Home > Full Text Reviews > Clopidogrel and Modified-Release... > Model risk parameter values and sources

PubMed Health. A service of the National Library of Medicine, National Institutes of Health.

Greenhalgh J, Bagust A, Boland A, et al. Clopidogrel and Modified-Release Dipyridamole for the Prevention of Occlusive Vascular Events (Review of Technology Appraisal No. 90): A Systematic Review and Economic Analysis. Southampton (UK): NIHR Journals Library; 2011 Sep. (Health Technology Assessment, No. 15.31.)

Appendix 10Model risk parameter values and sources

For patients surviving an ischaemic stroke, four long-term treatment options are available to prevent future occlusive vascular events: low-dose ASA, clopidogrel, MRD and ASA + MRD. For the other three patient groups (MI only, peripheral arterial disease only and multivascular disease) only ASA and clopidogrel are licensed for secondary prevention. In all cases it is also necessary to consider periods when no active long-term drug treatment is being taken to reduce the risk of occlusive vascular event.

NICE Clinical Guidance CG48:28 post myocardial infarction clopidogrel

For patients suffering a new MI, recommendations were made in CG4828 for the short-term use of clopidogrel + ASA to prevent early vascular events (primarily repeat MIs):

  • For patients experiencing a NSTEMI, clopidogrel + ASA is recommended for 12 months.
  • For patients experiencing a STEMI, clopidogrel + ASA is recommended for 4 weeks (30 days).

The CURE27 trial provides the evidence source for the first recommendation. This showed a significant protective effect in relation to repeat MIs, but not for strokes. The absolute risk reduction over 12 months was 1.47% (standard error 0.42%).

The recommendation for STEMI patients derives primarily from the COMMIT29 trial where a modest reduction was seen in the rate of re-infarctions, but not in strokes. During the 30-day follow-up, an absolute risk reduction of 0.33% was reported (standard error 0.14%).

To accommodate the likely impact of these guidelines a weighted-average effect has been estimated of 0.853% (standard error 0.207%), based on the balance of STEMI and NSTEMI patients in the GRACE118 study (54.2% and 45.8%, respectively). This reduction is applied to the transient effect risk parameter values shown below for a second MI event after surviving a non-fatal MI, but not to any other MI risks that are much smaller and where no transient effect was identified.

Risks of first occlusive vascular event

Haemorrhagic stroke as first event

The annual risks of suffering an haemorrhagic stroke are generally very low, but vary significantly between patient types and between different treatment options. Reviewing all of the data available, it appears that this risk is effectively constant over quite long periods of time. Evidence in some cases of a small additional early risk is not confirmed from other sources, and may in part be a consequence of differing qualifying criteria among trials, so that some early acute events (in hospital or in the immediate post-discharge period) are counted within some trials, but excluded in others. In estimating model parameters, such transient effects are ignored and only the longer-term annual event rate is used.

For ASA and clopidogrel treatments, risks are estimated from the CAPRIE26 trial; in the ischaemic stroke-only population, sufficient haemorrhagic stroke events were recorded to allow separate parameter values to be obtained, but for the other groups it was possible to derive only a single risk estimate for the population regardless of the treatment in use.

Haemorrhagic stroke risk for MRD + ASA treatment was estimated from the PRoFESS57 trial (noting that the clopidogrel arm in PRoFESS57 yielded a similar event rate to that in CAPRIE26). The risk appropriate for untreated patients was based on the ASA estimated RR for ‘no treatment’ vs ASA in an ATTC66 analysis of secondary prevention published in 2002: RR 1.22 (95% CI 1.03 to 1.44). Finally, the annual risk of haemorrhagic stroke when using MRD was set at the same level as ‘no treatment’, based on the finding of very similar risks reported from the ESPS-230 trial.

