International evaluation of an AI system for breast cancer screening.

Screening mammography aims to identify breast cancer at earlier stages of the disease, when treatment can be more successful1. Despite the existence of screening programmes worldwide, the interpretation of mammograms is affected by high rates of false positives and false negatives2. Here we present an artificial intelligence (AI) system that is capable of surpassing human experts in breast cancer prediction. To assess its performance in the clinical setting, we curated a large representative dataset from the UK and a large enriched dataset from the USA. We show an absolute reduction of 5.7% and 1.2% (USA and UK) in false positives and 9.4% and 2.7% in false negatives. We provide evidence of the ability of the system to generalize from the UK to the USA. In an independent study of six radiologists, the AI system outperformed all of the human readers: the area under the receiver operating characteristic curve (AUC-ROC) for the AI system was greater than the AUC-ROC for the average radiologist by an absolute margin of 11.5%. We ran a simulation in which the AI system participated in the double-reading process that is used in the UK, and found that the AI system maintained non-inferior performance and reduced the workload of the second reader by 88%. This robust assessment of the AI system paves the way for clinical trials to improve the accuracy and efficiency of breast cancer screening.

A clinically applicable approach to continuous prediction of future acute kidney injury.

The early prediction of deterioration could have an important role in supporting healthcare professionals, as an estimated 11% of deaths in hospital follow a failure to promptly recognize and treat deteriorating patients1. To achieve this goal requires predictions of patient risk that are continuously updated and accurate, and delivered at an individual level with sufficient context and enough time to act. Here we develop a deep learning approach for the continuous risk prediction of future deterioration in patients, building on recent work that models adverse events from electronic health records2-17 and using acute kidney injury-a common and potentially life-threatening condition18-as an exemplar. Our model was developed on a large, longitudinal dataset of electronic health records that cover diverse clinical environments, comprising 703,782 adult patients across 172 inpatient and 1,062 outpatient sites. Our model predicts 55.8% of all inpatient episodes of acute kidney injury, and 90.2% of all acute kidney injuries that required subsequent administration of dialysis, with a lead time of up to 48 h and a ratio of 2 false alerts for every true alert. In addition to predicting future acute kidney injury, our model provides confidence assessments and a list of the clinical features that are most salient to each prediction, alongside predicted future trajectories for clinically relevant blood tests9. Although the recognition and prompt treatment of acute kidney injury is known to be challenging, our approach may offer opportunities for identifying patients at risk within a time window that enables early treatment.

A root cause analysis system to establish the most plausible explanation for post-endoscopy upper gastrointestinal cancer.

BACKGROUND : Missing upper gastrointestinal cancer (UGIC) at endoscopy may prevent curative treatment. We have developed a root cause analysis system for potentially missed UGICs at endoscopy (post-endoscopy UGIC [PEUGIC]) to establish the most plausible explanations. METHODS : The electronic records of patients with UGIC at two National Health Service providers were examined. PEUGICs were defined as UGICs diagnosed 6-36 months after an endoscopy that did not diagnose cancer. An algorithm based on the World Endoscopy Organization post-colonoscopy colorectal cancer algorithm was developed to categorize and identify potentially avoidable PEUGICs. RESULTS : Of 1327 UGICs studied, 89 (6.7 %) were PEUGICs (patient median [IQR] age at endoscopy 73.5 (63.5-81.0); 60.7 % men). Of the PEUGICs, 40 % were diagnosed in patients with Barrett's esophagus. PEUGICs were categorized as: A - lesion detected, adequate assessment and decision-making, but PEUGIC occurred (16.9 %); B - lesion detected, inadequate assessment or decision-making (34.8 %); C - possible missed lesion, endoscopy and decision-making adequate (8.9 %); D - possible missed lesion, endoscopy or decision-making inadequate (33.7 %); E - deviated from management pathway but appropriate (5.6 %); F - deviated inappropriately from management pathway (3.4 %). The majority of PEUGICs (71 %) were potentially avoidable and in 45 % the cancer outcome could have been different if it had been diagnosed on the initial endoscopy. There was a negative correlation between endoscopists' mean annual number of endoscopies and the technically attributable PEUGIC rate (correlation coefficient -0.57; P = 0.004). CONCLUSION : Missed opportunities to avoid PEUGIC were identified in 71 % of cases. Root cause analysis can standardize future investigation of PEUGIC and guide quality improvement efforts.

