Value of some clinical parameters in predicting continuous renal replacement therapy-related hypotension

Abstract

To screen the clinical parameters in predicting continuous renal replacement therapy (CRRT)-related hypotension in the patients with renal failure. A retrospective analysis was conducted. Patients with renal failure received CRRT admitted to Qingdao Municipal Hospital from July 1st, 2012 to June 30th 2019 were enrolled. Clinical data was recorded for the patients, including gender, age, weight, parameters before CRRT [systolic pressure (SBP), diastolic pressure (DBP), mean arterial pressure (MAP), central venous pressure (CVP), heart rate, blood routine examination, blood biochemistry, B-type natriuretic peptide (BNP), cardiothoracic ratio, left ventricular ejection fraction (LVEF)], CRRT duration, blood flow velocity, replacement fluid volume, net ultrafiltration volume, ultrafiltration rate and 30-day prognosis. The patients who had CRRT-related hypotension or whose net ultrafiltration was zero were enrolled as intolerance ultrafiltration group. Others were enrolled in normal ultrafiltration group. The parameters of the patients in the two groups were compared, and their predictive values in CRRT-related hypotension were evaluated by receiver operating characteristic (ROC) curve analysis. There was significant difference in age, BNP, CVP, CRRT duration, net ultrafiltration volume, and ultrafiltration rate between the normal ultrafiltration group (n = 98) and intolerance ultrafiltration group [n = 13; age (years old): 66.21±12.21 vs. 74.54±5.93, BNP (ng/L): 1 617.57 (755.00, 2 296.25) vs. 398.95 (76.80, 703.00), CVP (cmH2O, 1 cmH2O = 0.098 kPa): 14.99±2.28 vs. 12.60±1.14, CRRT duration (hours): 23.71±11.48 vs. 14.51±8.99, net ultrafiltration volume (mL): 3 120.98 (1 863.75, 3 515.00) vs. 793.85 (0, 1 125.00), ultrafiltration rate (mL/h): 133.44 (98.72, 156.64) vs. 68.47 (0, 105.21), all P < 0.05]. ROC curve analysis showed that age, BNP, CVP, CRRT duration, net ultrafiltration volume and ultrafiltration rate had good predictive value for the CRRT-related hypotension [the area under ROC curve (AUC) and 95% confidence interval (95%CI): age was 0.734 (0.626-0.842), BNP was 0.868 (0.776-0.960), CVP was 0.846 (0.757-0.935), CRRT duration was 0.746 (0.595-0.897), net ultrafiltration volume was 0.926 (0.845-1.000), and ultrafiltration rate was 0.794 (0.620-0.969), respectively]. Furthermore, according to the cut-off point of ultrafiltration rate, the patients were divided into ≥ 99.50 mL/h group and < 99.50 mL/h group. There were statistical differences in BNP, CVP and net ultrafiltration volume between the two groups. ROC curve analysis showed that BNP and CVP had good prediction value for hypotension when ultrafiltration rate exceeds ≥ 99.50 mL/h. The AUC (95%CI) of BNP and CVP were 0.716 (0.604-0.828) and 0.749 (0.651-0.847), the sensitivity was 70.0%, 78.8%, and the specificity was 35.5%, 45.2%, respectively. The values of age, BNP and CVP are more useful than other parameters in predicting CRRT-related hypotension before the start.