Cost-utility analysis: what is the most cost-effective target for cholesterol for secondary prevention of cardiovascular disease? Evidence to update NICE guideline CG181 - Cardiovascular disease: risk assessment and reduction, including lipid modification

Study type
Protocol
Date of Approval
Study reference ID
22_002517
Lay Summary

This study is estimating the costs and benefits associated with different target levels for cholesterol for people who had a cardiovascular disease (CVD) event in the past, such as a stroke.

People who had a CVD event are usually treated with a medicine called statin that reduces the level of “bad” cholesterol in the blood. High levels of bad cholesterol increase the chance of people having another CVD event. When a statin alone does not reduce bad cholesterol below a certain target, other more expensive treatments are available. However, the best target for bad cholesterol is uncertain as the lower the target the more the NHS will have to spend on medicines.

An economic model will be developed and will incorporate data from a variety of sources. CPRD primary care and linked hospital admissions data, which contains data on a representative sample of people in England, will be used to estimate the distribution of bad cholesterol across the English population who have had a CVD event , and how it changes over time. The same dataset will be used to measure the chance of having another CVD event and of dying because of it. Clinical trials will be used to estimate change in bad cholesterol resulting from each medicine. The model will calculate healthcare costs and people’s health (quality and length of life) to find the most cost-effective target in England. This will inform the guideline committee in setting a target that GPs should follow when treating people with CVD.

Technical Summary

A health economic model will be developed using data from a variety of sources including CPRD-HES-ONS. The analysis will estimate the cost-effectiveness of different cholesterol targets in primary care for people who have CVD and are on a statin. In the model, people whose cholesterol is above a specified target are given another cholesterol-lowering drug in addition to their statin.

The analysis requires CPRD data to inform baseline characteristics and risks of the population (that is the people that have not been given an additional drug). Censoring occurs at death or at the suspension of statin therapy or addition of the lipid modification therapy. The CPRD analysis will estimate the following:
-distribution of CVD diagnoses
-distribution of LDL and non-HDL cholesterol
-average change of LDL and non-HDL associated with ageing
-the rate of subsequent CVD events
-CVD and non-CVD mortality with or without a recent subsequent CVD event

Established CVD is defined as a diagnosis of: ischaemic stroke, TIA, peripheral artery disease (including non-coronary revascularisation), myocardial infarction, other coronary heart disease (including unstable angina, stable angina, coronary revascularisation).

Subsequent events are defined as hospitalisation for: ischaemic stroke, myocardial infarction, unstable angina, elective coronary revascularisation, non-coronary revascularisation.

Exposure is defined as having established CVD and being on a statin. The analysis is a descriptive study to identify baseline parameters that will be fed into the health economic model. As such, no comparator needs to be defined. Rates and average changes in cholesterol will be calculated with their 95% confidence intervals. No formal statistical tests will be conducted.

The health economic model will calculate healthcare costs and quality-adjusted life-years to assess the cost-effectiveness of different cholesterol targets in England. This will inform the guideline committee in setting targets for treating people with CVD, so that patient outcomes are improved.

Health Outcomes to be Measured

Intermediate outcomes:
- Distribution of established CVD event types* stratified by gender;
- The median age and proportion of people in different subgroups of patients defined by gender and calculated LDL-cholesterol level;
- The median age and proportion of people in different subgroups of patients defined by gender and calculated non-HDL-cholesterol level;
- Average annual change in LDL-cholesterol associated with ageing;
- Average annual change in non-HDL-cholesterol associated with ageing;

End-point outcomes (events per person-year)
- Rates of each type of subsequent CVD** event for patients currently on a statin with an established CVD event, by age and gender;
- CVD mortality rates for people with established CVD* who have not experienced a subsequent event** in the last 12 months stratified by age and gender
- Non-CVD mortality rates for people with established CVD* who have not experienced a subsequent event** in the last 12 months stratified by age and gender
- CVD mortality rates for people with an established CVD event* and who have experienced a subsequent event** in the last 12 months stratified by age and gender
- Non-CVD mortality rates for people with an established CVD event* and who have experienced a subsequent event** in the last 12 months stratified by age and gender

Each outcome will be a parameter in the health economic model. Therefore, we do not distinguish between primary and secondary outcomes.

*Established CVD is defined as a diagnosis of: ischaemic stroke, peripheral artery disease (including non-coronary revascularisation), myocardial infarction, other coronary heart disease (including unstable angina, stable angina, coronary revascularisation)
**Hospitalisation for: ischaemic stroke, myocardial infarction, unstable angina, elective coronary revascularisation, non-coronary revascularisation
*** CVD death is defined as: death caused by CVD including coronary heart disease, heart failure and ischaemic stroke
**** Non-CVD death is death due to any cause other than CVD

Collaborators

Alfredo Mariani - Chief Investigator - National Institute for Health and Clinical Excellence - NICE
Alfredo Mariani - Corresponding Applicant - National Institute for Health and Clinical Excellence - NICE
David Preiss - Collaborator - University of Oxford
David Wonderling - Collaborator - National Institute for Health and Clinical Excellence - NICE
Eleanor Yelland - Collaborator - CPRD
Jessie Oyinlola - Collaborator - CPRD
Mike Lonergan - Collaborator - CPRD
Patrick Muller - Collaborator - National Institute for Health and Clinical Excellence - NICE
Rachael Williams - Collaborator - CPRD
Riyaz Patel - Collaborator - Barts Health and UCLH NHS Trusts

Former Collaborators

Mike Lonergan - Collaborator - CPRD

Linkages

HES Admitted Patient Care;ONS Death Registration Data