The aim of the study was to evaluate the efficacy of various approaches to correcting and preventing hypotension episodes in patients on maintenance hemodialysis (HD).Material and methods. The study included 35 patients on maintenance hemodialysis in the Dialysis Center "MCHTP No. 1", which is a part of "B. Braun Avitum" network centers in the Russian Federation. All patients underwent an automatic blood pressure (BP) measurement using a machine-inbuilt option device for noninvasive blood pressure measurement. Prior to the study, all patients underwent a clinical test with a "dry weight" assessment and a bio-impedance analysis. The study had a cross-over design: first, all patients were treated using the standard methods for correcting hypotension episodes (at 4 initial procedures). Then, in the following 4 procedures, in addition to standard methods, a computer algorithm was used to automatically regulate the ultrafiltration (UF) rate: the automatic pressure monitoring system (biologic rr comfort) with continuous monitoring of blood pressure throughout the procedure; BP was recorded before and after the HD procedure, as well as at least once every 5 minutes during 3 initial procedures; and starting from the 4th procedure, the intervals were determined by the algorithm automatically. The average blood pressure values were analyzed during the dialysis procedures for the entire observation period. The duration of the study was 3 weeks for each patient.Results. The average predialysis blood pressures in the group with the standard approach to hemodynamic correction were 124.6 ± 27.7 and 74.5 ± 21.1 mm Hg, the postdialysis blood pressures were 114.4 ± 24.4 and 71.3 ± 16.3 mm Hg. With the use of automatic pressure monitoring system, the predialysis and postdialysis blood pressures were significantly higher than those with the standard approach: 133.2 ± 21.3 and 79.3 ± 15.8 mm Hg (p < 0.001 and p = 0.009), vs. 125.7 ± 23.9 and 75.9 ± 18.3 mm Hg (p < 0.001 and p < 0.001), respectively. Upon closer examination of the intradialysis pressure variations, the intradialysis blood pressures were 110.2 ± 17.3 and 68.3 ± 13.9 mm Hg when measured by using the standard approach, and significantly higher: 124 ± 20.5 and 75.9 ± 14.2 mm Hg when the automatic pressure monitoring system was used (p = 0.03; p = 0.02). Also, higher mean arterial pressures were noted: 82.5 ± 13.9 with the standard approach vs.91.5 ± 15.6 mm Hg (p = 0.01) with the automatic pressure monitoring system. Studying UF rates, we found that the UF rate was slightly higher without using the automatic pressure monitoring system (8.0 ml/kg/h vs. 7.9 ml/kg/h). Thus, the new approach used in addition to the standard methods of correcting hypotension was effective and safe. No significant differences were seen in Kt/V values. However, when automatic pressure monitoring system was used in patients, the target phosphate levels were achieved: the inorganic phosphorus value was 1.5 mmol/L when using the UF control algorithm vs. 1.8 mmol/L with a standard dialysis program. However, these data did not reach the statistical significance (p = 0.07).Conclusion. Intradialysis hypotension and high UF rates remain frequent and potentially dangerous complications of HD procedure, which worsen the long-term prognosis of patients on HD, mainly due to the increase in cardiovascular morbidity and mortality. The new approach to the prevention and correction of hypotension by using the automatic pressure monitoring system allows one to reduce the UF rate in a timely manner, preventing the development of hypotension episodes, reducing their rates, and improving the achievement of target blood pressure values, both preand postdialysis, as well as intradialysis blood pressure variations.
Partition Learning for Multiagent Planning