The risk of subsequent surgery following bowel resection for Crohn's disease in a national cohort of 19 207 patients.

AIM: Surgery is required for most patients with Crohn's disease (CD) and further surgery may be necessary if medical treatment fails to control disease activity. The aim of this study was to characterize the risk of, and factors associated with, further surgery following a first resection for Crohn's disease. METHODS: Hospital Episode Statistics from England were examined to identify patients with CD and a first recorded bowel resection between 2007 and 2016. Multivariable logistic regression was used to examine risk factors for further resectional surgery within 5 years. Prevalence-adjusted surgical rates for index CD surgery over the study period were calculated. RESULTS: In total, 19 207 patients (median age 39 years, interquartile range 27-53 years; 55% women) with CD underwent a first recorded resection during the study period. 3141 (16%) underwent a further operation during the study period. The median time to further surgery was 2.4 (interquartile range 1.2-4.6) years. 3% of CD patients had further surgery within 1 year, 14% by 5 years and 23% by 10 years. Older age (≥58), index laparoscopic surgery and index elective surgery (adjusted OR 0.65, 95% CI 0.54-0.77; 0.77, 0.67-0.88; and 0.77, 0.69-0.85; respectively) were associated with a reduced risk of further surgery by 5 years. Prior surgery for perianal disease (1.60, 1.37-1.87), an extraintestinal manifestation of CD (1.51, 1.22-1.86) and index surgery in a high-volume centre for CD surgery (1.20, 1.02-1.40) were associated with an increased risk of further surgery by 5 years. A 25% relative and 0.3% absolute reduction in prevalence-adjusted index surgery rates for CD was observed over the study period. CONCLUSIONS: Further surgery following an index operation is common in CD. This risk was particularly seen in patients with perianal disease, extraintestinal manifestations and those who underwent index surgery in a high-volume centre.

What is the Efficacy of Intra-Articular Platelet-Rich Plasma Injections for Knee Osteoarthritis in Clinical Practice? A "Real-Life" Prospective Cohort.

INTRODUCTION: Platelet-rich plasma (PRP) injections may improve symptoms in patients suffering from knee osteoarthritis. However, there is a lack of data on its effectiveness in a "real-life" cohort. This multi-site institutional registry study aimed to assess patients' longitudinal progress after PRP injection for knee osteoarthritis. MATERIALS AND METHODS: All patients receiving PRP injections for knee osteoarthritis at a large, integrated tertiary academic center (December 18, 2017 to March 1, 2021) were eligible. A prospective data collection instrument was used to collect patient demographics, procedural information, and patient-reported outcome measures. Overall, 97 patients met the inclusion criteria, and 53 were included in the analysis. RESULTS: One in four patients (26%) improved on all three Knee Injury and Osteoarthritis Outcome Score subscales: 17% in two subscales and 20% in one subscale, respectively. Overall, 64% of patients improved in at least one patient-reported outcomes measure. At six months post injection, 49% of patients were satisfied. CONCLUSION: PRP injection provides positive changes in two out of three patients in different magnitudes and characteristics with careful attention to clinically meaningful differences.

Rate of pancreatic cancer following a negative endoscopic ultrasound and associated factors.

