ObjectiveTo evaluate the predictive value of C-reactive protein/albumin ratio (CAR) for postoperative complications of pancreaticoduodenectomy (PD).MethodsThe clinical data of 134 patients with pancreaticoduodenectomy (PD) in the Department of Pancreatic Surgery of The First Affiliated Hospital of Xinjiang Medical University from 2015 to 2018 were retrospectively collected, including general conditions and postoperative complications. The predictive value of CAR was calculated.ResultsOf the 134 patients, 38 patients suffered from postoperative pancreatic fistula (POPF), 32 patients suffered from abdominal infection, 5 patients suffered from biliary fistula, 63 patients suffered from delayed gastric emptying (DGE), 13 patients suffered from post pancreatectomy hemorrhage (PPH), 8 patients suffered from wound infection, and 1 patient suffered from chyle leakage. There was no significant difference in general conditions such as gender, age, BMI, American Society of Anesthesiologists (ASA) score, tumor nature, tumor diameter, operative time, intraoperative blood loss, diabetes history, jaundice history, and drinking history (P>0.05), but the hospital stay in the complication group was longer than that of the non-complication group (P<0.05). The value of CAR in the pancreatic fistula and abdominal infection group were significantly higher than those in the non-complication group at 1 d, 3 d and 5 d (exclude 1 day after surgery on POPF), the difference was statistically significant (P<0.05). On the 3rd day after surgery, the sensitivity of CAR predicting POPF was 79.95%, the specificity was 86.46%; the sensitivity of predicting abdominal infection was 75.00%, the specificity was 81.37%, and the result was better than using procalcitonin (PCT) alone, but similar with C-reactive protein (CRP) alone or CRP+PCT.ConclusionPostoperative CAR can better predict POPF and abdominal infection after PD, and the effect is better than PCT alone.