TABLE 74Model parameter estimates for the risk of haemorrhagic stroke as first event

PopulationDetailASACLOPASA + MRDMRDNo treatment
IS onlyAnnual risk (%)0.49000.26100.43200.40200.4020
Standard error (%)0.02200.01700.01200.03800.0380
SourceCAPRIE26CAPRIE26PRoFESS57CAPRIE26/ATTC66
MI onlyAnnual risk (%)0.09560.0956N/AN/A0.0784
Standard error (%)0.00030.0003N/AN/A0.0069
SourceCAPRIE26N/AN/ACAPRIE26/ATTC66
PAD onlyAnnual risk (%)0.09100.0910N/AN/A0.0746
Standard error (%)0.01170.0117N/AN/A0.0114
SourceCAPRIE26CAPRIE26N/AN/ACAPRIE26/ATTC66
MVDAnnual risk (%)0.19600.1960N/AN/A0.1602
Standard error (%)0.01200.0120N/AN/A0.0170
SourceCAPRIE26CAPRIE26N/AN/ACAPRIE26/ATTC66

CLOP, clopidogrel; IS, ischaemic stroke; MVD, multivascular disease; N/A, not available; PAD, peripheral arterial disease.

Ischaemic stroke as first event

The risk of suffering a recurrent ischaemic stroke is relatively high for patients in the ‘ischaemic stroke-only’ and multivascular disease populations. In addition to a long-term steady risk level, an important transient increased risk is also present within the trial data, which applies for slightly different periods for each population.

For the ‘ischaemic stroke-only’ population model, parameter values have been estimated from CAPRIE26 for ASA and clopidogrel, and from a comparison of PRoFESS57 and CAPRIE26 for ASA + MRD. The ‘no-treatment’ risk was based on the ATTC66 RR for ASA versus ‘no treatment’ applicable to ischaemic stroke. Finally, the annual risk of ischaemic stroke when using MRD was based on the MRD + ASA estimate adjusted by the RRR (24.7%) compared with MRD reported in the ESPS-230 trial. No consistent differences were observed in any of the trials relating to gender.

In the ‘MI-only’ population, no consistent differences were found in the CAPRIE26 data for the choice of treatment (ASA vs clopidogrel), but long-term risks were much higher for females than for males. Therefore, parameters were estimated for two models (males and females separately), combining patients in the two trial arms.

TABLE 75Model parameter estimates for risk of ischaemic stroke as first event in the ‘ischaemic stroke-only’ population

PopulationDetailASACLOPASA + MRDMRDNo treatment
IS onlyLong-term annual risk (%)4.2013.9713.9715.2736.001
Standard error (%)0.0270.0270.0270.4840.247
Transient risk (%) (%)1.9621.7231.7232.2882.802
Standard error (%)0.0440.0470.0470.2290.127
Duration of transient risk (months)2.83.13.13.12.8
SourceCAPRIE26CAPRIE26PRoFESS57/CAPRIE26ProFESS57/CAPRIE26/ESPS–230CAPRIE26/ATTC66

TABLE 76Model parameter estimates for risk of ischaemic stroke as first event in the ‘MI-only’, ‘peripheral arterial disease-only’ and multivascular disease populations

PopulationDetailASACLOPNo treatment
MI only (females)Long-term annual risk (%)0.7740.7741.106
Standard error (%)0.0410.0410.074
Transient risk (%)0.3140.3140.449
Standard error (%)0.0550.0550.077
Duration of transient risk (months)0.30.30.3
SourceCAPRIE26CAPRIE26CAPRIE26/ATTC66
MI only (males)Long-term annual risk (%)0.3000.3000.429
Standard error (%)0.0250.0250.038
Transient risk (%)0.3230.3230.462
Standard error (%)0.0440.0440.065
Duration of transient risk (months)3.73.73.7
SourceCAPRIE26CAPRIE26CAPRIE26/ATTC66
PAD onlyLong-term annual risk (%)1.1451.1451.636
Standard error (%)0.0120.0120.067
Transient risk (%)−0.099−0.099−0.141
Standard error (%)0.0160.0160.023
Duration of transient risk (months)0.60.60.6
SourceCAPRIE26CAPRIE26CAPRIE26/ATTC66
MVD (females)Long-term annual risk (%)4.3163.8796.166
Standard error (%)0.0700.0860.272
Transient risk (%)0.4130.2650.591
Standard error (%)0.0970.1150.144
Duration of transient risk (months)0.030.50.03
SourceCAPRIE26CAPRIE26CAPRIE26/ATTC66
MVD (males)Long-term annual risk (%)3.3762.9034.823
Standard error (%)0.0300.0290.192
Transient risk (%)0.8080.6271.154
Standard error (%)0.0440.0440.079
Duration of transient risk (months)1.31.61.3
SourceCAPRIE26CAPRIE26CAPRIE26/ATTC66

