From the Department of Preventive Medicine, Northwestern University
Medical School (P.G.), Chicago, Ill; Center for Human Nutrition, University of
Texas Southwestern Medical Center at Dallas (S.G.); Cardiovascular Division,
Departments of Medicine, Cardiac Unit, Massachusetts General Hospital
(R.C.P.), Boston, Mass; and National Heart, Lung, and Blood Institute (C.L.),
Bethesda, Md.
Correspondence to Philip Greenland, MD, Department of Preventive Medicine, Northwestern University Medical School, 680 N Lake Shore Dr, Suite 1102, Chicago, IL 60611. E-mail p-greenland{at}nwu.edu
Assessment of risk
and reduction of risk are well-accepted responsibilities of the
physician. The pathway from assessment of risk to reduction of risk
basically involves three steps: (1) measurement of risk factors and
collection of clinical data relevant to patient risk; (2)
interpretation of risk-related data with estimation of risk in absolute
terms (eg, risk of an event per year) as well as relative terms (ie,
low, intermediate, or high compared with others of the same age and
sex); and (3) on the basis of risk estimation results, intervention to
minimize disease risk or to prevent risk factor development in the
future. Although the process seems reasonably straightforward, problems
occur at each step that weaken the link between risk assessment and
risk reduction. Such problems occur in assessment of CVD risk
estimation and reduction just as in most other areas of medical
practice in spite of the availability of excellent data relating to CVD
risk estimation from the Framingham Heart Study and other similar data
sets.1
Periodic measurement of CVD risk factors in healthy people (step
1) is routinely recommended by the AHA, the ACC, and the NHLBI, in
addition to other authorities on disease
prevention.2 3 4 Blood lipid measurements, blood
pressure readings, age, sex, cigarette smoking and diabetic status, ECG
findings, and other risk predictors can be recorded or measured in
the office setting and can be entered into risk assessment
algorithms.1 5 Unfortunately, studies show that
physicians frequently fail to collect these simple and well-accepted
data elements in the course of usual medical
care.6 Even such basic clinical assessments as
the blood cholesterol level, blood pressure, and cigarette
smoking status are not routinely obtained,6 and
lack of such data obviously impedes risk assessment.
Interpretation of risk factor data (step 2) also poses problems. Over
the past three decades, cardiovascular risk profiles
and risk instruments have been available from long-term observational
studies.5 Such tools were expected to improve the
ease and quality of risk-related medical decisions by making risk
assessment more quantitative. Indeed, a 1976 Framingham Heart Study
publication7 stated: "The 10 percent of persons
identified with the use of this [risk assessment] function as at
highest risk accounted for about one fifth of the 8 year incidence of
coronary heart disease and about one third of the 8 year
incidence of atherothrombotic brain infarction, hypertensive heart
disease and intermittent claudication. Hence, the function provides an
economic and efficient method of identifying persons at high
cardiovascular risk who need preventive treatment and
persons at low risk who need not be alarmed about one moderately
elevated risk characteristic."
In this issue of Circulation, an updated version of the
Framingham Risk Prediction Score and a new clinically usable instrument
are published1 along with an extensive commentary
from the AHA Task Force on Risk Reduction.8 The
new Framingham Risk Score uses recently adopted clinical cut points
from NCEP2 and from the
JNC-VI.3 The commentary8
revisits the concept that risk scores such as those from Framingham
should prove useful to healthcare professionals engaged in risk factor
reductions for individual patients and might assist in selection of
specific therapies.
The goal of targeting risk-reducing treatments to persons most likely
to benefit is an appropriate one. Promotion of this goal was the focus
of the ACC's 27th Bethesda Conference: matching the intensity of risk
factor management with the hazard for coronary disease
events.9 An advanced risk prediction instrument
was prepared as a part of that report, but there is little evidence
that it has received attention outside of the conference. Despite the
availability of risk estimation tools such as those developed over many
years from Framingham, physicians cannot readily interpret the results
(step 2 of the process). Risk scores can be based on coronary
morbidity as well as mortality, and there is no general agreement which
to use or whether to use a combination of the two. Newer risk factors,
such as lipoprotein(a), homocysteine, LDL particle size, and thrombotic
markers, are not included in the available risk equations. It is
uncertain whether such new markers would improve risk assessment and
treatment selection. Risk scores, including the Framingham Score, may
or may not be accurate risk predictors in populations that differ from
those in which the original data were collected, eg, lower-risk
populations such as Asians or Latin Americans. Some risk assessment
tools, including the Framingham Risk Score, find no increase in
coronary heart disease risk for women with increasing age from
55 to 75 years, a finding that is not in accord with other
epidemiological data sets.10 11 Thus, risk
interpretation is also problematic.
