INTRAINDIVIDUAL VARIABILITY OF FIBRINOGEN LEVELS AND CARDIOVASCULAR RISK PROFILE

Prospective population studies have established that fibrinogen is an independent predictor for ischemic heart disease and stroke. These stu dy conclusions have prompted recommendations that fibrinogen determina tions be included in the cardiovascular risk profile. The routine avai lability of fibrinogen measurements may result in widespread screening prior to establishing the validity of a single fibrinogen level as an accurate descriptor for individual subjects. The objectives of this s tudy were to describe the methodological and intraindividual component s of variability in fibrinogen measurements determined by using the Cl auss method; to establish the usefulness of a single fibrinogen measur ement on risk stratification and retest reproducibility; and to determ ine the influence of intraindividual fibrinogen variability on sample size estimates. Fibrinogen levels were measured by a modification of t he Clauss method. Three cohorts of apparently healthy, nonsmoking volu nteers were recruited. The single-day intraindividual component of fib rinogen variability was determined in 39 subjects. For the 5-day intra individual component of fibrinogen variability, 32 subjects were recru ited, and in the 6-week intraindividual study, 28 subjects were includ ed. The coefficient of variation for the methodological component of f ibrinogen variability was 5.8% as determined from batch analyses, but the intraindividual coefficient of variation for replicate measures on a single day was 10.7%. The 5-day intraindividual coefficient of vari ation was 14.2%, and for the 6-week period it was 17.8%. Based on the 6-week data, an average of four fibrinogen measures is required to red uce misclassification error to less than 10%. Sample size estimates we re made based on predetermined levels of statistical power and the 6-w eek intraindividual and interindividual variability estimates. The abi lity to stratify an individual patient based on a single measurement o f fibrinogen is limited by methodological and intraindividual componen ts of variability. Clinical studies reporting the influence of specifi c therapeutic interventions on fibrinogen concentrations must account not only for interindividual variability but for methodological and in traindividual components of variability as well.