CLOP, clopidogrel; MVD, multivascular disease; PAD, peripheral arterial disease.

In the ‘peripheral arterial disease-only’ population, there was no evidence of differences by either gender or treatment, so a single model was calibrated covering all CAPRIE26 trial patients.

In the multivascular disease population, there was equivocal evidence in CAPRIE26 suggesting that females are at greater risk than males, and that ASA may be less effective than clopidogrel at preventing recurrent ischaemic stroke; however, the differences appeared to be quite small. In this case, four separate models were calibrated to ensure that even small differences would be reflected in the economic results.

In all cases, risks for patients not receiving any prophylaxis were estimated by adjusting the ASA rates using the RR from the ATTC66 meta-analysis.

Myocardial infarction as first event

The risk of suffering a MI is relatively high for patients in the ‘MI-only’ and multivascular disease populations. In addition to a long-term steady risk level, an important transient increased risk is also present in some cases within the trial data, which applies for different periods for each population.

For the ‘ischaemic stroke-only’ population model, parameter values have been estimated from CAPRIE26 for ASA and clopidogrel where no difference was observed within the trial. A comparison of PRoFESS57 and CAPRIE26 allowed estimation of the long-term risk when receiving treatment with MRD + ASA. The ‘no-treatment’ risk was based on the ATTC66 RR for ASA vs ‘no treatment’ applicable to MI. Finally, the annual risk of MI when using MRD is assumed to be equal to that of ‘no treatment’ based on comparable event rates reported in the ESPS-230 trial. No consistent differences were observed in any of the trials relating to gender.

TABLE 77Model parameter estimates for risk of MI as first event in the ‘ischaemic stroke-only’ population

PopulationDetailASACLOPASA + MRDMRDNo treatment
IS onlyLong-term annual risk (%)0.4920.4920.3630.6560.656
Standard error (%)0.0060.0060.0060.0190.019
Transient risk (%)N/AN/AN/AN/AN/A
Standard error
Duration of transient risk (months)
SourceCAPRIE26CAPRIE26PRoFESS57/CAPRIE26CAPRIE26/ESPS–230CAPRIE26/ATTC66

CLOP, clopidogrel; IS, ischaemic stroke; N/A, not available.

In the ‘MI-only’ and ‘peripheral arterial disease-only’ populations, separate estimates of risk were obtained from the CAPRIE data for treatment with ASA and clopidogrel. No differences were apparent between male and female patients.

For the multivascular disease population, there was some evidence in the CAPRIE26 data supporting risk differences by both gender and treatment. Four separate models were calibrated to ensure that even small differences would be reflected in the economic results. Transient risks were only evident for ASA treatment.

In all cases, risks for patients not receiving any prophylaxis were estimated by adjusting the ASA rates using the RR from the ATTC66 meta-analysis.

