[Preliminary study of the arm equilibrium pressure to predict the effect of fluid challenge on urine output in oliguric intensive care unit patients].

Objective: To evaluate whether arm equilibrium pressure (Parm) is helpful to predict the effect of fluid load in improving oliguria in intensive care unit(ICU) patients. Methods: Hemodynamically stable patients [mean artery pressure (MAP)>65 mmHg (1 mmHg=0.133 kPa), heart rate (HR)<120 beats/min, lactic acid<2 mmol/L] with urine output (UO)<0.5 ml·kg(-1)·h(-1) for 3 consecutive hours were enrolled. The fluid loading was performed by infusion of ringer's lactate 500 ml within 30 minute after baseline hemodynamic data were recorded. The positive renal response was defined as UO increased more than 0.5 ml·kg(-1)·h(-1) 1 hour after fluid challenge, otherwise was negative. Results: A total of 30 oliguric ICU patients were enrolled including 17 males and 13 females with median age (54.2±16.3) years. After fluid load, patients' HR decreased[(84±13)beat/min vs. (80±10) beat/min, P<0.01], central venous pressure (CVP) increased[(7.0±2.4)mmHg vs. (8.8±2.6) mmHg, P<0.01], 30s Parm [(33.4±5.3) mmHg vs. (35.4±5.8) mmHg, P<0.01] and 60s Parm [(26.9±4.5) mmHg vs. (28.7±5.0) mmHg, P<0.01] increased, and UO [(18.5±8.8)ml/h vs. (64.1±38.3)ml/h, P<0.01] increased significantly, while MAP and lactic acid did not change (P>0.05). There were eighteen renal responders and 12 patients did not response. In responding group, MAP[(78.1±10.7) mmHg vs. (91.2±11.7) mmHg, P<0.01], 30s Parm[(30.4±3.8) mmHg vs. (38.0±3.7) mmHg, P<0.01] and 60s Parm [(24.3±2.5) mmHg vs. (30.8±4.0) mmHg, P<0.01] before fluid load were lower than those in negative group. HR, CVP, lactic acid, age and body weight were comparable between two groups (P>0.05). After volume loading, MAP, 30s and 60s Parm in positive group were still lower than those in negative group (P<0.05), while HR, CVP and lactic acid were similar (P>0.05). Correlation analysis showed that baseline 30s Parm (r=-0.75, P<0.01), 60s Parm (r=-0.69, P<0.01), and MAP (r=-0.46, P<0.05) were negatively correlated with 1 h UO after fluid load, but HR and CVP were not (P>0.05). The receiver operating curve (ROC) showed that 30s Parm had the largest area under curve (AUC) of 0.94 (95% CI 0.84-1.05, P<0.01), which 35.5 mmHg was the best threshold with sensitivity 94.4% and specificity 91.7%(likelihood ratio 11.37). Conclusion: In hemodynamically stable oliguric ICU patients, if Parm is lower than normal reference value, volume expansion is more likely to increase UO. Thus Parm can be used to predict the effect of fluid loadon UO.