AKI indicates a rapid decrease in renal function with a broad spectrum of severity, ranging from mild serum creatinine elevation to the complete loss of renal function. Postoperative AKI accounted for 18–47% of all causes of hospital-acquired AKI [3] and associated with considerable morbidity and mortality [21]. Elderly, smoking, high American society of anesthesiologists classification, comorbidities such as hypertension, diabetes mellitus, and coronary artery disease (CAD) were all risk factors for AKI after thoracic surgery [3], One study showed that the incidence of AKI was 1.8% in patients undergoing thoracoscopic lobectomy, but elder patients and patients with CAD were excluded [22]. Another study showed that the incidence of AKI was 6% after open lobectomy surgery [23]. Because of the mini-invasive method, the incidence of AKI may be lower after thoracoscopic surgery. The incidence of AKI in our study was 6.2%, because the patients in our study were elderly, with more comorbidities, and the operation duration was more than 2 hours, so the incidence of AKI was higher in our study. There were several diagnostic criteria of AKI [24], the diagnostic indicators included serum creatinine and urine output. The postoperative urine output record was inaccurate because the urinary catheter was removed immediately after anesthesia recovery in our study, studies also showed that urine output was unrelated to intraoperative infusion volume and perioperative kidney function [25, 26], so only creatinine level was used to diagnosis AKI in our study.
The relationship between fluid administration and kidney function have been studied. Compared to intraoperative infusion rate of 10.9 ml/kg/h, intraoperative infusion rate of 6.5 ml/kg/h was associated with a higher AKI incidence after major abdominal surgery [27]. Another study of 92094 patients undergoing noncardiac surgery showed an increased incidence of AKI when intraoperative fluid volume was less than 900 ml for a 3-hour operation [28]. In open thoracic surgery, compared to intraoperative net infusion rate of more than 6 ml/kg/h, the AKI incidence was higher when intraoperative net infusion rate of less than 3 ml/kg/h [23]. In our study, the total infusion volume was larger in GDFT group than that in RFT group, though there was no significant difference of AKI incidence between the two groups, but the postoperative increase degree of serum creatinine was lower in GDFT group than that in RFT group. Our result showed decreasing trend of AKI with increasing fluid volume.
GDFT using fluids, inotropes and vasopressors, but not fluids alone, could reduce the postoperative complications, because the bundle protocol could improve hemodynamics and tissue perfusion [29]. Compared to intraoperative infusion rate of 6.5 ml/kg/h, patients who received infusion rate of 10.9 ml/kg/h during major abdominal surgery had higher CI at the end of operation, because the higher CI could increase perfusion in organs sensitive to hypovolemia, this may explain the lower rate of AKI in the liberal fluid group [30]. In our study, intraoperative CI were higher in GDFT group than that in RFT group, the higher CI may be another explanation of lower increase degree of postoperative serum creatinine in GDFT group.
Both colloids and crystalloids can be used in GDFT protocol. Previous studies showed that intraoperative colloid usage was not associate with an increased risk of AKI after cardiac surgery [31], liver transplantation surgery [32], and thoracoscopic lobectomy [22]. Meta-analysis showed that postoperative kidney dysfunction was similar between colloid-based GDFT and crystalloid-based GDFT after major non-cardiac surgery [33]. Colloid was more effective at expanding blood volume and stabilizing hemodynamic parameters [34], in our study, colloid was chosen for fluid resuscitation in GDFT group, thus avoiding excessive fluid infusion.
Excessive perioperative fluid administration was a risk factor for postoperative pulmonary complications (PPCs). Study showed that intraoperative crystalloid fluid more than 6 ml/kg/h was independent risk factor of PPCs after thoracoscopic lung resection [35], the PPCs was very high after lung resections when intraoperative infusion rate was more than 8 ml/kg/h [36]. In our study, the infusion rate was less than 6 ml/kg/h in both groups, the incidence of PPCs was low and consistent between the two groups. The cardiac complications after pulmonary lobectomy were low and more related to surgical trauma [37], so there were no significant differences of cardiac complications between the two groups.
Our study also has limitations. First, fluid management after operation was not standardized, it may skew the effect of intraoperative fluid optimization. Second, the implementation rate of some ERAS program items was low, it may affect the results. Third, the postoperative AKI may be associated with continuing renal dysfunction for up to 3 months, however, our study did not have long-term follow-up, the long term benefit of intraoperative fluid management was not investigated.