TABLE 78Model parameter estimates for risk of MI as first event in the ‘MI-only’, ‘peripheral arterial disease-only’ and multivascular disease populations

PopulationDetailASACLOPNo treatment
MI onlyLong-term annual risk (%)2.0391.6292.719
Standard error (%)0.0190.0190.076
Transient risk (%)1.4771.5891.969
Standard error (%)0.0290.0290.065
Duration of transient risk (months)2.22.52.2
SourceCAPRIE26CAPRIE26CAPRIE26/ATTC66
PAD onlyLong-term annual risk (%)0.9640.9531.285
Standard error (%)0.0310.0300.055
Transient risk (%)0.181−0.3980.241
Standard error (%)0.0430.0450.058
Duration of transient risk (months)6.62.66.6
SourceCAPRIE26CAPRIE26CAPRIE26/ATTC66
MVD (females)Long-term annual risk (%)2.3861.4973.182
Standard error (%)0.0710.0720.127
Transient risk (%)0.464N/A0.619
Standard error (%)0.102N/A0.141
Duration of transient risk (months)0.7N/A0.7
SourceCAPRIE26CAPRIE26CAPRIE26/ATTC66
MVD (males)Long-term annual risk (%)2.7942.4863.726
Standard error (%)0.0250.0180.105
Transient risk (%)0.713N/A0.951
Standard error (%)0.037N/A0.054
Duration of transient risk (months)1.9N/A1.9
SourceCAPRIE26CAPRIE26CAPRIE26/ATTC66

CLOP, clopidogrel; MVD, multivascular disease; PAD, peripheral arterial disease.

Other vascular death as first event

The incidence of other vascular death as a first event in the ‘ischaemic stroke-only’ population was estimated directly jointly from the CAPRIE26 trial data for ASA and clopidogrel treatments, where no meaningful differences were observed related to either choice of treatment or to gender. Analysis of the PRoFESS57 trial results similarly show no differences between clopidogrel and MRD + ASA. Occlusive vascular disease was not reported in other trials, but the ESPS-230 report allowed calculation of total deaths excluding fatal strokes and this was considered a reasonable proxy for other vascular death, allowing RR multipliers to be calculated for MRD and ‘no treatment’ compared with ASA + MRD.

TABLE 79Model parameter estimates for risk of other vascular death as first event in the ‘ischaemic stroke-only’ population

PopulationDetailASACLOPASA + MRDMRDNo treatment
IS onlyLong-term annual risk (%)1.0501.0501.0501.0251.156
Standard error0.0260.0260.0260.1000.094
Transient risk (%)−0.457−0.457−0.457−0.446−0.503
Standard error0.0490.0490.0490.0640.067
Duration of transient risk (months)6.96.96.96.96.9
SourceCAPRIE26CAPRIE26PRoFESS57/CAPRIE26CAPRIE26/ESPS–230CAPRIE26/ESPS–230

In the ‘MI-only’ population, separate estimates of risk were obtained from the CAPRIE26 data for treatment with ASA and clopidogrel, and for both genders.

In the ‘peripheral arterial disease-only’ population, no differences were observed by gender, so combined estimates were obtained for ASA and clopidogrel after combining results for male and female patients.

For the multivascular disease population, there was clear evidence in the CAPRIE26 data supporting risk differences by gender, but not by treatment. Therefore, two models were calibrated for male and female patients.

In all cases, risks for patients not receiving any prophylaxis were estimated by adjusting the ASA rates using the RR from ESPS-230 trial as described above.

TABLE 80Model parameter estimates for risk of other vascular death as a first event in the ‘MI-only’, ‘peripheral arterial disease-only’ and multivascular disease populations

