Cost-effectiveness of statin therapies for older people in the UK

Aims and objectives

The use of statin therapy in older people has been more limited than in other patient categories68,69 despite increasing CVD risks with increasing age and the growing number and share of older people. In particular, there is less definitive evidence of statin benefit among older people without previous CVD,5 and guidelines often fall short of making specific recommendations on the use of statins for primary CVD prevention in older people and instead suggest patient–physician discussion about individual patient circumstances.15,70,71

We set out to assess the lifetime net health effect and cost-effectiveness of statin treatment in older people in the present-day UK using the UK CVD microsimulation model, together with the effects of statin in older people reported by an IPD meta-analysis of large statin trials.5

Methods

Study population

We assessed the lifetime effectiveness and cost-effectiveness of statin therapy in categories of UK adults ≥ 70 years old in the UKB31 and the Whitehall II32 studies by sex, previous CVD and pre-treatment LDL-C levels. All UKB participants ≥ 70 years old at recruitment into the study or at any subsequent study resurvey were included in the present study from their earliest eligible attendance. All Whitehall II participants ≥ 70 years old at Phase 9 of the study were included in the present study.

Cardiovascular disease microsimulation model validation in older people

The development and assessment of the CVD microsimulation policy model have been reported in Chapters 2 and 3 of this report and the healthcare costs and QoL models in Chapters 4 and 5, respectively. Here, the model performance was further assessed among participants ≥ 70 years old in UKB and Whitehall II studies using their linked electronic death records, hospital admissions, primary care records (UKB only) and cancer registrations to identify MI, strokes, CRV (UKB only), incident diabetes (UKB only) and incident cancers and deaths during follow-up.

Cost-effectiveness of statin therapy in older people

The effects and costs of standard and higher-intensity statin therapy are described in Table 13 and Chapter 6.

In the base-case analysis of cost-effectiveness of statin therapy in older people, we assumed that similar rate ratio (RR) reductions in cardiovascular events per 1-mmol/l LDL-C reduction were achieved independently of age at statin initiation as reported in the IPD meta-analysis.5 In a key scenario analysis, we applied RR reductions in cardiovascular end points per 1 mmol/l LDL-C reduction with statin therapy from age 75 years onwards, informed from data only among people > 75 years of age in that meta-analysis. Finally, to further explore the limited randomised evidence in older people without previous CVD, we report a further scenario analysis employing the effect of statin therapy on the risk of major vascular event per 1-mmol/l LDL-C reduction, as reported in this category of trial participants by the CTTC (RR 0.92, 99% CI 0.73 to 1.16)5 for the risks of MI and stroke of statin-treated individuals in the model. In this scenario analysis, we also assumed that statin treatment did not affect risks of CRV and VD.

Our assessment of parameter uncertainty and the sensitivity analyses for statin cost-effectiveness in older people followed the approach described in Chapter 6 and Appendix 6, Table 30, for people 40–70 years old. Further scenario analyses with doubled risk of NVD, with general QoL reduced by 0.1 and with both assumptions together, were executed to assess sensitivity to further reduced potential in older people to benefit from preventive cardiovascular treatment.

Results

The baseline characteristics of participants ≥ 70 years old identified from both UKB and Whitehall II studies are presented in Table 16. There were 15,019 (52% men; mean age 72.5 years) participants without previous CVD and 5103 (66% men; mean age 72.9 years) with previous CVD. The derived untreated mean LDL-C levels were 4.2 mmol/l (SD 0.78 mmol/l) and 4.3 mmol/l (SD 0.98 mmol/l) among participants without and participants with previous CVD, respectively.

In model validation, the cumulative event rates predicted by the CVD microsimulation model, using the baseline characteristics of participants ≥ 70 years in UKB and Whitehall II cohorts, corresponded well to observed event rates (Figure 11).

FIGURE 11 Cardiovascular disease microsimulation model validation among UKB and Whitehall II participants aged ≥ 70 years. In the Whitehall II study, no linked data for CRV and diabetes were available and, therefore, CRV and diabetes were excluded from model validation. (Figure not reproduced here; see the original article.)

