Current approaches to predict cardiovascular risk fail to identify many people who would benefit from preventive treatment, while others receive unnecessary intervention. Machine-learning offers opportunity to improve accuracy by exploiting complex interactions between risk factors. We assessed whether machine-learning can improve cardiovascular risk prediction. Prospective cohort study using routine clinical data of 378,256 patients from UK family practices, free from cardiovascular disease at outset. Four machine-learning algorithms (random forest, logistic regression, gradient boosting machines, neural networks) were compared to an established algorithm (American College of Cardiology guidelines) to predict first cardiovascular event over 10-years. Predictive accuracy was assessed by area under the ‘receiver operating curve’ (AUC); and sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV) to predict 7.5% cardiovascular risk (threshold for initiating statins). 24,970 incident cardiovascular events (6.6%) occurred. Compared to the established risk prediction algorithm (AUC 0.728, 95% CI 0.723–0.735), machine-learning algorithms improved prediction: random forest +1.7% (AUC 0.745, 95% CI 0.739–0.750), logistic regression +3.2% (AUC 0.760, 95% CI 0.755–0.766), gradient boosting +3.3% (AUC 0.761, 95% CI 0.755–0.766), neural networks +3.6% (AUC 0.764, 95% CI 0.759–0.769). The highest achieving (neural networks) algorithm predicted 4,998/7,404 cases (sensitivity 67.5%, PPV 18.4%) and 53,458/75,585 non-cases (specificity 70.7%, NPV 95.7%), correctly predicting 355 (+7.6%) more patients who developed cardiovascular disease compared to the established algorithm. Machine-learning significantly improves accuracy of cardiovascular risk prediction, increasing the number of patients identified who could benefit from preventive treatment, while avoiding unnecessary treatment of others.

Background: Heavy menstrual bleeding (HMB) is a common problem, yet evidence to inform decisions about initial medical treatment is limited.Objectives: To assess the clinical effectiveness and cost-effectiveness of the levonorgestrel-releasing intrauterine system (LNG-IUS) (Mirena^®, Bayer) compared with usual medical treatment, with exploration of women’s perspectives on treatment.Design: A pragmatic, multicentre randomised trial with an economic evaluation and a longitudinal qualitative study.Setting: Women who presented in primary care.Participants: A total of 571 women with HMB. A purposeful sample of 27 women who were randomised or ineligible owing to treatment preference participated in semistructured face-to-face interviews around 2 and 12 months after commencing treatment.Interventions: LNG-IUS or usual medical treatment (tranexamic acid, mefenamic acid, combined oestrogen–progestogen or progesterone alone). Women could subsequently swap or cease their allocated treatment.Outcome measures: The primary outcome was the patient-reported score on the Menorrhagia Multi-Attribute Scale (MMAS) assessed over a 2-year period and then again at 5 years. Secondary outcomes included general quality of life (QoL), sexual activity, surgical intervention and safety. Data were analysed using iterative constant comparison. A state transition model-based cost–utility analysis was undertaken alongside the randomised trial. Quality-adjusted life-years (QALYs) were derived from the European Quality of Life-5 Dimensions (EQ-5D) and the Short Form questionnaire-6 Dimensions (SF-6D). The intention-to-treat analyses were reported as cost per QALY gained. Uncertainty was explored by conducting both deterministic and probabilistic sensitivity analyses.Results: The MMAS total scores improved significantly in both groups at all time points, but were significantly greater for the LNG-IUS than for usual treatment [mean difference over 2 years was 13.4 points, 95% confidence interval (CI) 9.9 to 16.9 points;p < 0.001]. However, this difference between groups was reduced and no longer significant by 5 years (mean difference in scores 3.9 points, 95% CI –0.6 to 8.3 points;p = 0.09). By 5 years, only 47% of women had a LNG-IUS in place and 15% were still taking usual medical treatment. Five-year surgery rates were low, at 20%, and were similar, irrespective of initial treatments. There were no significant differences in serious adverse events between groups. Using the EQ-5D, at 2 years, the relative cost-effectiveness of the LNG-IUS compared with usual medical treatment was £1600 per QALY, which by 5 years was reduced to £114 per QALY. Using the SF-6D, usual medical treatment dominates the LNG-IUS. The qualitative findings show that women’s experiences and expectations of medical treatments for HMB vary considerably and change over time. Women had high expectations of a prompt effect from medical treatments.Conclusions: The LNG-IUS, compared with usual medical therapies, resulted in greater improvement over 2 years in women’s assessments of the effect of HMB on their daily routine, including work, social and family life, and psychological and physical well-being. At 5 years, the differences were no longer significant. A similar low proportion of women required surgical intervention in both groups. The LNG-IUS is cost-effective in both the short and medium term, using the method generally recommended by the National Institute for Health and Care Excellence. Using the alternative measures to value QoL will have a considerable impact on cost-effectiveness decisions. It will be important to explore the clinical and health-care trajectories of the ECLIPSE (clinical effectiveness and cost-effectiveness of levonorgestrel-releasing intrauterine system in primary care against standard treatment for menorrhagia) trial participants to 10 years, by which time half of the cohort will have reached menopause.Trial registration: Current Controlled Trials ISRCTN86566246.Funding: This project was funded by the NIHR Health Technology Assessment programme and will be published in full inHealth Technology Assessment; Vol. 19, No. 88. See the NIHR Journals Library website for further project information

