(®) (Ovarian) was well calibrated across all tenths of risk and over all age. The 10% of women with the highest predicted risks included 64% of all ovarian cancer diagnoses over the next 2 years. QCancer(®) (Ovarian) appears to be a useful tool for identifying undetected cases of ovarian cancer in primary care in the UK for early referral and investigation.

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. © 2015 The Authors. Statistics in Medicine Published by John Wiley and Sons Ltd.