PopulationDetailASACLOPNo treatment
MI only (females)Long-term annual risk (%)0.8631.4440.951
Standard error0.1370.2340.167
Transient risk (%)0.7090.6580.780
Standard error0.1190.1180.139
Duration of transient risk (months)0.81.40.8
SourceCAPRIE26CAPRIE26CAPRIE26/ESPS–230
MI only (males)Long-term annual risk (%)0.6461.0800.711
Standard error0.0190.0390.060
Transient risk (%)0.5300.4920.583
Standard error (%)0.0250.0480.054
Duration of transient risk (months)0.81.40.8
SourceCAPRIE26CAPRIE26CAPRIE26/ESPS–226
PAD onlyLong-term annual risk (%)1.4990.5831.650
Standard error (%)0.3920.0590.447
Transient risk (%)−1.226−0.161−1.351
Standard error (%)1.5610.1111.751
Duration of transient risk (months)16.63.416.6
SourceCAPRIE26CAPRIE26CAPRIE26/ESPS–226
MVD (females)Long-term annual risk (%)1.4271.4271.571
Standard error (%)0.0640.0640.144
Transient risk (%)0.7010.7010.772
Standard error (%)0.1090.1090.137
Duration of transient risk (months)2.32.32.3
SourceCAPRIE26CAPRIE26CAPRIE26/ESPS–226
MVD (males)Long-term annual risk (%)2.6532.6532.922
Standard error (%)0.0160.0160.232
Transient risk (%)−0.230−0.230−0.254
Standard error (%)0.0270.0270.035
Duration of transient risk (months)2.42.42.4
SourceCAPRIE26CAPRIE26CAPRIE26/ESPS–226

CLOP, clopidogrel; MVD, multivascular disease; PAD, peripheral arterial disease.

Risks of subsequent occlusive vascular events

For patients surviving a first occlusive vascular event within the key trials (CAPRIE26 and PRoFESS57), the number of patients suffering a second or third event are very small. In a few cases it is feasible to estimate parameter values relating to specific second events, but in many cases the data are insufficient so it has been necessary to make assumptions based on the available evidence.

Following non-fatal ischaemic stroke as first event: risk of second ischaemic stroke event

A number of patients who survived an ischaemic stroke in the CAPRIE26 trial went on to experience a second ischaemic stroke event. No significant differences in incidence rates were apparent relating to the choice of treatment. However, those belonging to the ‘ischaemic stroke-only’ population experienced a lower level of risk than other patients. The same approach to extending these parameters to cover other treatments used as for ischaemic stroke first events.

TABLE 81Model parameter estimates for risk of ischaemic stroke as a second event following non-fatal ischaemic stroke as a first event

PopulationDetailASA, CLOP ASA + MRDaMRDNo treatment
IS onlyLong-term annual risk (%)7.3239.72510.462
Standard error (%)0.6941.2771.069
Transient risk (%)7.0399.34910.056
Standard error (%)1.4012.0691.997
Duration of transient risk (months)6.26.26.2
SourcePRoFESS57/CAPRIE26ProFESS26/CAPRIE26/ESPS–226CAPRIE26/ATTC26
MI only, PAD only and MVDLong-term annual risk (%)11.627N/A16.610
Standard error (%)0.2010.714
Transient risk (%)3.3354.764
Standard error (%)0.2240.365
Duration of transient risk (months)1.41.4
SourceCAPRIE26CAPRIE26/ATTC66

CLOP, clopidogrel; IS, ischaemic stroke; MVD, multivascular disease; N/A, not applicable; PAD, peripheral arterial disease.

a

Not applicable to populations other than ‘ischaemic stroke-only’.

Following non-fatal ischaemic stroke as first event: risk of myocardial infarction event

Very few ischaemic stroke survivors suffered a subsequent MI in the CAPRIE26 trial. A single overall linear regression hazard model was calibrated for all patient groups, extended additional treatments as before for first MI events.

TABLE 82Model parameter estimates for risk of MI as a second event following non-fatal ischaemic stroke as a first event

PopulationDetailASA, CLOPASA + MRDMRD, no treatment
All patientsLong-term annual risk (%)1.2120.8921.616
Standard error (%)0.1810.2200.243
Transient risk (%)N/AN/AN/A
Standard error (%)N/AN/AN/A
Duration of transient risk (months)N/AN/AN/A
SourceCAPRIE26PRoFESS/CAPRIE26CAPRIE26/ESPS–230/ATTC66

CLOP, clopidogrel; N/A, not available.

Following non-fatal ischaemic stroke as first event: risk of other vascular death event

Very few patients who survived an ischaemic stroke in the CAPRIE26 trial suffered a subsequent other vascular death event. A single projection model was calibrated for all patient groups, extended additional treatments as before for primary other vascular death events.