While ethnic disparities in health and health care are increasing, evidence on how to enhance quality of care and reduce inequalities remains limited. Despite growth in the scope and application of guidelines on “cultural competence,” remarkably little is known about how practising health professionals experience and perceive their work with patients from diverse ethnic communities. Using cancer care as a clinical context, we aimed to explore this with a range of health professionals to inform interventions to enhance quality of care. We conducted a qualitative study involving 18 focus groups with a purposeful sample of 106 health professionals of differing disciplines, in primary and secondary care settings, working with patient populations of varying ethnic diversity in the Midlands of the UK. Data were analysed by constant comparison and we undertook processes for validation of analysis. We found that, as they sought to offer appropriate care, health professionals wrestled with considerable uncertainty and apprehension in responding to the needs of patients of ethnicities different from their own. They emphasised their perceived ignorance about cultural difference and were anxious about being culturally inappropriate, causing affront, or appearing discriminatory or racist. Professionals' ability to think and act flexibly or creatively faltered. Although trying to do their best, professionals' uncertainty was disempowering, creating a disabling hesitancy and inertia in their practice. Most professionals sought and applied a knowledge-based cultural expertise approach to patients, though some identified the risk of engendering stereotypical expectations of patients. Professionals' uncertainty and disempowerment had the potential to perpetuate each other, to the detriment of patient care. This study suggests potential mechanisms by which health professionals may inadvertently contribute to ethnic disparities in health care. It identifies critical opportunities to empower health professionals to respond more effectively. Interventions should help professionals acknowledge their uncertainty and its potential to create inertia in their practice. A shift away from a cultural expertise model toward a greater focus on each patient as an individual may help.

Prognostic modelling using standard methods is well-established, particularly for predicting risk of single diseases. Machine-learning may offer potential to explore outcomes of even greater complexity, such as premature death. This study aimed to develop novel prediction algorithms using machine-learning, in addition to standard survival modelling, to predict premature all-cause mortality. A prospective population cohort of 502,628 participants aged 40–69 years were recruited to the UK Biobank from 2006–2010 and followed-up until 2016. Participants were assessed on a range of demographic, biometric, clinical and lifestyle factors. Mortality data by ICD-10 were obtained from linkage to Office of National Statistics. Models were developed using deep learning, random forest and Cox regression. Calibration was assessed by comparing observed to predicted risks; and discrimination by area under the ‘receiver operating curve’ (AUC). 14,418 deaths (2.9%) occurred over a total follow-up time of 3,508,454 person-years. A simple age and gender Cox model was the least predictive (AUC 0.689, 95% CI 0.681–0.699). A multivariate Cox regression model significantly improved discrimination by 6.2% (AUC 0.751, 95% CI 0.748–0.767). The application of machine-learning algorithms further improved discrimination by 3.2% using random forest (AUC 0.783, 95% CI 0.776–0.791) and 3.9% using deep learning (AUC 0.790, 95% CI 0.783–0.797). These ML algorithms improved discrimination by 9.4% and 10.1% respectively from a simple age and gender Cox regression model. Random forest and deep learning achieved similar levels of discrimination with no significant difference. Machine-learning algorithms were well-calibrated, while Cox regression models consistently over-predicted risk. Machine-learning significantly improved accuracy of prediction of premature all-cause mortality in this middle-aged population, compared to standard methods. This study illustrates the value of machine-learning for risk prediction within a traditional epidemiological study design, and how this approach might be reported to assist scientific verification.

