Elsevier

Atherosclerosis

Volume 181, Issue 1, July 2005, Pages 93-100
Atherosclerosis

Clinical Research
A comparison of the PROCAM and Framingham point-scoring systems for estimation of individual risk of coronary heart disease in the Second Northwick Park Heart Study

https://doi.org/10.1016/j.atherosclerosis.2004.12.026Get rights and content

Abstract

We have compared the predictive value of the PROCAM and Framingham risk algorithms in healthy UK men from the Second Northwick Park Heart Study (NPHS-II) (50–64 years at entry), followed for a median of 10.8 years for coronary heart disease (CHD) events. For PROCAM, the area under the receiver operating characteristic (ROC) curve was 0.63 (95% CI, 0.59–0.67), and not significantly different (p = 0.46) from the Framingham score, 0.62 (0.58–0.66). Sensitivities for a 5% false-positive rate (DR5) were 13.8 and 12.4%, respectively. Calibration analysis for PROCAM gave a ratio of observed to expected events of 0.46 (Hosmer–Lemeshow test, p < 0.0001) and 0.47 for Framingham (p < 0.0001). Using measures taken at 5 years of high-density lipoprotein cholesterol and (estimated) low-density lipoprotein cholesterol levels increased the ROC by only 1%. An NPHS-II risk algorithm, developed using a 50% random subset, and including age, triglyceride, total cholesterol, smoking status, and systolic blood pressure at recruitment, gave an ROC of 0.64 (0.58–0.70) with a DR5 of 10.7% when applied to the second half of the data. Adding family history and diabetes increased the DR5 to 18.4% (p = 0.28). Adding lipoprotein(a) >26.3 mg/dL (relative risk 1.6, 1.1–2.4) gave a DR5 of 15.5% (p = 0.55), while adding fibrinogen levels (relative risk for 1S.D. increase = 1.5, 1.1–2.0) had essentially no additional impact (DR5 = 16.9%, p = 0.95). Thus, the PROCAM algorithm is marginally better as a risk predictor in UK men than the Framingham score, but both significantly overestimate risk in UK men. The algorithm based on NPHS-II data performs similarly to those for PROCAM and Framingham with respect to discrimination, but gave an improved ratio of observed to expected events of 0.80 (p = 0.01), although no score had a high sensitivity. Any novel factors added to these algorithms will need to have a major impact on risk to increase sensitivity above that given by classical risk factors.

Introduction

Many individual characteristics contribute to the risk of clinical coronary heart disease (CHD) including gender, age, blood lipid concentrations, blood pressure, glucose tolerance, adiposity, and cigarette smoking. The complexity of the inter-relations between these risk factors makes assessment of individual ‘global’ risk difficult to evaluate in routine clinical practice, and statistical approaches have been developed, based on survival regression methods (e.g. Cox proportional hazards regression) or logistic regression. To simplify this approach for everyday use, point-scoring systems have been developed that permit the impact of several risk factors to be considered simultaneously [1], [2]. The population distribution of each risk factor is divided into several categories (e.g. cigarette smoker: yes/no; high-density lipoprotein cholesterol (HDLc): <35, 35–54, 55+ mg/dL), and each category is given a risk score. These scores are totalled and the result converted into 10-year risk of a coronary event from tables.

Point-scoring schemes have been developed from the Framingham study in the USA [1] and the PROCAM study in Germany [2]. The PROCAM system includes age, high-density lipoprotein cholesterol, low-density lipoprotein cholesterol (LDLc), triglyceride, smoking, diabetes, family history of CHD, and systolic blood pressure as risk factors. The Framingham system does not include information on family history, diabetes, triglyceride, or LDLc, but does include total cholesterol and interactions of age with smoking and cholesterol. Both systems use acute CHD events as the end point. Not surprisingly, the Framingham system was not as accurate as the PROCAM system when applied to the PROCAM data set, and the proposal has been made that valid comparison of performance requires their application to a third and independent data set [2]. The Second Northwick Park Heart Study (NPHS-II), a prospective cardiovascular study of healthy middle-aged men, has provided the opportunity both for development of a point-scoring system using conventional and novel coronary risk factors [3], [4], [5], and comparison with Framingham and PROCAM in a British setting.

Section snippets

Subjects

Full details of the recruitment methods, participant characteristics, and baseline measurements have been published previously [3], [4], [5]. Serum HDLc was measured using polyethylene glycol 8000 and enzymatic colorimetry on the sample of plasma taken at year 5 [5] and values used to estimate LDLc for each subject using the Friedewald equation as described [5]. Briefly, NPHS-II is a prospective study of 3052 healthy middle-aged Caucasian men (50–64 years) recruited from nine United Kingdom

Application of PROCAM and Framingham scores to NPHS-II

The PROCAM and Framingham scoring systems were applied to the 2732 NPHS-II men with complete data. Serum HDLc and LDLc were not measured at baseline in NPHS-II and so levels for these variables were set to the average observed in a subset of over 2000 NPHS-II men after 5 years of follow-up (LDLc 4.0 mmol/L and HDLc 0.8 mmol/L). The ability of the scores to separate men with and without disease was assessed using ROC curve analysis (Fig. 1). The ROC area using PROCAM was 0.63 (95% CI, 0.59–0.67).

