Determinants of lipid clinic referral and attendance in a multi-ethnic adult population in South London

Problem

Genetic or primary dyslipidaemias are a collection of inherited disorders of lipoprotein metabolism characterised by high circulating blood lipids early in life, covering hypercholesterolaemias (including familial hypercholesterolaemia, FH), hypertriglyceridaemias and combined hyperlipidaemias. Primary dyslipidaemias are thought to account for 60% of all premature cardiovascular disease. However, early diagnosis can significantly reduce cardiovascular risk with lipid-lowering medicines. Patients identified with high risk primary dyslipidaemia are recommended for lipid clinic referral to inform patient management. Improving FH identification (currently greatly underdiagnosed), and addressing ethnic disparities in dyslipidaemia management, is a current NHS priority.

Approach

Aim: To assess determinants of lipid clinic referral or attendance in ethnically diverse adults.Retrospective cross-sectional study using the Lambeth DataNet primary care database, containing anonymised adult patient data (≥ 18 years) from 41 general practices (312,359 adult patients) in South London. The odds of lipid clinic referral was the main outcome variable, and this was assessed across ten ethnic groups and other patient-level factors: demographic (age and gender); socio-economic indicators (non-English speakers, income quintile); lifestyle factors (obesity and smoking); co-morbidities (type 2 diabetes, hypertension, ischaemic heart disease, chronic kidney disease, CKD); and practice factors (consultation frequency and practice list size). Stata 17 was used to run sequential models of logistic regression which were adjusted for practice effects using multi-level modelling.

Findings

A total of 312,359 adult patients were included in the analysis, of which 1001 (0.3%) were coded as either referred to or seen in lipid clinic. The fully adjusted model for odds of lipid clinic referral or attendance, OR (95% CI, p-value) showed the following associations: age (years) OR 1.03 (1.02-1.03, p<0.001), male gender OR 1.17 (1.03-1.33, p<0.001), Black African ethnicity OR 0.67 (0.54-0.84, p=0.001), Black Caribbean ethnicity OR 0.72 (0.57-0.89, p=0.003), obesity 1.53 (1.32-1.76, p<0.001), ex-smoker 1.55 (1.34-1.84, p<0.001), current smoker 1.25 (1.05-1.50, p=0.01), hypertension OR 1.21 (1.02-1.44, p=0.03), diabetes OR 1.63 (1.37-1.94, p<0.001), CKD OR 0.68 (0.51-0.90, p=0.008) and frequency of GP attendance ≥7 visits/last 12-months OR 2.14 (1.83-2.52, p<0.001). Sensitivity analyses restricted to FH coded patients showed a non-significant increased likelihood of lipid clinic referral for Indian OR 2.64 (0.69-10.17, p=0.16), Black Caribbean OR 1.79 (0.70-4.55, p=0.22) OR and Black African OR 1.76 (0.75-4.11, p=0.19) individuals, compared to White ethnic groups.

Consequences

Factors associated with a higher likelihood of lipid clinic referral or attendance included older age, male gender, obesity, ex or current smokers, hypertension, diabetes and high frequency attenders to primary care. Individuals recorded as being of Black African and Black Caribbean ethnicity, and individuals with CKD were less likely to have a coded lipid clinic referral or attendance, after adjusting for other determinants including deprivation measures, suggesting large inequalities in health care access in these groups.