Machine learning identifies individuals at higher risk of incident cardio-renal-metabolic diseases and cardiovascular death who have unrealised opportunities to reduce future cardiovascular risk

Abstract

Abstract Background Machine learning may be able to identify individuals at risk of cardio-renal-metabolic events using routinely-collected data, and these individuals may be suitable for targeted preventative strategies.(1, 2) Purpose To train and test a machine learning algorithm to identify individuals at higher risk of incident cardio-renal-metabolic diseases and cardiovascular death, and then establish if there are opportunities to reduce their future cardiovascular risk. Methods We trained a random classifier (OPTIMISE) in UK primary care EHR data from 2 081 139 individuals aged ≥30 years (Jan 2, 1998, Nov 30, 2018), randomly divided into training (80%) and testing (20%) datasets. We calculated the cumulative incidence rate for ten cardio-renal-metabolic diseases and death. Fine and Gray’s models with competing risk of death were fit for each outcome between higher and lower predicted risk. In a multi-centre pilot interventional single arm study consenting individuals aged ≥30 years at higher predicted risk received cardio-renal-metabolic phenotyping and assessment for guideline target attainment. Results In the testing dataset (n = 416 228), individuals at higher predicted risk had higher long-term risk of heart failure (HR 12.54), aortic stenosis (HR 9.98), AF (HR 8·75), stroke/TIA (HR 8.07), CKD (HR 6.85), PVD (HR 6.62), valvular heart disease (HR 6.49), MI (HR 5.02), diabetes (HR 2.05) and COPD (HR 2.02) (Figure 1). This cohort were also at higher risk of death (HR 10.45), accounting for 74% of cardiovascular deaths (8 582 of 11 676) during 10-year follow up. Of 82 higher risk patients in the pilot study (mean age 71.6 years (SD 7.5), 50% women), the prevalence of cardio-renal-metabolic disease was high (Table 1), and there were opportunities to reduce future cardiovascular risk. Of higher risk patients with hypertension, 58.5% (31/53) of those aged <80 years had a systolic blood pressure (SBP)>140mmHg, and 54.5% (6/11) of those aged ≥80 years had a SBP >150mmHg. Of those with type 2 diabetes and co-existent ASCVD, only 23.1% (3/13) were on SGLT2 inhibitor therapy. Of higher risk patients on statin therapy, 37.0% (20/54) had LDL-cholesterol >1.8 mmol/L, and 52.0% (12/25) of patients with previous ASCVD had an LDL-cholesterol >1.4mmol/L. Furthermore, 19.5% (16/82) of the higher risk cohort had undiagnosed moderate or high risk CKD; who were infrequently prescribed a statin (41.7%; 5/12), ACE-i/ARB therapy with co-existent hypertension (61.5%. 8/13), or SGLT2 inhibitor with co-existent diabetes (83.3% (5/6)). Obesity was present in 49%, and 17% (14/82) were eligible for GLP-1 RA therapy. Conclusions The machine learning OPTIMISE algorithm can identify people at higher risk of cardio-renal-metabolic diseases and death using routinely collected data. On prospective evaluation higher risk individuals have unrecorded and undertreated cardio-renal-metabolic diseases, which are actionable targets for preventative care.Figure 1