The Swiss Medical Board (SMB) published a report showing costs per quality adjusted life years (cost per quality-adjusted life year [cost/QALY]) to be extremely unfavourable (CHF 210 000/QALY) for statins in primary care in Switzerland when administered for a 5-year period. In its final report, the SMB recommended that statins should be used in primary prevention only if the SCORE risk is at least 7.5% over 10 years . This is in contrast to current Swiss guidelines, which recommend that patients with a risk of 5% in 10 years for fatal cardiovascular disease may be offered statin medication.
In a Framingham Offspring Study, improved coronary risk stratification by the use of lower thresholds of coronary risk in younger subjects showed increased sensitivity at somewhat maintained specificity . Further, higher levels of low density lipoprotein (LDL) cholesterol at the same risk levels improved prediction of who might benefit more from statins (an observation from the U.S. NHANES study ). In the Copenhagen General Population Study, a SCORE threshold of 2.4% appeared sufficiently sensitive for the treatment of those with future cardiovascular events . In 3172 subjects hospitalised for a first myocardial infarction in Switzerland, the sensitivity of a SCORE threshold of 5% was only 37%, and only 16% of these 3172 subjects were on statin treatment when the myocardial infarction occurred . Therefore, clinical evidence suggests that the SCORE threshold for starting statin treatment, based on prevention of cardiovascular events, should be lower than 5% in 10 years, but costs/QALY have not been reported for lower decision thresholds in Switzerland and Germany.
We first aimed to determine cost/QALY for various thresholds of SCORE risk, using the SMB assumptions, for treatment periods of 5 and of 10 years. Second, we looked at the risk distribution for SCORE in two large primary care populations with advanced carotid atherosclerosis and tried to define a sufficiently sensitive SCORE threshold.
The two groups of subjects were examined at a practice-based level. In the Swiss Imaging Centre in Olten, subjects were referred by their general practitioner (GP) or self-referred (56% and 44%, respectively) to the Vascular Risk Foundation (Olten, Switzerland). In the German Imaging Centre in Koblenz, all subjects were referred within a working medicine setting. Subjects had to be free of cardiovascular symptoms or diseases. Diabetic subjects and those with previous cardiovascular disease were excluded from this analysis because they have a high coronary risk by definition. Laboratory values and blood pressure were measured locally and these results, with the medical history, were entered into a data spreadsheet (Excel, Microsoft, Richmond, USA).
Presence and amount of longitudinal carotid plaque surface was imaged with a high-resolution ultrasound linear transducer using a 7.5–12.0 MHz probe, which identified all plaques defined by intimal thickening ≥1.0 mm. The longitudinal area of all plaques was summed up to compute the value for the total plaque area (TPA) in mm2. All TPA measurements were made by A.A. in Koblenz and by M.R. in Olten. A TPA ≥80 mm2 (TPA80) defined a coronary risk equivalent (risk >20% for fatal and nonfatal myocardial infarction within 10 years) . Intraobserver reproducibility (M.R.) was tested for the right carotid artery in 57 patients with a correlation coefficient of r2 0.964 (left carotid artery: r2 0.944, both arteries r2 0.986). For the cutoffs of TPA 0–9 mm2, 10–49 mm2, 50–99 mm2 and greater than 100 mm2, the kappa value was 0.69 (95% confidence interval [CI] 0.54–0.84). The sensitivity to detect TPA80 was calculated for three SCORE thresholds (2.5, 5.0 and 7.5%).
Computation of risk
Cardiovascular risk was computed with the published risk formulas in an excel spreadsheet (Excel, Microsoft, Richmond, USA), with use of SCORE for low risk populations .
