The drowning whipple: perioperative fluid balance and outcomes following pancreaticoduodenectomy

Abstract

Background and objectives: Given the high incidence of postoperative morbidity following pancreaticoduodenectomy (PD), efforts at improving patient outcomes are vital. We sought to determine the impact of perioperative fluid balance on outcomes following PD in order to identify a targeted strategy for reducing morbidity.

Methods: A retrospective review of consecutive PDs from 2008 to 2012 was completed. Cumulative fluid balances were recorded at 0, 24, 48, and 72 hr postoperatively and patients were divided into quartiles. Multivariate analyses were performed accounting for age, gender, diagnosis, ASA class, estimated blood loss, colloid and blood product use, and hemoglobin nadir. The predefined primary outcome measures were 90-day morbidity (Clavien grade ≥ III), mortality, and hospital readmission.

Results: One hundred sixty-nine PDs were performed during the study period. The 90-day morbidity and mortality rates for the cohort were 40.2% and 3.0%, respectively, while hospital length of stay was 13.6 ± 6.7 days (mean ± SD). Higher fluid balance at 48 and 72 hr postoperatively was an independent predictor of morbidity and length of stay on multivariate analysis.

Conclusions: Higher postoperative fluid balance is associated with increased postoperative morbidity and longer hospital stay following PD. Efforts at maintaining a fluid-restrictive strategy should be emphasized in this population.

Keywords: fluid overload; morbidity; pancreaticoduodenectomy; surgical outcomes.