TABLE 83Model parameter estimates for risk of other vascular death as a second event following non-fatal ischaemic stroke as a first event

PopulationDetailASA, CLOP, ASA + MRDMRDNo treatment
All patientsLong-term annual risk (%)1.8531.8092.041
Standard error (%)0.1420.2180.232
Transient risk (%)2.3542.2972.592
Standard error (%)0.2110.3000.310
Duration of transient risk (months)2.02.02.0
SourceCAPRIE26PRoFESS57/CAPRIE26CAPRIE26/ESPS–230/ATTC66

Following non-fatal ischaemic stroke as first event: risk of haemorrhagic stroke event

Insufficient haemorrhagic stroke events occurred among ischaemic stroke survivors to allow any subdivision by patient subgroups or treatments.

TABLE 84Model parameter estimates for risk of haemorrhagic stroke as a second event following non-fatal ischaemic stroke as a first event

PopulationDetailAll treatmentsNo treatment
All patientsLong-term annual risk (%)1.0540.864
Standard error (%)0.0900.108
Transient risk (%)0.2500.205
Standard error (%)0.0590.049
Duration of transient risk (months)0.10.1
SourceCAPRIE26CAPRIE26/ATTC66

Following non-fatal myocardial infarction as first event: risk of myocardial infarction event

No differences in MI risk were detectable by treatment in the CAPRIE26 trial data, but the risk among the multivascular disease population was more than double the risk in the other groups.

TABLE 85Model parameter estimates for risk of MI as a second event following non-fatal MI as a first event

PopulationDetailASA, CLOPASA + MRDMRDNo treatment
IS only, MI only and PAD onlyLong-term annual risk (%)5.7874.2617.7167.716
Standard error (%)0.1900.8170.3270.327
Transient risk (%)a3.2873.0984.3834.383
Standard error (%)0.2390.6050.3400.340
Duration of transient risk (months)1.61.61.61.6
SourceCAPRIE26PRoFESS57/CAPRIE26CAPRIE26/ESPS–230CAPRIE26/ATTC66
MVDLong-term annual risk (%)12.228N/AN/A16.303
Standard error (%)0.513N/AN/A0.819
Transient risk (%)a8.713N/AN/A11.617
Standard error (%)0.462N/AN/A0.734
Duration of transient risk (months)0.8N/AN/A0.8
SourceCAPRIE26CAPRIE26/ATTC66

CLOP, clopidogrel; IS, ischaemic stroke; MVD, multivascular disease; N/A, not available; PAD, peripheral arterial disease.

a

These transient risks are further reduced by 0.853% for the short-term impact of the CG48 guidance,28 as described above.

Following non-fatal myocardial infarction: risk of ischaemic stroke event

The risk of suffering an ischaemic stroke event following a non-fatal MI was found to be very low and a single projective model was calibrated using all available CAPRIE26 data.

TABLE 86Model parameter estimates for risk of ischaemic stroke as a second event following non-fatal MI as a first event

PopulationDetailASA, CLOPASA + MRDMRDNo treatment
All patientsLong-term annual risk (%)1.8371.8372.4402.624
Standard error (%)0.2670.2670.4170.394
Transient risk (%)1.6081.6082.1352.297
Standard error (%)0.3070.3070.4520.431
Duration of transient risk (months)2.22.22.22.2
SourceCAPRIE26PRoFESS57/CAPRIE26CAPRIE26/ESPS–230CAPRIE26/ATTC66

Following non-fatal myocardial infarction: risk of other vascular death event

Although it was not possible to detect any difference in risk by treatment type in the CAPRIE26 data, it was clear that patients with multivascular disease suffered a threefold risk of other vascular death cause following a non-fatal MI compared with other groups.