BACKGROUND : Data are limited regarding pancreatic cancer diagnosed following a pancreaticobiliary endoscopic ultrasound (EUS) that does not diagnose pancreatic cancer. We have studied the frequency of, and factors associated with, post-EUS pancreatic cancer (PEPC) and 1-year mortality. METHODS : Between 2010 and 2017, patients with pancreatic cancer and a preceding pancreaticobiliary EUS were identified in a national cohort using Hospital Episode Statistics. Patients with a pancreaticobiliary EUS 6-18 months before a later pancreatic cancer diagnosis were the PEPC cases; controls were those with pancreatic cancer diagnosed within 6 months of pancreaticobiliary EUS. Multivariable logistic regression models examined the factors associated with PEPC and a Cox regression model examined factors associated with 1-year cumulative mortality. RESULTS : 9363 pancreatic cancer patients were studied; 93.5 % identified as controls (men 53.2 %; median age 68 [interquartile range (IQR) 61-75]); 6.5 % as PEPC cases (men 58.2 %; median age 69 [IQR 61-77]). PEPC was associated with older age (≥ 75 years compared with < 65 years, odds ratio [OR] 1.42, 95 %CI 1.15-1.76), increasing co-morbidity (Charlson co-morbidity score > 5, OR 1.90, 95 %CI 1.49-2.43), chronic pancreatitis (OR 3.13, 95 %CI 2.50-3.92), and diabetes mellitus (OR 1.58, 95 %CI 1.31-1.90). Metal biliary stents (OR 0.57, 95 %CI 0.38-0.86) and EUS-FNA (OR 0.49, 95 %CI 0.41-0.58) were inversely associated with PEPC. PEPC was associated with a higher cumulative mortality at 1 year (hazard ratio 1.12, 95 %CI 1.02-1.24), with only 14 % of PEPC patients (95 %CI 12 %-17 %) having a surgical resection, compared with 21 % (95 %CI 20 %-22 %) of controls. CONCLUSIONS : PEPC occurred in 6.5 % of patients and was associated with chronic pancreatitis, older age, more co-morbidities, and specifically diabetes mellitus. PEPC was associated with a worse prognosis and lower surgical resection rates.

AOASM Position Statement on Esports, Active Video Gaming, and the Role of the Sports Medicine Physician.

ABSTRACT: Electronic sports, or esports, has a global audience of over 300 million fans and is increasing in popularity, resulting in projected revenue of over $1 billion by the end of this past year. The global pandemic of 2020 had little to no effect on these increasing numbers because athletes have been able to continue to engage in sports because of its electronic nature and fans have been able to follow them virtually. Esports has been recognized as an organized sport by the International Olympic Committee, the US National Collegiate Athletic Association, and several secondary school athletic associations within the United States. In addition, professional teams have been established in several major cities within the United States, Canada, Europe, and Australia. With the growth of esports, the necessity of incorporating esports medicine into the practice of sports medicine physicians has become paramount. Esports can be played on a monitor or screen and played using physical activity in what has become known as active video gaming. Within both of these platforms, there have emerged certain conditions unique to esports. There are also certain conditions seen in other sports applicable to esports athletes. This document will review the evaluation of the esports athlete, introduce conditions unique to these athletes and review common conditions seen in esports, discuss diagnostics used in the evaluation of esports athletes, introduce treatment options for conditions unique to esports and review those for commonly seen injuries in esports, discuss prevention of injuries in esports, and introduce a framework for the future development of esports medicine that can be introduced into the daily practice of the sports medicine physician.

Achalasia Is Associated With Atopy in Patients Younger Than 40 Years of Age.

INTRODUCTION: Case reports describe individuals with achalasia features subsequently diagnosed with eosinophilic esophagitis (an atopic disorder). We have examined associations between achalasia and atopic and autoimmune conditions. METHODS: This is a UK cohort study of 2,593 subjects with achalasia matched to 10,402 controls. RESULTS: At diagnosis, achalasia was associated with autoimmune conditions (odds ratio 1.39; 95% confidence interval 1.02-1.90) and atopic conditions (1.40; 1.00-1.95) in those aged younger than 40 years. DISSCUSSION: Our findings support an autoimmune etiology in achalasia but also suggest a possible atopic etiology in younger subjects.

A national study of artificial urinary sphincter and male sling implantation after radical prostatectomy in England.

OBJECTIVES: To consider the provision of post-radical prostatectomy (RP) continence surgery in England. MATERIALS AND METHODS: Patients with an Office of Population Census and Surveys Classification of Interventions and Procedures, version 4 code for an artificial urinary sphincter (AUS) or male sling between 1 January 2010 and 31 March 2018 were searched for within the Hospital Episode Statistics (HES) dataset. Those without previous RP were excluded. Multivariable logistic regressions for repeat AUS and sling procedures were built in stata. Further descriptive analysis of provision of procedures was performed. RESULTS: A total of 1414 patients had received index AUS, 10.3% of whom had undergone prior radiotherapy; their median follow-up was 3.55 years. The sling cohort contained 816 patients; 6.7% of these had received prior radiotherapy and the median follow-up was 3.23 years. Whilst the number of AUS devices implanted had increased each year, male slings peaked in 2014/2015. AUS redo/removal was performed in 11.2% of patients. Patients in low-volume centres were more likely to require redo/removal (odds ratio [OR] 2.23 95% confidence interval [CI] 1.02-4.86; P = 0.045). A total of 12.0% patients with a sling progressed to AUS implantation and 1.3% had a second sling. Patients with previous radiotherapy were more likely to require a second operation (OR 2.03 95% CI 1.01-4.06; P = 0.046). Emergency re-admissions within 30 days of index operation were 3.9% and 3.6% fewer in high-volume centres, for AUS and slings respectively. The median time to initial continence surgery from RP was 2.8 years. Increased time from RP conferred no reduced risk of redo surgery for either procedure. CONCLUSION: There is a volume effect for outcomes of AUS procedures, suggesting that they should only be performed in high-volume centres. Given the known impact of incontinence on quality of life, patients should be referred sooner for post-prostatectomy continence surgery.