Effect model of the SMB
The SMB model for calculating cost/QALY is as follows. For one fatal cardiovascular event (myocardial infarction, stroke, coronary revascularisation), 4.5 nonfatal events occur. The cost is CHF 8500 per fatal event and CHF 25 000 per nonfatal event in the first year and CHF 8000 in subsequent years. Loss of QALY is 1.0 for fatal and 0.2 for nonfatal events. The annual preventive medical cost per individual, including statin costs, is CHF 470, all cardiovascular events occur uniformly after 50% of the total observation time of five years. Loss of QALY at 2.5 years was therefore 2 × 2.5 ×1 = 5.0 QALY for fatal events and 9 × 2.5 × 0.2 = 4.5 QALY for nonfatal events, and thus 5.0 + 4.5 = 9.5 QALY in 1000 persons or 0.0095 QALY per person. When this effect model is applied to a 10-year period, then 4 fatal events and 18 nonfatal events can be prevented; therefore, 4 × 5 × 1 = 20 QALY for fatal and 18 × 5 × 0.2 = 18 QALY for nonfatal events, or a total of 38 QALY losses, can be prevented in 1000 persons, which is 0.038 QALY per person. Therefore, the effect model is 4 times higher in 10 years when compared with 5 years. The SMB based its assumptions on the Cholesterol Treatment Trialists’ (CTT) study published in 2012 and on the effect of statins on LDL, which is assumed to be a reduction of 1.0 mmol/l to obtain the above mentioned preventive effects . The SMB QALY calculator can be accessed online for further details and for calculation of examples .
Effects of statins in primary care
The aggregated relative risk reduction per 1 mmol/l LDL reduction used by the SMB is 22% [8, 10]. This 22% relative risk reduction was chosen by the SMB despite the fact, that in low risk subjects, relative risk reduction was shown to be 30% for major vascular events (CTT Appendix 2012, web figure 5, statins versus placebo only, 22 trials) . In order to calculate cost/QALY, we used relative risk reductions of 22% and of 30%.
For practice-based subjects referred by GPs or within the working setting, informed consent for imaging of carotid arteries and measuring coronary risk factors was the reason for encounter; further preventive therapies were left to the decision of the referring GP. The self-referred subjects of the Vascular Risk Foundation privided informed written consent. The study (Cordicare II) was approved by the joint ethics committee of the cantons of Solothurn and Aargau.
Sensitivity (see table 4) and number needed to treat (NNT, see table 2) were calculated with established formulas. The SMB calculations are outlined in the SMB effect model described above.
Study population description
On average, the Swiss group (n = 2203) included older subjects than the German group (n = 2942, 57 ± 9 vs 46 ± 10 years) and more women (49 vs 34%). The assessment of 10-year risk for both groups showed that most subjects were in the low risk category. The prevalence of TPA80 was higher in Switzerland (22 vs 15%). Lipid profiles were comparable. Average coronary risk was low (table 1).
|Table 1: Baseline characteristics, prevalence of risk and average total plaque area for Switzerland and Germany.|
|Number of subjects (n)||2203|| || ||2942|| || |
|Female (n, %)||1083|| ||49%||989|| ||34%|
|Age, years (mean ± 1 SD)||57||±||9||46||±||10|
|Family history of CAD (n, %)||386|| ||18%||660|| ||22%|
|Current smoker (n, %)||458|| ||21%||770|| ||26%|
|Blood pressure systolic, mm Hg (mean ± 1 SD)||129||±||16||123||±||16|
|TPA mm2 (mean ± 1 SD)||52||±||50||36||±||50|
|Individuals with TPA ≥80 mm2 (n, %)||484|| ||22%||452|| ||15%|
|Total cholesterol, mmol/l (mean ± 1 SD)||5.9||±||1.2||5.9||±||1.2|
|HDL cholesterol, mmol/l (mean ± 1 SD)||1.5||±||0.5||1.4||±||0.4|
|LDL cholesterol, mmol/l (mean ± 1 SD)||3.7||±||1.0||3.8||±||0.9|
|Triglycerides, mmol/l (mean ± 1 SD) ||1.5||±||0.9||1.7||±||1.2|
|SCORE average risk in 10 years (% ± 1 SD)||1.5||±||1.7||0.6||±||0.3|
|PROCAM average risk in 10 years (% ± 1 SD)||6.2||±||7.4||4.3||±||3.5|
|AGLA average risk in 10 years (% ± 1 SD)||4.3|| ||5.2|| ||–|| |
|CAD = coronary artery disease; HDL = high-density lipoprotein; LDL = low density lipoprotein; SD = standard deviation; TPA = total plaque area|
For a 5-year statin treatment period, cost/QALY as calculated by the SMB is CHF 210 279 (table 2). These costs are applicable to 1000 statin-treated persons with a risk of 5% for cardiovascular events, corresponding to a SCORE risk of 0.91% according to the SMB assumption that for every two fatal events, nine nonfatal events occur (1:4.5 ratio). In these 1000 persons, 11 cardiovascular events would be prevented, based on a relative risk reduction of 22%. With a relative risk reduction of 30%, 15 cardiovascular events could have been prevented at reduced cost/QALY of CHF 144 380.