PROCAM and Framingham comparison

Assessment of CHD risk is commonly used to identify patients who may benefit from primary prevention, and these assessments have frequently been based on equations derived from the Framingham study. In the present study, the risk scoring systems developed from both Framingham and PROCAM data have been applied to a UK-based sample of middle-aged men. Since this sample was comprised of European men, it might be expected that the PROCAM system (derived for subjects in Germany) would predict their

Acknowledgements

The following general practices collaborated in the study: The Surgery, Aston Clinton, Upper Gordon Road, Camberley; The Health Centre, Carnoustie; Whittington Moor Surgery, Chesterfield; The Market Place Surgery, Halesworth; The Health Centre, Harefield; Potterells Medical Centre, North Mymms; Rosemary Medical Centre, Parkstone, Poole; The Health Centre, St. Andrews. NPHS-II was supported by the UK Medical Research Council, the US National Institutes of Health (grant NHLBI 33014) and Du Pont

References (30)

  • M. Seed et al.

    Lipoprotein(a) as a predictor for myocardial infarction in middle-aged men

    Am J Med

    (2001)
  • V. Salomaa et al.

    Soluble thrombomodulin as a predictor of incident coronary heart disease and symptomless carotid artery atherosclerosis in the Atherosclerosis Risk in Communities (ARIC) study: a case-control study

    Lancet

    (1999)
  • Executive Summary of the Third Report of the National Cholesterol Education Program (NCEP) Expert Panel on Detection,...
  • G. Assman et al.

    Simple scoring scheme for calculating the risk of acute coronary events based on the 10-year follow-up of the prospective cardiovascular Munster (PROCAM) study

    Circulation

    (2002)
  • J.A. Cooper et al.

    Comparison of novel haemostatic factors and conventional risk factors for prediction of coronary heart disease

    Circulation

    (2000)
  • G.J. Miller et al.

    Increased activation of the haemostatic system in men at high risk for fatal coronary heart disease

    Thromb Haemost

    (1996)
  • P.J. Talmud et al.

    Non-fasting apolipoprotein B and triglyceride levels as a useful predictor of coronary heart disease risk in middle-aged UK men

    Arterioscler Thromb Vasc Biol

    (2002 (November 1))
  • World Health Organization Regional Office for Europe. Myocardial Infarction Community Registers. Copenhagen, Denmark:...
  • R.J. Prineas et al.

    The Minnesota Code Manual of Electrocardiographic Findings

    (1982)
  • E.R. Delong et al.

    Comparing the areas under two or more correlated receiver operating curves: a non-parametric approach

    Biometrics

    (1988)
  • B. Rosner et al.

    Correction of logistic regression relative risk estimates and confidence intervals for random within-person measurement error

    Am J Epidem

    (1992)
  • J.P. Empana et al.

    Are the Framingham and PROCAM coronary heart disease risk functions applicable to different European populations. The prime study

    Eur Heart J

    (2003)
  • J.W. Stephens et al.

    Cardiovascular risk and diabetes: are the methods of risk prediction satisfactory?

    Eur J Cardiovasc Prev Rehab

    (2004)
  • H-W. Hense et al.

    Framingham risk function overestimates risk of coronary heart disease in men and women from Germany: results from the MONICA Ausburg and the PROCAM cohorts

    Eur Heart J

    (2003)
  • A. Menotti et al.

    Comparison of the Framingham risk function-based coronary chart with risk function from an Italian poulation study

    Eur Heart J

    (2000)
  • Cited by (106)

    • Genetic risk analysis of coronary artery disease in Pakistani subjects using a genetic risk score of 21 variants

      2017, Atherosclerosis
      Citation Excerpt :

      These risk scores lack accuracy and may overestimate the risk in low risk subjects or underestimate the risk in subjects at high risk [6,14]. Almost 15–20% cases who developed CAD in their later lives were underestimated [15] with most of the cases occurring in intermediate and low risk subjects [16,17]. The variability in disease susceptibility in individuals exposed to similar environmental factors and having almost similar CRFs may be attributed to the genetic variations [18].

    • Genetics of atherosclerosis

      2017, Coronary Artery Disease: From Biology to Clinical Practice
    • MicroRNAs as regulators of endothelial cell functions in cardiometabolic diseases

      2016, Biochimica et Biophysica Acta - Molecular and Cell Biology of Lipids
    View all citing articles on Scopus
    View full text