Risk prediction continue to ignore treatment effects
Risk prediction that ignores treatment effects is dangerous nonsense:
it may
mislead clinicians and patients. After a systematic review carried out in
2005
noted over-prediction in more recent cohorts, the NICE lipid guidelines
(1)
recommended that risk prediction studies should consider the impact of
treatment. However, Collins and Altman (2) correctly point out a dilemma:
a
treatment naïve cohort is impractical and unethical.
Recent cardiovascular risk scores have attempted to adjust for
treatment by a
few methods: (a) excluding those on treatment from the study e.g. QRISK2
excludes those on statins (3), (b) incorporating the use of
antihypertensives
as a risk predictor e.g. ARIC (4), QRISK (3, 5), and (c) adjusting for the
impact
of antihypertensive treatment on systolic blood pressure in the model e.g.
Framingham 2008 (6). However, these adjustments have focused on
antihypertensive treatment at baseline. No CVD risk scores have considered
the effects of other treatments such as platelet inhibitors or statin or,
more
importantly, treatment started after baseline examination. Hence CVD risk
scores will tend to underestimate CVD risk, and this underestimation is
greater in more recent cohorts (and more recent tools).
What can be done? As a start, risk prediction studies should collect
and
report data on participants on treatment at baseline and those started
during
the course of the study; this should not only include antihypertensives
but
also other disease modifying medications. Methods to adjust for the effect
of
treatment are poorly developed and need to be improved.
Researchers appear to have decided that the best way to deal with the
problem is to ignore the effect of treatment despite the findings of the
review
and the recommendations of the guidelines.
Reference:
1. Beswick AD, Brindle P, Fahey T, Ebrahim S. A systematic review of
risk
scoring methods and clinical decision aids used in the primary prevention
of
coronary heart disease. CG67 Lipid Modification: guideline appendix K.
London: National Collaborating Centre for Primary Care and Royal College
of
General Practitioners, 2008.
2. Collins GS, Altman DG. An independent and external validation of
QRISK2
cardiovascular disease risk score: a prospective open cohort study. BMJ
2010;
340: c2442.
3. Hippisley-Cox J, Coupland C, Vinogradova Y, Robson J, Minhas R,
Sheikh A,
et al. Predicting cardiovascular risk in England and Wales: prospective
derivation and validation of QRISK2. BMJ 2008;336:a332.
4. Chambless LE, Folsom AR, Sharrett AR, Sorlie P, Couper D, Szklo M,
Nieto
FJ. Coronary heart disease risk prediction in the Atherosclerosis Risk in
Communities (ARIC) study. J Clin Epidemiol 2003; 56(9): 880-890.
5. Hippisley-Cox J, Coupland C, Vinogradova Y, Robson J, May M,
Brindle P.
Derivation and validation of QRISK, a new cardiovascular disease risk
score
for the United Kingdom: prospective open cohort study. BMJ 2007;335:136.
6. D’Agostino RB, Sr, Vasan RS, Pencina MJ, Wolf PA, Cobain M,
Massaro JM,
Kannel WB. General cardiovascular risk profile for use in primary care:
the
Framingham Heart Study. Circulation, 2008; 117 (6):743-753.
Competing interests:
None declared
Competing interests:
No competing interests
03 June 2010
Su May Liew
DPhil Student (Oxford)/ Senior Lecturer (University of Malaya)
Paul Glasziou, Professor
Department of Primary Health Care, Oxford University OX3 7LF/ University of Malaya, 50603 Malaysia
Rapid Response:
Risk prediction continue to ignore treatment effects
Risk prediction that ignores treatment effects is dangerous nonsense:
it may
mislead clinicians and patients. After a systematic review carried out in
2005
noted over-prediction in more recent cohorts, the NICE lipid guidelines
(1)
recommended that risk prediction studies should consider the impact of
treatment. However, Collins and Altman (2) correctly point out a dilemma:
a
treatment naïve cohort is impractical and unethical.
Recent cardiovascular risk scores have attempted to adjust for
treatment by a
few methods: (a) excluding those on treatment from the study e.g. QRISK2
excludes those on statins (3), (b) incorporating the use of
antihypertensives
as a risk predictor e.g. ARIC (4), QRISK (3, 5), and (c) adjusting for the
impact
of antihypertensive treatment on systolic blood pressure in the model e.g.
Framingham 2008 (6). However, these adjustments have focused on
antihypertensive treatment at baseline. No CVD risk scores have considered
the effects of other treatments such as platelet inhibitors or statin or,
more
importantly, treatment started after baseline examination. Hence CVD risk
scores will tend to underestimate CVD risk, and this underestimation is
greater in more recent cohorts (and more recent tools).
What can be done? As a start, risk prediction studies should collect
and
report data on participants on treatment at baseline and those started
during
the course of the study; this should not only include antihypertensives
but
also other disease modifying medications. Methods to adjust for the effect
of
treatment are poorly developed and need to be improved.
Researchers appear to have decided that the best way to deal with the
problem is to ignore the effect of treatment despite the findings of the
review
and the recommendations of the guidelines.
Reference:
1. Beswick AD, Brindle P, Fahey T, Ebrahim S. A systematic review of
risk
scoring methods and clinical decision aids used in the primary prevention
of
coronary heart disease. CG67 Lipid Modification: guideline appendix K.
London: National Collaborating Centre for Primary Care and Royal College
of
General Practitioners, 2008.
2. Collins GS, Altman DG. An independent and external validation of
QRISK2
cardiovascular disease risk score: a prospective open cohort study. BMJ
2010;
340: c2442.
3. Hippisley-Cox J, Coupland C, Vinogradova Y, Robson J, Minhas R,
Sheikh A,
et al. Predicting cardiovascular risk in England and Wales: prospective
derivation and validation of QRISK2. BMJ 2008;336:a332.
4. Chambless LE, Folsom AR, Sharrett AR, Sorlie P, Couper D, Szklo M,
Nieto
FJ. Coronary heart disease risk prediction in the Atherosclerosis Risk in
Communities (ARIC) study. J Clin Epidemiol 2003; 56(9): 880-890.
5. Hippisley-Cox J, Coupland C, Vinogradova Y, Robson J, May M,
Brindle P.
Derivation and validation of QRISK, a new cardiovascular disease risk
score
for the United Kingdom: prospective open cohort study. BMJ 2007;335:136.
6. D’Agostino RB, Sr, Vasan RS, Pencina MJ, Wolf PA, Cobain M,
Massaro JM,
Kannel WB. General cardiovascular risk profile for use in primary care:
the
Framingham Heart Study. Circulation, 2008; 117 (6):743-753.
Competing interests:
None declared
Competing interests: No competing interests