For a 10-year statin treatment period, and with treated baseline risks of SCORE 2.5, 5.0 and 7.5%, cost/QALY decreased to CHF 62 057, CHF 17 081 and CHF 2089, respectively, when a relative risk reduction of 22% was used in the calculation, and decreased to CHF 38 070, CHF 5088 and CHF –5906, respectively, when the evidence-based purely relative risk reduction of 30% was used. Therefore, cost/QALY is acceptable at a SCORE risk of 2.5%, and three times lower than the recommendation of the SMB (table 2).
|Table 2: Cost per quality adjusted life year (QALY) for 5 and 10 years and for different levels of coronary risk , with relative risk reductions of 30% and 22%.|
| ||Relative risk reduction 30%|| ||Relative risk reduction 22%|
|SMB cost/QALY model||5 years||10 years|| ||5 years||10 years|
|Fatal event||1.0||1.0||1.0||1.0|| ||1.0||1.0||1.0||1.0|
|Nonfatal event||4.5||4.5||4.5||4.5|| ||4.5||4.5||4.5||4.5|
|Cost of fatal event||8500||8500||8500||8500|| ||8500||8500||8500||8500|
|Cost of nonfatal event (1st year)||25 000||25 000||25 000||25 000|| ||25 000||25 000||25 000||25 000|
|Cost of nonfatal event (after 1st year)||8000||8000||8000||8000|| ||8000||8000||8000||8000|
|Statin and monitoring cost (per year)||470||470||470||470|| ||470||470||470||470|
|Effect [(improvement of life) * (quality)]||13.0||71.3||142.5||213.8|| ||9.5||52.3||104.5||156.8|
|Total cost (per 1000 individuals)||477 750||1 987 500||3 975 000||5 962 500|| ||350 350||1 457 500||2 915 000||4 372 500|
|Total cost (per individual)||478||1988||3975||5963|| ||350||1458||2915||4373|
|Statin and monitoring cost (observation years)||2350||4700||4700||4700|| ||2350||4700||4700||4700|
|Avoided health care costs||1872||2713||725||–1263|| ||2000||3243||1785||328|
|Cost:efficiency ratio (cost per QALY)||144 380||38 070||5088||–5906|| ||210 279||62 057||17 081||2089|
|Fatal risk in % in years||0.91||2.50||5.00||7.50|| ||0.91||2.50||5.00||7.50|
|Number of individuals||1000||1000||1000||1000|| ||1000||1000||1000||1000|
|Expected fatal events||9||25||50||75|| ||9||25||50||75|
|Expected nonfatal events||41||113||225||338|| ||41||113||225||338|
|Total number of events (fatal + nonfatal)||50||138||275||413|| ||50||138||275||413|
|Avoidable fatal events||2.7||8||15||23|| ||2.0||5.5||11.0||16.5|
|Avoidable nonfatal events||12.3||33.8||67.5||101.3|| ||9.0||24.8||49.5||74.3|
|Total number of avoidable events||15.0||41.3||82.5||123.8|| ||11.0||30.3||60.5||90.8|
|Absolute risk||5.0||14||28||41|| ||5.0||14||28||41|
|Avoidable risk||1.5||4||8||12|| ||1.1||3||6||9|
|Number needed to treat||67||24||12||8|| ||91||33||17||11|
Performance of various SCORE thresholds
to detect TPA80
The proportion of subjects with TPA80 increased with increasing levels of SCORE (table 3). The proportion of subjects with TPA80 was substantial in all age categories (11%, 25%, 45% in Switzerland and 13%, 41% in Germany). Of note, in the age group 40–65 years, most TPA80 were found in those with a SCORE risk below 2.5%. The rate of those with TPA80 (true positives) and those missed (false negatives) differed by SCORE thresholds of 2.5%, 5.0% and 7.5% in the age group 40–65 years: sensitivity was 28%, 9% and 2% in Switzerland and 28%, 5% and 1% in Germany (table 4).