In participant categories by sex, previous CVD and pre-treatment LDL-C level, in the base-case analysis standard statin therapy was projected to increase individual QALYs (undiscounted) by 0.24–0.70, and higher-intensity statin therapy by a further 0.04–0.13 QALYs (Table 17). Across these categories, standard statin therapy compared with no statin had an incremental cost per QALY gained ranging from £120 to £3500 and higher-intensity compared with standard statin therapy ranging from £2210 to £11,780 (Table 18, and see Appendix 8, Table 33). In our key scenario analysis, the cost-effectiveness was not materially affected if relative risk reductions of cardiovascular events were equal to those reported only in the subgroup of participants > 75 years old, although the gains in QALYs were somewhat smaller (0.15–0.48 with standard statin and 0.02–0.09 with higher-intensity statin across participant categories) (see Tables 17 and 18). In the further scenario analysis using statin effects among older people without previous CVD for risks of MI and stroke and no statin effects on risks of CRV and VD, a smaller benefit was projected with lifetime statin therapy, although standard statin therapy remained cost-effective in participant categories by sex and level of pre-treatment LDL-C (see Tables 17 and 18).

The analyses of parameter uncertainty indicated that, in the base-case analysis, either standard or higher-intensity statin therapy was certain to be cost-effective at willingness-to-pay thresholds as low as £5000 per QALY (Figure 12, and see Appendix 8, Table 33). At a willingness-to-pay threshold of £20,000 per QALY, higher-intensity statin therapy had the highest probability of being cost-effective across all categories of men and women ≥ 70 years old in the base-case cost-effectiveness analysis (see Figure 12). At a cost-effectiveness threshold of £5000 per QALY, however, standard statin therapy had the highest probability of being cost-effective among women with pre-treatment LDL-C lower than 4.1 mmol/l and men with pre-treatment LDL-C lower than 3.4 mmol/l (see Figure 12). In the key scenario analysis, although at the £20,000-per-QALY threshold higher-intensity statin therapy retained the highest probability of being cost-effective, at £5000 per QALY standard statin therapy had the highest probability of cost-effectiveness in most categories of older men and women (see Figure 12). In the further scenario analysis among older people without previous CVD, at the £20,000-per-QALY threshold, higher-intensity statin therapy had the highest probability of being cost-effective for men and women with LDL-C ≥ 3.4 mmol/l but standard statin therapy was most probably cost-effective at lower LDL-C level (see Figure 12). In this scenario, at the £5000-per-QALY threshold, standard statin therapy was cost-effective only for older men with LDL-C ≥ 3.4 mmol/l and older women with LDL-C ≥ 4.1 mmol/l. The analyses of parameter uncertainty indicated that in the key and further scenario analyses standard or higher-intensity statin therapy remained more likely to be cost-effective than no statin treatment across all patient categories at a willingness-to-pay threshold of £20,000 per QALY, but the level of uncertainty was larger (see Figure 12 and Appendix 8, Table 34).

FIGURE 12 Probability that lifetime statin therapy is cost-effective in older people. (a) Men; and (b) women. (Figure not reproduced here; see the original article.)

The cost-effectiveness results remained robust in a wide range of sensitivity analyses for the base-case scenario analysis (Table 19) with higher sensitivity noted for higher-intensity statin at a price five times higher. The results remained robust in the sensitivity analyses for the key scenario analysis, including with further annual decline in RR reductions of cardiovascular events with statin therapy, although higher sensitivity was noted for cost-effectiveness of higher-intensity statin in older women and in older men with further declining risk reductions and lower pre-treatment LDL-C level at a price five times higher (Table 20). The results were not materially different in scenario analyses with doubled risk of NVD, with general QoL reduced by 0.1 and with both (results not shown).

Summary of findings

This assessment of the long-term effectiveness and cost-effectiveness of statin therapy in people aged ≥ 70 years in the UK used a validated CVD microsimulation model and the totality of randomised evidence of effects of statin treatment effects in older people. It concluded that lifetime statin treatment was likely to increase QoL adjusted survival in older men and women, irrespective of their CVD history or level of LDL-C and, at the current UK cost of generic statins, was likely to be cost-effective for all patient categories studied. Higher-intensity statin therapy was the strategy likely to bring the highest health benefit cost-effectively, although standard statin regimens would also achieve most of the benefit and were preferred in some scenarios and sensitivity analyses. These cost-effectiveness results remained robust in sensitivity analyses, including those with somewhat smaller CVD risk reductions with statin therapy in older individuals. It is noted that reductions of major CVD events with statin therapy have not been independently established among older people without prior CVD in randomised clinical trials and, therefore, the conclusions for cost-effectiveness may be less certain in this category.

Report Supplementary Material 1 includes the Consolidated Health Economic Evaluation Reporting Standards 2022 (CHEERS 2022) Checklist.