The third step in risk reduction is intervention, and problems exist
here as well. Survey data show that physicians frequently fail to
achieve target levels for optimal risk reduction of
hypercholesterolemic or hypertensive
patients.12 13 Even when the risk factors have
been measured properly, patients have been diagnosed appropriately, and
interventions have been initiated, problems with patient compliance or
lack of appropriate follow-up can lead to inadequate risk factor
control. It is not surprising that multiple risk factors pose a
particular challenge in clinical care when single risk factors are
often not treated to target levels that are well accepted and widely
disseminated, such as those of the NCEP and JNC-VI.
Because they are aware of the many problems on the pathway from risk
assessment to risk reduction, the AHA, ACC, and NHLBI plan to undertake
a review of available cardiovascular risk assessment
methods to advise clinicians how to improve risk assessment and risk
reduction processes in clinical practice. Data such as those from
Framingham1 will be exceedingly important in
informing the deliberations of these three policy-forming groups. At
least three steps are necessary before any risk assessment tool can
achieve widespread use: (1) development of a tool that incorporates all
or most measures of risk that are already widely available; (2)
validation of the usefulness of such a tool in clinical practices; and
(3) discovery of ways to improve risk measurement and incorporate risk
assessment more readily into a busy clinician's daily routine. It is
hoped that risk assessment tools such as the Framingham Risk Score can
soon be made widely available, will be clinically confirmed as
accurate, will be found by clinicians to be acceptable and convenient
to use, and will ultimately improve the quality of patient care.
Selected Abbreviations and Acronyms
Footnotes
The opinions expressed in this editorial are not necessarily those of the editors or of the American Heart Association.
(Circulation. 1998;97:1761-1762.)
References
© 1998 American Heart Association, Inc.
Editorial
Problems on the Pathway From Risk Assessment to Risk Reduction
Key Words: Editorials • risk factors • coronary disease
ACC
=
American College of Cardiology
AHA
=
American Heart Association
CVD
=
cardiovascular disease
JNC-VI
=
Joint National Committee on Hypertension Detection, Treatment, and
Control
NCEP
=
National Cholesterol Education Program
NHLBI
=
National Heart, Lung, and Blood Institute
This article has been cited by other articles:
 |
|
 |
|
  R. C. Pasternak, J. Abrams, P. Greenland, L. A. Smaha, P. W. F. Wilson, and N. Houston-Miller Task force #1--identification of coronary heart disease risk: is there a detection gap? J. Am. Coll. Cardiol., June 4, 2003; 41(11): 1863 - 1874. [Full Text] [PDF]
|
|
|
|
 |
|
 P. Greenland, S. C. Smith Jr, and S. M. Grundy Improving Coronary Heart Disease Risk Assessment in Asymptomatic People: Role of Traditional Risk Factors and Noninvasive Cardiovascular Tests Circulation, October 9, 2001; 104(15): 1863 - 1867. [Full Text] [PDF]
|
|
|
|
 |
|
 W.B. Kannel Quest for an optimal population cardiovascular risk factor burden Eur. Heart J., January 2, 2001; 22(2): 105 - 107. [PDF]
|
|
|
|
|
|
A Rosengren, A Dotevall, H Eriksson, and L Wilhelmsen Optimal risk factors in the population: prognosis, prevalence, and secular trends. Data from Goteborg population studies Eur. Heart J., January 2, 2001; 22(2): 136 - 144. [Abstract] [PDF]
|
|
|
|
 |
|
 I.U. Haq, L.E. Ramsay, P.R. Jackson, and E.J. Wallis Prediction of coronary risk for primary prevention of coronary heart disease: a comparison of methods QJM, July 1, 1999; 92(7): 379 - 385. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. P. McCormack, M. Levine, and R. E. Rangno An Easier Approach to Estimating Risk of Coronary Heart Disease and Stroke Circulation, April 27, 1999; 99(16): 2219 - 2222. [Full Text] [PDF]
|
|
|
|
|
|
K. A. Eagle Estimation of Risk Reduction Circulation, April 13, 1999; 99 (14): 1922 - 1926. [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||