Quality assessment of the screening test for predictors of coronary heart disease

Abstract

Background. Coronary heart disease (CHD) ranks first among the causes of death, morbidity, and disablement. The development of innovative methods for predicting CHD will reduce these losses.The aim of the work. To assess the quality of the screening test for predictors of coronary heart disease using statistical quality control of a verifiable diagnostic test (with binary outcomes).Materials and methods. In 2008–2013, 70 cases of CHD were registered in a groupof 7959 initially healthy men 18–66 years old who were the members of locomotive crews. Statistical analysis identified CHD predictors: arterial hypertension; psychosocial stress; hyperglycemia; dyslipidemia; excessive alcohol consumption; obesity of degree I–III; age 34–66 years; microalbuminuria; thickening of the intima- media complex/atherosclerotic plaque (IMC/ASP); pulse wave velocity (PWV) > 12 m/s; left ventricular hypertrophy; grade I–II retinopathy; atherosclerosis of aorta. DiagStat (Russian Federation) software determined their predictive ability when used in screening tests to predict CHD. We demonstrated the use of this method to assess the predictive ability of risk factors for any disease.Results. CHD predictors have high to moderate specificity for the absence of CHD in individuals who test negative for the above-listed factors. IMC/ASP, microalbuminuria, PWV > 12 m/s, grade III obesity moderately increase the posterior odds of developing CHD versus its absence in comparison with the prior odds after receiving a positive result of the verifiable diagnostic test for these factors. Age 34–66 years moderately increases the posterior odds in favor of the absence of CHD versus its occurrence compared with the prior odds after receiving a negative result of the verifiable diagnostic test.Conclusion. When assessing the result of the verifiable diagnostic test, we should focus on both the probability of occurrence and the absence of CHD in the presence or absence of a predictor in the patient. Since the determination of PWV > 12 m/s, atherosclerosis of aorta, microalbuminuria, stress, and excessive alcohol consumption among workers of locomotive crews is not mandatory, it is necessary to conduct a targeted search for them