Behavioral economics informed message content in text message reminders to improve cervical screening participation: Two pragmatic randomized controlled trials.

The objective of the reported research was to assess the impact of text message (SMS) reminders and their content on cervical screening rates. Women invited for cervical screening in Northwest London from February-October 2015 were eligible. 3133 women aged 24-29 (Study 1) were randomized (1, 1) to 'no SMS' (control), or a primary care physician (PCP) endorsed SMS (SMS-PCP). 11,405 women aged 30-64 (Study 2), were randomized (1, 1:1:1:1:1:1) to either: no SMS, an SMS without manipulation (SMS), the SMS-PCP, an SMS with a total or proportionate social norm (SMS-SNT or SMS-SNP), or an SMS with a gain-framed or loss-framed message (SMS-GF and SMS-LF). The primary outcome was participation at 18 weeks. In Study 1 participation was significantly higher in the SMS-PCP arm (31.4%) compared to control (26.4%, aOR, 1.29, 95%CI: 1.09-1·51; p = 0.002). In Study 2 participation was highest in the SMS-PCP (38.4%) and SMS (38.1%) arms compared to control (34.4%), (aOR: 1.19, 95%CI: 1.03-1.38; p = 0.02 and aOR: 1.18, 95%CI: 1.02-1.37; p = 0.03, respectively). The results demonstrate that behavioral SMSs improve cervical screening participation. The message content plays an important role in the impact of SMS. The results from this trial have already been used to designing effective policy for cervical cancer screening. The NHS Cervical Screening Programme started running a London-wide screening SMS campaign which was based on the cervical screening trial described here. According to figures published by Public Health England, after six months attendance increased by 4.8%, which is the equivalent of 13,400 more women being screened at 18 weeks.

Gamer's Health Guide: Optimizing Performance, Recognizing Hazards, and Promoting Wellness in Esports.

Electronic sports (esports), or competitive video gaming, is a rapidly growing industry and phenomenon. While around 90% of American children play video games recreationally, the average professional esports athlete spends 5.5 to 10 h gaming daily. These times and efforts parallel those of traditional sports activities where individuals can participate at the casual to the professional level with the respective time commitments. Given the rapid growth in esports, greater emphasis has been placed on identification, management, and prevention of common health hazards that are associated with esports participation while also focusing on the importance of health promotion for this group of athletes. This review outlines a three-point framework for sports medicine providers, trainers, and coaches to provide a holistic approach for the care of the esports athlete. This esports framework includes awareness and management of common musculoskeletal and health hazards, opportunities for health promotion, and recommendations for performance optimization.

Key challenges for delivering clinical impact with artificial intelligence.

