Symptom dynamics assessment of the disease by methods of survival analysis

Abstract

The article discusses typical tasks of quantitative comparison and analysis of the effectiveness of diagnostic medical technologies. Indicators of clinical informativeness of diagnostic methods, quantitative methods for their calculation and interpretation, their importance for making diagnostic decisions are discussed. Using one of the diagnostic methods as an example, algorithms for comparing the test diagnostic test with the gold standard test according to clinical information indicators are described: sensitivity, specificity, positive and negative predictive value of test results, accuracy, likelihood ratio of positive and negative test results. The problems of constructing ROC curves by the example of diagnostic indicators of nasal obstruction: airflow and pressure, as well as the calculation and presentation of AUC (in one figure), finding diagnostic threshold points for two tests (changes in flow and resistance), and testing the statistical hypothesis about the equality of AUC of these two tests, creating a nomogram for calculating the post-test probability of illness. It is shown how, using the presented technique, it is possible to efficiently calculate all the standard operational characteristics of diagnostic medical technologies and additional useful indicators. All calculations were performed in the statistical program R. The text of the article presents program codes of the R language with explanations.