Upload
doannhu
View
219
Download
0
Embed Size (px)
Citation preview
Medical Costs per QALY of Statins Using the Swiss Medical Board (SMB) assumptions: Observed Effects in Two Large Primary Prevention Cohorts from Germany and Switzerland
Michel Romanens1, Franz Ackermann1, Thomas Szucs2, Isabella Sudano3, Ansgar Adams4
1 Vascular Risk Foundation (Varifo), Olten 2 European Centre of Pharmaceutical Medicine (ECPM), Basel 3 University Heart Center Cardiology, University Hospital Zürich, Zürich,
Switzerland,4 Gesundheitsvorsorge und Sicherheitstechnik GmbH, Bonn, Germany
Medical Costs per QALY of Statins Using SMB assumptionsSGIM Annual Meeting 2015 (FM277): M. Romanens et al / www.varifo.ch
Medical Costs per QALY of Statins Using SMB assumptionsSGIM Annual Meeting 2015 (FM277): M. Romanens et al / www.varifo.ch
Background
- In June 2014 the SMB publishes a report showing costs per quality
adjusted life years (Costs/QALY) to be extremely unfavorable (210’000
SFr) for statins in primary care in subjects having an ESC risk of 0.9% in
5 years.
- As a consequence, in otherwise cardiovascular healthy subjects without
diabetes mellitus or familial hypercholesterolemia, the SMB recommends
statins in primary care only in subjects with a calculated ESC risk of
≥7.5% in 10 years.
- The rationale for this apparent gap warrants further elucidation.
- Further, costs/QALY at various cutoffs and it’s effect in cardiovascular
disease prevention at the primary care level have not yet been reported.
- The SMB Statin report is available at http://www.medical-
board.ch/fileadmin/docs/public/mb/Fachberichte/2014-07-
21_Bericht_Statine_Final_Anpassung.pdf
Medical Costs per QALY of Statins Using SMB assumptionsSGIM Annual Meeting 2015 (FM277): M. Romanens et al / www.varifo.ch
AIMS
- We aim to determine Costs/QALY for various cutoffs of ESC risk using
the SMB assumptions.
- We aim to determine, whether Costs/QALY at different ESC risk levels
correlate with the number needed to treat (NNT) in the SMB
assumptions.
- Based on various NNT, we aim to derive the ideal ESC risk cutoff for
costs/QALY, allowing an NNT of 25 to be adequate.
- We aim to test various ESC risk cutoffs in two large healthy populations
from Switzerland (CH) and Germany (DE) to detect a coronary risk
equivalents defined by the total carotid plaque burden, a highly sensitive
and specific marker of future fatal and nonfatal myocardial infarction.
Medical Costs per QALY of Statins Using SMB assumptionsSGIM Annual Meeting 2015 (FM277): M. Romanens et al / www.varifo.ch
METHODS (1)
- The EFFECT MODEL of the SMB to calculate costs/QALY is:
- for 1 fatal AMI, 4.5 nonfatal AMI occur
- relative risk reduction per 1 mmol/l LDL is 22%
- cost per fatal AMI is CHF 8'500, per nonfatal AMI is CHF 25'000 in
the first year and CHF 8'000 in subsequent years
- loss of QALY is 1.0 for fatal and 0.2 for nonfatal AMI
- annual preventive medical cost per individual including statin costs
CHF 470 SFr
- all AMI events occur uniformly after 50% of the total observation
time.
- The calculations are available at www.varifo.ch/QALYVarifo.xlsx
Medical Costs per QALY of Statins Using SMB assumptionsSGIM Annual Meeting 2015 (FM277): M. Romanens et al / www.varifo.ch
METHODS (2)
- We defined the numbers of subjects exhibiting a coronary risk equivalent
using carotid atherosclerotic plaque imaging and calculated the
sensitivity (SENS) and the specificity (SPEC) of various ESC cutoffs to
detect these subjects.