TABLE 87Model parameter estimates for risk of other vascular death as a second event following non-fatal MI as a first event

PopulationDetailASA, CLOPASA + MRDMRDNo treatment
MI only, IS only and PAD onlyLong-term annual risk (%)3.1103.1103.0353.425
Standard error (%)0.1520.1520.3180.317
Transient risk (%)N/AN/AN/AN/A
Standard error (%)
Duration of transient risk (months)
SourceCAPRIE26PRoFESS57/CAPRIE26CAPRIE26/ESPS–230CAPRIE26/ATTC66
MVDLong-term annual risk (%)10.850N/AN/A11.949
Standard error (%)0.3041.000
Transient risk (%)N/AN/A
Standard error (%)
Duration of transient risk (months)
SourceCAPRIE26CAPRIE26/ATTC66

CLOP, clopidogrel; IS, ischaemic stroke; MVD, multivascular disease; N/A, not available; PAD, peripheral arterial disease.

Following non-fatal myocardial infarction: risk of a haemorrhagic stroke event

The risk of haemorrhagic stroke following an initial MI event was found to be extremely low.

TABLE 88Model parameter estimates for risk of haemorrhagic stroke as a second event following non-fatal MI as a first event

PopulationDetailAll treatmentsNo treatment
All patientsLong-term annual risk (%)0.1900.156
95% confidence limits (LCL, UCL)0.005% to 0.699%0.006% to 0.853%
Transient risk (%)N/AN/A
Standard error (%)N/AN/A
Duration of transient risk (months)N/AN/A
SourceCAPRIE26CAPRIE26/ATTC66

LC, lower confidence limit; N/A, not available; UCL, upper confidence limit.

Following non-fatal haemhorrhagic stroke as a first event

There were too few events of any type recorded in the CAPRIE26 trial to patients surviving an initial haemorrhagic stroke. However, in order to provide parameters for this part of the model, a simple device was employed: the overall event rate was subdivided among the possible four types of event (ischaemic stroke, haemorrhagic stroke, MI and other vascular death) in proportion to their frequency among CAPRIE26 first events, and the figure converted to a single average event rate for each event.

TABLE 89Model parameter estimates for risk of second events following haemorrhagic stroke as first event

PopulationEventDetailAll treatmentsNo treatment
All patientsISLong-term annual risk (%)2.8754.107
Standard error (%)0.4890.726
Haemorrhagic strokeLong-term annual risk (%)1.9441.594
Standard error (%)0.3310.298
MILong-term annual risk (%)0.1820.243
Standard error (%)0.0310.042
Other vascular deathLong-term annual risk (%)1.4391.585
Standard error (%)0.2450.311
SourceCAPRIE26CAPRIE26/ATTC66

Risk modifiers

Cox's proportional hazard regressions were carried out on the CAPRIE26 data to identify the influence of age and stroke-related disability (using the modified Rankin Score) on the key first events in the trial. From these results event modifying factors were derived to allow the risk values described above to be adjusted to the characteristics of individual patients.

TABLE 90Risk modifiers for age and stroke-related disability

EventAge modifier (per year)Stroke disability (modified Rankin Score)
Not disabled (0–2)Disabled (3+)
IS1.0200.9451.201
Haemorrhagic stroke1.0100.8551.653
MI1.0410.9811.064
Other vascular death1.0430.7742.283
Non-vascular death1.0730.8621.614
© 2011, Crown Copyright.

Included under terms of UK Non-commercial Government License.

Cover of Clopidogrel and Modified-Release Dipyridamole for the Prevention of Occlusive Vascular Events (Review of Technology Appraisal No. 90): A Systematic Review and Economic Analysis
Clopidogrel and Modified-Release Dipyridamole for the Prevention of Occlusive Vascular Events (Review of Technology Appraisal No. 90): A Systematic Review and Economic Analysis.
Health Technology Assessment, No. 15.31.
Greenhalgh J, Bagust A, Boland A, et al.
Southampton (UK): NIHR Journals Library; 2011 Sep.

Download

PubMed Health Blog...

read all...

Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...