Machine Learning Applied to Cholesterol-Lowering Pharmacotherapy: Proof-of-Concept in High-Risk Patients Treated in Primary Care.

Objectives: Machine learning has potential to improve the management of lipid disorders. We explored the utility of machine learning in high-risk patients in primary care receiving cholesterol-lowering medications. Methods: Machine learning algorithms were created based on lipid management guidelines for England [National Institute for Health and Care Excellence (NICE) CG181] to reproduce the guidance with >95% accuracy. Natural language processing and therapy identification algorithms were applied to anonymized electronic records from six South London primary care general practices to extract medication information from free text fields. Results: Among a total of 48,226 adult patients, a subset of 5630 (mean ± standard deviation, age = 67 ± 13 years; male:female = 55:45) with a history of lipid-lowering therapy were identified. Additional major cardiometabolic comorbidities included type 2 diabetes in 13% (n = 724) and hypertension in 32% (n = 1791); all three risk factors were present in a further 28% (n = 1552). Of the 5630 patients, 4290 (76%) and 1349 (24%) were in primary and secondary cardiovascular disease prevention cohorts, respectively. Statin monotherapy was the most common current medication (82%, n = 4632). For patients receiving statin monotherapy, 71% (n = 3269) were on high-intensity therapy aligned with NICE guidance with rates being similar for the primary and secondary prevention cohorts. In the combined cohort, only 46% of patients who had been prescribed lipid-lowering therapy in the previous 12 months achieved the NICE treatment goal of >40% reduction in non-high-density lipoprotein cholesterol from baseline pretreatment levels. Based on the most recent data entry for patients not at goal the neural network recommended either increasing the dose of statin, adding complementary cholesterol-lowering medication, or obtaining an expert lipid opinion. Conclusions: Machine learning can be of value in (a) quantifying suboptimal lipid-lowering prescribing patterns, (b) identifying high-risk patients who could benefit from more intensive therapy, and (c) suggesting evidence-based therapeutic options.

Evolution of Type 2 Diabetes Management from a Glucocentric Approach to Cardio-Renal Risk Reduction: The New Paradigm of Care.

For many years, clinical studies could not show that lowering glucose in patients with type 2 diabetes leads to better macrovascular outcomes. In the past few years, new data have shown that treatment with two classes of dugs developed as "glucose-lowering agents," SGLT2 inhibitors and GLP-1 receptor agonists, can reduce macrovascular and renal complications. These studies have prompted debate about the main aim of type 2 diabetes management. In this review, three eras of diabetes management are described according to the treatment recommendations, such as the ADA/EASD consensus, moving from a pure glucocentric view into the present cardio-renal outcome-oriented approach, this has been endorsed by major diabetes and cardiology societies. While in the first era normalizing HbA1c was the only focus (e.g., UK Prospective Diabetes Study), failing to show a reduction in cardiovascular morbidity and mortality, further studies analyzing the pros and cons of intensified control such as ACCORD, VADT, ADVANCE recognized that treatment intensification was associated with weight gain and hypoglycemia, thereby potentially reducing the benefits of glycemic control. Therefore, the focus in the second area was on controlling HbA1c without these unwanted effects. The consistent beneficial results of several cardiovascular outcome trials with SGLT2 inhibitors and GLP-1 receptor agonists showing significantly improved cardio-renal outcomes, induced a paradigm shift: a change from (only) control of HbA1c to an organ-protective approach with the main focus now on cardio-renal risk; this is now considered as the third era. Recent data indicating beneficial effects of glucose-lowering agents in particular SGLT2 inhibitors even in subjects without diabetes, improving hospitalization for heart failure and renal outcomes might reveal another new era, which could then be considered a fourth era. While current international guidelines call for this paradigm shift, registry data show that we are still far from translating this objective into real-world practice.

Glucose: archetypal biomarker in diabetes diagnosis, clinical management and research.

The clinical utility of diabetes biomarkers can be considered in terms of diagnosis, management and prediction of long-term vascular complications. Glucose satisfies all of these requirements. Thresholds of hyperglycemia diagnostic of diabetes reflect inflections that confer a risk of developing long-term microvascular complications. Degrees of hyperglycemia (impaired fasting glucose, impaired glucose tolerance) that lie below the diagnostic threshold for diabetes identify individuals at risk of progression to diabetes and/or development of atherothrombotic cardiovascular disease. Self-measured glucose levels usefully complement hemoglobin A1c levels to guide daily management decisions. Continuous glucose monitoring provides detailed real-time data that is of value in clinical decision making, assessing response to new diabetes drugs and the development of closed-loop artificial pancreas technology.

Adipocytokine profiles in a putative novel postmenopausal polycystic ovary syndrome (PCOS) phenotype parallel those in premenopausal PCOS: the Rancho Bernardo Study.

The objective was to investigate whether the associations between leptin, adiponectin, andadiposity reported in classic polycystic ovary syndrome (PCOS) are also observed in elderly women with a novel putative postmenopausal PCOS phenotype. We studied 713 postmenopausal community-dwelling women. Diagnosis of the novel phenotype required the presence of ≥3 diagnostic features including: 1) a personal history of oligomenorrhea; 2) history of infertility or miscarriage; 3) current or past clinical or hormonal evidence of hyperandrogenism; 4) central obesity; 5) biochemical evidence of insulin resistance. Women in the control group had ≤2 of these components. Mean age (±SD) was 74±8 years for the study cohort. Sixty-six women (9.3%) had the putative PCOS phenotype. Serum leptin was higher (mean 25.70±15.67 vs 14.94+9.89 ng/mL, P<.01) and adiponectin lower (mean 11.72±4.80 vs 17.31±7.45 µg/mL, P<.01) in cases vs controls. Leptin was positively, and adiponectin inversely, associated with an increasing number of phenotype features (P<.01 for linearity). In age-adjusted regression analysis, adjustment for waist circumference eliminated the association between leptin and the PCOS phenotype, but not the association between adiponectin and the PCOS phenotype. In this novel postmenopausal PCOS phenotype, adipocytokine profiles and their associations with adiposity parallel those reported in younger women with classic PCOS. These results support our hypothesis that a putative phenotype analogous to PCOS can be identified in postmenopausal women using clinical and biochemical criteria. Use of this novel phenotype could provide a basis for studies of the delayed consequences of PCOS in older women.