BACKGROUND: Artificial intelligence (AI) research in healthcare is accelerating rapidly, with potential applications being demonstrated across various domains of medicine. However, there are currently limited examples of such techniques being successfully deployed into clinical practice. This article explores the main challenges and limitations of AI in healthcare, and considers the steps required to translate these potentially transformative technologies from research to clinical practice. MAIN BODY: Key challenges for the translation of AI systems in healthcare include those intrinsic to the science of machine learning, logistical difficulties in implementation, and consideration of the barriers to adoption as well as of the necessary sociocultural or pathway changes. Robust peer-reviewed clinical evaluation as part of randomised controlled trials should be viewed as the gold standard for evidence generation, but conducting these in practice may not always be appropriate or feasible. Performance metrics should aim to capture real clinical applicability and be understandable to intended users. Regulation that balances the pace of innovation with the potential for harm, alongside thoughtful post-market surveillance, is required to ensure that patients are not exposed to dangerous interventions nor deprived of access to beneficial innovations. Mechanisms to enable direct comparisons of AI systems must be developed, including the use of independent, local and representative test sets. Developers of AI algorithms must be vigilant to potential dangers, including dataset shift, accidental fitting of confounders, unintended discriminatory bias, the challenges of generalisation to new populations, and the unintended negative consequences of new algorithms on health outcomes. CONCLUSION: The safe and timely translation of AI research into clinically validated and appropriately regulated systems that can benefit everyone is challenging. Robust clinical evaluation, using metrics that are intuitive to clinicians and ideally go beyond measures of technical accuracy to include quality of care and patient outcomes, is essential. Further work is required (1) to identify themes of algorithmic bias and unfairness while developing mitigations to address these, (2) to reduce brittleness and improve generalisability, and (3) to develop methods for improved interpretability of machine learning predictions. If these goals can be achieved, the benefits for patients are likely to be transformational.

Aortic conduit aneurysm and enteric fistula formation in a post-liver transplant patient: A potential causative role for cytomegalovirus?

Hepatic arterial aortic conduits can be used as an alternative means of revascularizing the donor liver when the native recipient hepatic artery (HA) cannot be used. Cytomegalovirus (CMV) is a common Herpesviridae infection in patients who have undergone solid organ transplants. It can be asymptomatic but may cause fever and invasive disease affecting any organ system. Here we describe the first case in the literature of an aortic conduit aneurysm and concurrent CMV viremia following liver transplantation. We speculate on a causative role for CMV in the development of the aneurysm.

Changing health behaviors using financial incentives: a review from behavioral economics.

BACKGROUND: Incentives are central to economics and are used across the public and private sectors to influence behavior. Recent interest has been shown in using financial incentives to promote desirable health behaviors and discourage unhealthy ones. MAIN TEXT: If we are going to use incentive schemes to influence health behaviors, then it is important that we give them the best chance of working. Behavioral economics integrates insights from psychology with the laws of economics and provides a number of robust psychological phenomena that help to better explain human behavior. Individuals' decisions in relation to incentives may be shaped by more subtle features - such as loss aversion, overweighting of small probabilities, hyperbolic discounting, increasing payoffs, reference points - many of which have been identified through research in behavioral economics. If incentives are shown to be a useful strategy to influence health behavior, a wider discussion will need to be had about the ethical dimensions of incentives before their wider implementation in different health programmes. CONCLUSIONS: Policy makers across the world are increasingly taking note of lessons from behavioral economics and this paper explores how key principles could help public health practitioners design effective interventions both in relation to incentive designs and more widely.

Implementation of a Digitally Enabled Care Pathway (Part 2): Qualitative Analysis of Experiences of Health Care Professionals.

BACKGROUND: One reason for the introduction of digital technologies into health care has been to try to improve safety and patient outcomes by providing real-time access to patient data and enhancing communication among health care professionals. However, the adoption of such technologies into clinical pathways has been less examined, and the impacts on users and the broader health system are poorly understood. We sought to address this by studying the impacts of introducing a digitally enabled care pathway for patients with acute kidney injury (AKI) at a tertiary referral hospital in the United Kingdom. A dedicated clinical response team-comprising existing nephrology and patient-at-risk and resuscitation teams-received AKI alerts in real time via Streams, a mobile app. Here, we present a qualitative evaluation of the experiences of users and other health care professionals whose work was affected by the implementation of the care pathway. OBJECTIVE: The aim of this study was to qualitatively evaluate the impact of mobile results viewing and automated alerting as part of a digitally enabled care pathway on the working practices of users and their interprofessional relationships. METHODS: A total of 19 semistructured interviews were conducted with members of the AKI response team and clinicians with whom they interacted across the hospital. Interviews were analyzed using inductive and deductive thematic analysis. RESULTS: The digitally enabled care pathway improved access to patient information and expedited early specialist care. Opportunities were identified for more constructive planning of end-of-life care due to the earlier detection and alerting of deterioration. However, the shift toward early detection also highlighted resource constraints and some clinical uncertainty about the value of intervening at this stage. The real-time availability of information altered communication flows within and between clinical teams and across professional groups. CONCLUSIONS: Digital technologies allow early detection of adverse events and of patients at risk of deterioration, with the potential to improve outcomes. They may also increase the efficiency of health care professionals' working practices. However, when planning and implementing digital information innovations in health care, the following factors should also be considered: the provision of clinical training to effectively manage early detection, resources to cope with additional workload, support to manage perceived information overload, and the optimization of algorithms to minimize unnecessary alerts.