Objective To assess low-density lipoprotein cholesterol (LDL-C) response in patients after initiation of statins, and future risk of cardiovascular disease (CVD). Methods Prospective cohort study of 165 411 primary care patients, from the UK Clinical Practice Research Datalink, who were free of CVD before statin initiation, and had at least one pre-treatment LDL-C within 12 months before, and one post-treatment LDL-C within 24 months after, statin initiation. Based on current national guidelines, <40% reduction in baseline LDL-C within 24 months was classified as a sub-optimal statin response. Cox proportional regression and competing-risks survival regression models were used to determine adjusted hazard ratios (HRs) and sub-HRs for incident CVD outcomes for LDL-C response to statins. Results 84 609 (51.2%) patients had a sub-optimal LDL-C response to initiated statin therapy within 24 months. During 1 077 299 person-years of follow-up (median follow-up 6.2 years), there were 22 798 CVD events (12 142 in sub-optimal responders and 10 656 in optimal responders). In sub-optimal responders, compared with optimal responders, the HR for incident CVD was 1.17 (95% CI 1.13 to 1.20) and 1.22 (95% CI 1.19 to 1.25) after adjusting for age and baseline untreated LDL-C. Considering competing risks resulted in lower but similar sub-HRs for both unadjusted (1.13, 95% CI 1.10 to 1.16) and adjusted (1.19, 95% CI 1.16 to 1.23) cumulative incidence function of CVD. Conclusions Optimal lowering of LDL-C is not achieved within 2 years in over half of patients in the general population initiated on statin therapy, and these patients will experience significantly increased risk of future CVD.

Heterozygous familial hypercholesterolaemia (FH) is a common autosomal dominant disorder. The vast majority of affected individuals remain undiagnosed, resulting in lost opportunities for preventing premature heart disease. Better use of routine primary care data offers an opportunity to enhance detection. We sought to develop a new predictive algorithm for improving identification of individuals in primary care who could be prioritised for further clinical assessment using established diagnostic criteria. Data were analysed for 2,975,281 patients with total or LDL-cholesterol measurement from 1 Jan 1999 to 31 August 2013 using the Clinical Practice Research Datalink (CPRD). Included in this cohort study were 5050 documented cases of FH. Stepwise logistic regression was used to derive optimal multivariate prediction models. Model performance was assessed by its discriminatory accuracy (area under receiver operating curve [AUC]). The FH prediction model (FAMCAT), consisting of nine diagnostic variables, showed high discrimination (AUC 0.860, 95% CI 0.848-0.871) for distinguishing cases from non-cases. Sensitivity analysis demonstrated no significant drop in discrimination (AUC 0.858, 95% CI 0.845-0.869) after excluding secondary causes of hypercholesterolaemia. Removing family history variables reduced discrimination (AUC 0.820, 95% CI 0.807-0.834), while incorporating more comprehensive family history recording of myocardial infraction significantly improved discrimination (AUC 0.894, 95% CI 0.884-0.904). This approach offers the opportunity to enhance detection of FH in primary care by identifying individuals with greatest probability of having the condition. Such cases can be prioritised for further clinical assessment, appropriate referral and treatment to prevent premature heart disease.