BACKGROUND: Serum C-reactive protein (CRP) is an established discriminating factor for necrotizing pancreatitis. In this study, the CRP response with respect to inflammatory postoperative complications was examined in a large, homogeneous series of pancreatic resections in order to define a relevant clinical parameter for early detection of inflammatory postoperative complications. METHODS: 688 consecutive pancreatic resections with jejunal anastomosis were screened for inflammatory postoperative complications based on a prospective database. Ninety-one patients had at least one inflammatory postoperative complication and were compared to a subgroup of 60 consecutive patients with uneventful postoperative courses. RESULTS: In the postoperative setting after pancreatic resection, CRP peaked on postoperative day (POD) 3 with a median serum CRP of 132 mg/L, and gradually decreased thereafter in patients with an uncomplicated postoperative course. In complicated cases (with the exception of cholangitis), increase in CRP was significantly greater, peaked on POD 3 (median CRP 173 mg/L), and persisted thereafter, whereas white blood cell count and body temperature did not differ significantly from uneventful courses until POD 6. The median day of diagnosis of inflammatory postoperative complications was POD 9. A cutoff CRP value of 140 mg/dL on POD 4 yielded a positive predictive value of 89.1% (adjusted to the prevalence of inflammatory postoperative complications: 48.7%) with a specificity of 87.1% and a sensitivity of 69.5% for inflammatory postoperative complications. CONCLUSION: Persistence of CRP elevation above 140 mg/dL on POD 4 is predictive of inflammatory postoperative complications and should prompt an intense clinical search for major septic processes (e.g. pancreatic fistula or abscess) if pneumonia and wound infection are unlikely or excluded.
BACKGROUND: Serum C-reactive protein (CRP) is an established discriminating factor for necrotizing pancreatitis. In this study, the CRP response with respect to inflammatory postoperative complications was examined in a large, homogeneous series of pancreatic resections in order to define a relevant clinical parameter for early detection of inflammatory postoperative complications. METHODS: 688 consecutive pancreatic resections with jejunal anastomosis were screened for inflammatory postoperative complications based on a prospective database. Ninety-one patients had at least one inflammatory postoperative complication and were compared to a subgroup of 60 consecutive patients with uneventful postoperative courses. RESULTS: In the postoperative setting after pancreatic resection, CRP peaked on postoperative day (POD) 3 with a median serum CRP of 132 mg/L, and gradually decreased thereafter in patients with an uncomplicated postoperative course. In complicated cases (with the exception of cholangitis), increase in CRP was significantly greater, peaked on POD 3 (median CRP 173 mg/L), and persisted thereafter, whereas white blood cell count and body temperature did not differ significantly from uneventful courses until POD 6. The median day of diagnosis of inflammatory postoperative complications was POD 9. A cutoff CRP value of 140 mg/dL on POD 4 yielded a positive predictive value of 89.1% (adjusted to the prevalence of inflammatory postoperative complications: 48.7%) with a specificity of 87.1% and a sensitivity of 69.5% for inflammatory postoperative complications. CONCLUSION: Persistence of CRP elevation above 140 mg/dL on POD 4 is predictive of inflammatory postoperative complications and should prompt an intense clinical search for major septic processes (e.g. pancreatic fistula or abscess) if pneumonia and wound infection are unlikely or excluded.
Authors: Yakup Kulu; Bruno M Schmied; Jens Werner; Pietro Muselli; Markus W Büchler; Jan Schmidt Journal: HPB (Oxford) Date: 2009-09 Impact factor: 3.647