- SENS: sensitivity deals with the diseased subjects
- [true positives] / [true positives and false negatives]
- SPEC: specificity deals with the healthy subjects
- [true negatives] / [true negatives and false positives]
- NNT: 1 / absolute risk reduction
- Examples:
- 10% absolute risk reduction = 100/10 = NNT 10
- 1% absolute risk reduction = 100/1 = NNT 100
Medical Costs per QALY of Statins Using SMB assumptionsSGIM Annual Meeting 2015 (FM277): M. Romanens et al / www.varifo.ch
METHODS (3)
- Carotid imaging were obtained by ultrasound (linear probe 7-14 MHz)
- Imaging was performed with the identical imaging technique in CH and
DE
- Total Carotid Plaque Burden was determined from both carotid arteries
- Each carotid plaque was traced longitudinally and added together,
ending up with the total plaque area (TPA) in mm2.
- we used TPA ≥80 mm2 (TPA80) to define high long-term (≥20% in ≥10
years) coronary risk (Arterioscler Thromb Vasc Biol. 2014;34:226-230)
Medical Costs per QALY of Statins Using SMB assumptionsSGIM Annual Meeting 2015 (FM277): M. Romanens et al / www.varifo.ch
RESULTS (1)
- 1
Medical Costs per QALY of Statins Using SMB assumptionsSGIM Annual Meeting 2015 (FM277): M. Romanens et al / www.varifo.ch
RESULTS (1)
- 1
Medical Costs per QALY of Statins Using SMB assumptionsSGIM Annual Meeting 2015 (FM277): M. Romanens et al / www.varifo.ch
Results (2)
CH
2203
1083 (49%)
57±7
22%
60/79
30/93
11/98
4/99
DE
2942
989 (34%)
46±10
15%
30/95
7/100
2/100
0.5/100
Population (N)
Female (%)
Mean age (years±SD)
TPA ≥80 mm2
SENS/SPEC ESC 1.8% 10 years
SENS/SPEC ESC 3.3% 10 years
SENS/SPEC ESC 5.0% 10 years
SENS/SPEC ESC 7.5% 10 years
Sensitivity and specificity of ESC risk cutoffs the detect a
coronary risk equivalent defined by carotid plaque imaging
(TPA)
Medical Costs per QALY of Statins Using SMB assumptionsSGIM Annual Meeting 2015 (FM277): M. Romanens et al / www.varifo.ch
Discussion (1)
- Using the Costs/QALY assumptions of the SMB we find that the effect of
statins correlates with the NNT.
- Using a model with an ESC risk of 0.9% in 5 years, costs/QALY are
210’000 SFr. (NNT 91)
- Using a Model with an ESC risk of 7.5% in 10 years, costs/QALY are
2’089 SFr. (NNT 11)
- The middle way cutoff may be more suitable:
- Using a Model with an ESC risk of 3.3% in 10 years, costs/QALY are
40’261 SFr. (NNT 25)
NNT
Costs
/QA
LY
Medical Costs per QALY of Statins Using SMB assumptionsSGIM Annual Meeting 2015 (FM277): M. Romanens et al / www.varifo.ch
Discussion (2)
- The conclusion of the SMB has to be questioned, since there is a gap
between the presented costs/QALY of 210’000 SFr and the
recommended cutoff of ESC 7.5% to treat with statins.
- Using the SMB cutoff of ESC 7.5% would leave near 100% of healthy
subjects with a coronary risk equivalent untreated with statins.
- Using the SMB cutoff of ESC 7.5% creates an obsolescence for
measuring Cholesterol in primary care.
Medical Costs per QALY of Statins Using SMB assumptionsSGIM Annual Meeting 2015 (FM277): M. Romanens et al / www.varifo.ch
Conclusion
- We find statins to have costs per 1 mmol/l of LDL reduction of CHF
40'000/QALY (NNT 25) for an ESC risk of 3.3%, when we use the SMB
assumptions.
- With ESC 7.5% (SMB guide), many subjects with confirmed high risk
atherosclerosis would presumably not be treated (SENS 4% in CH, 0.5%
in DE), creating a situation, where there is an obsolescence for
Cholesterol measurements (because high LDL would not be treated
anyway).
- Further studies are needed to test the SMB statin effect assumptions,
since the SMB assumptions have not been evidenced in reality.
- In the future, costs per QALY should be calculated with inclusion of
medical and social costs.
- In the mean time, we strongly recommend to adhere to the international
guidelines for initiation of statin treatments in healthy subjects.