|Table 3: Coronary risk categories for SCORE by age groups, by the distribution of subjects with total plaque area ≥80 mm2 (TPA80) and by country (Switzerland, n = 2124, or Germany, n = 2280).|
|CH-SCORE %||40–55 (n)||TPA80 (n)||TPA80 (%)||56–65 (n)||TPA80 (n)||TPA80 (%)||65–75 (n)||TPA80 (n)||TPA80 (%)|
|0.00–0.99||706||55|| 8||288|| 27|| 9|| 12|| 3||25|
|2.50–4.99|| 26||10||38||141|| 62||44||181|| 89||49|
|5.00–7.49|| 2|| 1||50|| 20|| 8||40|| 41|| 25||61|
|7.50+|| 0|| 0|| 0|| 7|| 5||71|| 23|| 16||70|
| || || || || || || || || || |
|DE-SCORE %||40–55 (n)||TPA80 (n)||TPA80 (%)||56–65 (n)||TPA80 (n)||TPA80 (%)|| || || |
|0.00–0.99||1397|| 99|| 7||137|| 31||23|| || || |
|1.00–2.49|| 292|| 91|| 31||266||104||39|| || || |
|2.50–4.99|| 37|| 28|| 76||123|| 73||59|| || || |
|5.00–7.49|| 3|| 3||100|| 18|| 14||78|| || || |
|7.50+|| 1|| 1|| 0|| 6|| 5|| 0|| || || |
|Total||1730||222|| 13||550||227||41|| || || |
|Table 4: Sensitivity of various SCORE thresholds to detect TPA80 in those aged 40–65 years, for Switzerland and Germany separately.|
|SCORE threshold CH||2.50%||5.00%||7.50%|
|True positives (n)||86||14||5|
|False negatives (n)||221||146||302|
|SCORE threshold DE||2.50%||5.00%||7.50%|
|True positives (n)||124||23||6|
|False negatives (n)||325||426||443|
First, we found good cost efficiency of CHF 38 000/QALY with statin treatment of those at a SCORE risk above 2.5%, if treated for a 10-year period and with a relative risk reduction of 30%. The relative risk reduction of 30% instead of 22% is based on the evidence of 22 trials in primary care comparing statins with placebo [8, 10]. The cost-efficient SCORE risk of 2.5% is three times lower than the SCORE threshold of 7.5% recommended by the SMB and corresponds to a cardiovascular risk of 14% when nonfatal events are added. This acceptable level of cost efficiency was of a similar order (CHF 62 057/QALY) when we used the relative risk reduction of 22% proposed by the SMB (table 2).
Because the number of QALYs increased by a factor four when patients are treated for 10 instead of 5 years, the cost efficiency is exponentially influenced by the duration of treatment. Risk calculations are typically based on a 10-year period and, therefore, it is clinically more convincing to derive cost efficiency from 10 instead of 5 treatment periods in chronic noncommunicable diseases.
