Iaremyna I. Medical and social substantiation of economic mechanisms to optimize the hospital care at secondary level (on the basis of Vinnitsa city in-patient departments).

Dissertation laid scientific substantiation to reduce the risk of hospital readmissions on the basis of specially developed econometric model of hospital physician's behavior under different patient load with Slutsky decomposition. Analytically derived effects substantiated that the increase of the competition leads to: a) reduction of the number of services provided by a hospital physician; b) forward integration towards activities enhancing the quality of treatment; с) increase in the expected (average) length of hospital stay. These theoretically derived conclusions have been verified empirically. In particular, we elucidated current trends that are: a) reduction in the number of physicians at inpatient departments; b) increased patients load and patient-days load per physician; c) decrease in the average length of stay in hospital. According to the results of three specification of Cox mixed PH model based on panel data of 2220 patients the increase in the patients load per physician significantly (p <0.05) reduce the average stay of patient in hospital. Estimates of the effects by frailty survival model have supported that each additional day of stay safeguarded significant increase of period between consecutive hospitalizations by 2.5% being adjusted for age, diagnosis and other potent confounders. Substantiated economic mechanisms to reduce the risk of recurrent hospitalizations were suggested, namely: a) compensation of the shift of physician's profit curve by introduction of mixed payment mechanisms supported quality of services; b) an increase in the consumer price elasticity of demand for inpatient services; c) reduction of the monopoly power of hospitals as producer of inpatient services; d) strengthening the monopsony power of consumers of inpatient medical services. Implementation of elements of the substantiated system proved to be effective by data of expertise and natural experiment.