Implementation of a Digitally Enabled Care Pathway (Part 1): Impact on Clinical Outcomes and Associated Health Care Costs.

BACKGROUND: The development of acute kidney injury (AKI) in hospitalized patients is associated with adverse outcomes and increased health care costs. Simple automated e-alerts indicating its presence do not appear to improve outcomes, perhaps because of a lack of explicitly defined integration with a clinical response. OBJECTIVE: We sought to test this hypothesis by evaluating the impact of a digitally enabled intervention on clinical outcomes and health care costs associated with AKI in hospitalized patients. METHODS: We developed a care pathway comprising automated AKI detection, mobile clinician notification, in-app triage, and a protocolized specialist clinical response. We evaluated its impact by comparing data from pre- and postimplementation phases (May 2016 to January 2017 and May to September 2017, respectively) at the intervention site and another site not receiving the intervention. Clinical outcomes were analyzed using segmented regression analysis. The primary outcome was recovery of renal function to ≤120% of baseline by hospital discharge. Secondary clinical outcomes were mortality within 30 days of alert, progression of AKI stage, transfer to renal/intensive care units, hospital re-admission within 30 days of discharge, dependence on renal replacement therapy 30 days after discharge, and hospital-wide cardiac arrest rate. Time taken for specialist review of AKI alerts was measured. Impact on health care costs as defined by Patient-Level Information and Costing System data was evaluated using difference-in-differences (DID) analysis. RESULTS: The median time to AKI alert review by a specialist was 14.0 min (interquartile range 1.0-60.0 min). There was no impact on the primary outcome (estimated odds ratio [OR] 1.00, 95% CI 0.58-1.71; P=.99). Although the hospital-wide cardiac arrest rate fell significantly at the intervention site (OR 0.55, 95% CI 0.38-0.76; P<.001), DID analysis with the comparator site was not significant (OR 1.13, 95% CI 0.63-1.99; P=.69). There was no impact on other secondary clinical outcomes. Mean health care costs per patient were reduced by £2123 (95% CI -£4024 to -£222; P=.03), not including costs of providing the technology. CONCLUSIONS: The digitally enabled clinical intervention to detect and treat AKI in hospitalized patients reduced health care costs and possibly reduced cardiac arrest rates. Its impact on other clinical outcomes and identification of the active components of the pathway requires clarification through evaluation across multiple sites.

Improving the Quality of Ward-based Surgical Care With a Human Factors Intervention Bundle.

OBJECTIVE: This study aimed to explore the impact of a human factors intervention bundle on the quality of ward-based surgical care in a UK hospital. SUMMARY OF BACKGROUND DATA: Improving the culture of a surgical team is a difficult task. Engagement with stakeholders before intervention is key. Studies have shown that appropriate supervision can enhance surgical ward safety. METHODS: A pre-post intervention study was conducted. The intervention bundle consisted of twice-daily attending ward rounds, a "chief resident of the week" available at all times on the ward, an escalation of care protocol and team contact cards. Twenty-seven junior and senior surgeons completed validated questionnaires assessing supervision, escalation of care, and safety culture pre and post-intervention along with interviews to further explore the impact of the intervention. Patient outcomes pre and postintervention were also analyzed. RESULTS: Questionnaires revealed significant improvements in supervision postintervention (senior median pre 5 vs post 7, P = 0.002 and junior 4 vs 6, P = 0.039) and senior surgeon approachability (junior 5 vs 6, P = 0.047). Both groups agreed that they would feel safer as a patient in their hospital postintervention (senior 3 vs 4.5, P = 0.021 and junior 3 vs 4, P = 0.034). The interviews confirmed that the safety culture of the department had improved. There were no differences in inpatient mortality, cardiac arrest, reoperation, or readmission rates pre and postintervention. CONCLUSION: Improving supervision and introducing clear protocols can improve safety culture on the surgical ward. Future work should evaluate the effect these measures have on patient outcomes in multiple institutions.