Second, in our cross-sectional observation of otherwise healthy subjects with advanced carotid atherosclerosis from Switzerland and Germany, we showed that the statin treatment recommendation of the SMB would leave virtually all these patients untreated, because at the age of 40–65 years, the decision level of SCORE 7.5% to initiate statin therapy would treat only 5 of 302 subjects with advanced atherosclerosis in Switzerland (sensitivity 2%) and 6 of 443 subjects with advanced atherosclerosis in Germany (sensitivity 1%, table 4). We found that in subjects aged 40–65 years and with advanced carotid atherosclerosis, TPA80 was detected with a sensitivity of only 28% when a SCORE threshold of 2.5% was used (Germany: 28%). The sensitivity decreased to <10% for SCORE thresholds of 5.0% and 7.5% (table 4). By consequence, reservation of Statins to those with SCORE above 7.5% would lead to a rationing of a cost-effective intervention with Statins. Therefore, we expect the SMB recommendation to have disadvantage effects at the population level, where preventive therapy should start early in life [11–14].
Our results fit well into contemporary US guidelines, which recommend to use Statins (irrespective of the LDL level) based on the pooled risk equation (PCE) risk for major vascular events of greater than 7.5% and to consider Statins starting from a PCE risk of 5.0% . These numbers would correspond to an SCORE risk of 1.7% and 1.1%. The British Health Care System (NICE) recommends a similar threshold . Therefore, the SMB assumptions are not in line with these recommendations and raise the suspicion that the SMB is not sufficiently validated for correct cost/QALY calculations. Further, it does not incorporate indirect costs of an event, which further reduces its applicability.
In order to study the effect of the SMB threshold (SCORE risk 7.5% or more for an appropriate statin intervention in primary care), we used atherosclerosis imaging. Similar to coronary calcifications, the total carotid plaque burden contains important prognostic information and ability for reclassification . TPA80 is a high risk finding for incident myocardial infarction: 6,257 subjects with 894 incident cases of myocardial infarction were observed over a median follow-up time of 15.4 years, TPA of 40 ± 22 mm2 derived from the right carotid artery was associated with an unadjusted coronary risk of 23.9% (95%CI: 21.2–27.1) in 10 years. The hazard ratio per 1-SD increase in TPA (2.43 mm2) was 1.23 (95%CI: 1.15–1.32) using age as time scale and adjustments for sex, body mass index, smoking, total cholesterol, high-density lipoprotein cholesterol, diabetes mellitus, and hypertension .
By applying the SMB recommendation of SCORE 7.5% for an eventual statin eligibility and using a very conservative calculation for treatment effects of Statins, such medication would virtually be eliminated from primary care.
With regard to the cost-efficiency analysis, the results of greater than CHF 210 000 reported by the SMB can only be applied to a SCORE risk of less than 1% in 5 years, with substantial increases in cost efficiency from a SCORE risk of 2.5% in 10 years (CHF 38 070/QALY). The SMB recommendation is not derived from the cost model proposed and is very likely to pose a significant safety problem in subjects with low and intermediate cardiovascular risk, in whom 67% of first myocardial infarctions occur in Switzerland. Future guidelines should consider SCORE thresholds lower than 5% (e.g., 1.1–2.5%) for the initiation of statins in primary care for those aged 40–65 years.
No financial support and no other potential conflict of interest relevant to this article was reported.
Michel Romanens, MD
1 Felder S, Jüni P, Meier CA, et al. SMB Statin Recommendation. 2014. http://www.medical-board.ch/fileadmin/docs/public/mb/fachberichte/2014-07-21_bericht_statine_final_anpassung.pdf.
2 Navar-Boggan AM, Peterson ED, D’Agostino RB, Sr, Pencina MJ, Sniderman AD. Using age- and sex-specific risk thresholds to guide statin therapy: one size may not fit all. J Am Coll Cardiol. 2015;65(16):1633–9.
3 Thanassoulis G, Williams K, Altobelli KK, Pencina MJ, Cannon CP, Sniderman AD. Individualized Statin Benefit for Determining Statin Eligibility in the Primary Prevention of Cardiovascular Disease. Circulation. 2016;133(16):1574–81.
