Forecasting of need in postcardiosurgery hospitalizations on the basis of symptom-syndrome structuring of factors

Background. Study of predictive need in the future hospitalizations still was not carried out though it can be useful for solution of both medical and social-economic issues of health care.  Objective . Study of a possibility of forecasting of repeated hospitalizations in various terms after cardiac surgery on a set of characteristics of the patients in pre- and postoperative (hospital) periods. Design and methods . Results of long regular postoperative follow up (in terms up to 12 years) for 114 patients (80 men and 34 women) with the infective endocarditis (IE) after valve replacement of various localizations are analysed. The initial database included 133 factors — demographic, clinical, laboratory and tool characteristics of patients. Frequency and the reasons of hospitalizations, symptoms of disease complications — the prosthesis depended and independent were considered. In order to solve the prognostic problem, there were used a symptom-syndrome approach, providing expansion of the factor space due to the most significant factor combinations and the method of projective classification of patients, which provides a reduction of the degree in predictive randomness. Results . Three integrated factors, the most significant for forecasting of repeated hospitalizations are allocated. The first interpreted as “a factor of adequate treatment” included belonging to a male gender, maintaining working capacity and no need in postoperative use of cardiac glycosides (CG); the second one unites such variants as existence of either regular atrial fibrillation or need in implementation of temporary cardiostimulation in patients with HF III–IV (NYHA). The third integrated factor consisted of a combination of aortic valve prosthesis, nonphysical work and need in postsurgery inotrop support. Combination of two and more integrated factors resulted in high probability (the importance of distinction 0.7) allocations of groups with low and high risk of repeated hospitalizations — 9.5 % (4 of 43 patients) and 58.1 % (18 patients from 31) correspondingly. Conclusion . The proposed technique can be used for similar purposes in various categories of patients for advance planning of treatment, preventive, social and economic measures.