Quality assessment of the screening test for cerebral stroke predictors

Introduction. Cerebral stroke is an important unsolved problem. An employee of a locomotive crew operating a locomotive cannot perform his functions and ensure the safety of the transportation process in case of a cerebral stroke. Medical expert commissions do not guarantee the absence of a cerebral stroke in an employee during his work. Aim. To assess the quality of the screening test for predictors of stroke using statistical quality control of the verified diagnostic test (with binary outcomes). Materials and methods. 19 cases of cerebral stroke were registered among the respondents during the 2008-2013 follow-up of 22 clinical and anamnestic items of 7,959 initially healthy men who were employees of locomotive crews of the Trans-Baikal Railway aged 18-66 years old. Statistical analysis identified the following predictors of cerebral stroke: smoking, hyperglycemia, arterial hypertension, left ventricular myocardial hypertrophy, excessive alcohol consumption, pulse wave velocity over 12 m/s, age 46-66 years, as well as critical age periods of 46 and 60 years old, the latter being studied as separate factors. DiagStat software evaluated their predictive ability for use in screening tests. Results. Predictors of cerebral stroke have high and moderate specificity regarding the absence of latent stroke formation in those who do not have them with a probability of 99.6-99.9%. The «age of 60 years old» predictor increases the odds of stroke occurrence by 68.8 times, against its absence, after the patient receives a positive result of a verified diagnostic test in a patient. Conclusion. When making a decision regarding an employee who operates a locomotive, one should focus on both the probability of a stroke and the absence of this probability in the presence or the absence of the risk factor cerebral stroke in a patient. It is necessary to carry out a targeted search for the predictors of a cerebral stroke such as a pulse wave velocity of more than 12 m/s and excessive alcohol consumption, since the detection of these factors among workers of locomotive crews is not stipulated. Other predictors of cerebral stroke need to be identified and eliminated, as there are techniques for this.