4 Mortensen MB, Nordestgaard BG, Afzal S, Falk E. ACC / AHA guidelines superior to ESC / EAS guidelines for primary prevention with statins in non-diabetic Europeans: the Copenhagen General Population Study. Eur Heart J. 2017;38:586–94. doi:10.1093/eurheartj/ehw426.
5 Selby K, Nanchen D, Auer R, Gencer B, Räber L, Klingenberg R, et al. Low statin use in adults hospitalized with acute coronary syndrome. Prev Med. 2015;77:131–6.
6 Hald EM, Lijfering WM, Mathiesen EB, Johnsen SH, Løchen ML, Njølstad I, et al. Carotid atherosclerosis predicts future myocardial infarction but not venous thromboembolism: the Tromso study. Arterioscler Thromb Vasc Biol. 2014;34(1):226–30. PubMed
7 Conroy RM, Pyörälä K, Fitzgerald AP, Sans S, Menotti A, De Backer G, et al.; SCORE project group. Estimation of ten-year risk of fatal cardiovascular disease in Europe: the SCORE project. Eur Heart J. 2003;24(11):987–1003. PubMed
8 Cholesterol Treatment Trialists’ (CTT) Collaborators, Mihaylova B, Emberson J, Blackwell L, Keech A, Simes J, Barnes EH, Voysey M, et al.
The effects of lowering LDL cholesterol with statin therapy in people at low risk of vascular disease: meta-analysis of individual data from
27 randomised trials. Lancet. 2012;380(9841):581–90.
9 Michel Romanens. CTT and QALY caclulator. 2017. www.docfind.ch/CTT2012.xlsx.
10 Cholesterol Treatment Trialists’ Ctt Collaborators. CTT Appendix Online 2012. Append online. https://researchonline.lshtm.ac.uk/1649027/1/mmc1.pdf.
11 McEvoy JW, Diamond GA, Detrano RC, Kaul S, Blaha MJ, Blumenthal RS, et al. Risk and the physics of clinical prediction. Am J Cardiol. 2014;113(8):1429–35. PubMed
12 Gottesman RF, Schneider ALC, Albert M, Alonso A, Bandeen-Roche K, Coker L, et al. Midlife hypertension and 20-year cognitive change: the atherosclerosis risk in communities neurocognitive study. JAMA Neurol. 2014;71(10):1218–27. PubMed
13 Navar-Boggan AM, Peterson ED, D’Agostino RB, Sr, Neely B, Sniderman AD, Pencina MJ. Hyperlipidemia in early adulthood increases long-term risk of coronary heart disease. Circulation. 2015;131(5):451–8. PubMed
14 Kokkinos PF, Faselis C, Myers J, Panagiotakos D, Doumas M. Interactive effects of fitness and statin treatment on mortality risk in veterans with dyslipidaemia: a cohort study. 2012;6736:10–15.
15 Goff DC, Jr, Lloyd-Jones DM, Bennett G, Coady S, D’Agostino RB, Gibbons R, et al.; American College of Cardiology/American Heart Association Task Force on Practice Guidelines. 2013 ACC/AHA guideline on the assessment of cardiovascular risk: a report of the American College of Cardiology/American Heart Association Task Force on Practice Guidelines. Circulation. 2014;129(25, Suppl 2):S49–73
16 Robson J. Lipid modification: cardiovascular risk assessment and the modification of blood lipids for the primary and secondary prevention of cardiovascular disease. Natl Inst Clin Excell. 2014. http://www.nice.org.uk/guidance/cg181/resources/guidance-lipid-modification-cardiovascular-risk-assessment-and-the-modification-of-blood-lipids-for-the-primary-and-secondary-prevention-of-cardiovascular-disease-pdf.
17 Baber U, Mehran R, Sartori S, Schoos MM, Sillesen H, Muntendam P, et al. Prevalence, impact, and predictive value of detecting subclinical coronary and carotid atherosclerosis in asymptomatic adults: the BioImage study. J Am Coll Cardiol. 2015;65(11):1065–74. PubMed