Taking Theory to Practice: Quality Improvement for Pancreaticoduodenectomy and Development and Integration of the Fistula Risk Score.

Abstract

Improvements in surgical outcomes are predicated on recognizing effective practices with subsequent adaptation. It is unknown whether risk assessment for pancreatic fistula (clinically relevant postoperative pancreatic fistula [CR-POPF]) after pancreaticoduodenectomy (PD) translates to improved patient outcomes at the practice level. A prospectively collected, single-surgeon career experience (2003 to 2018) of 455 consecutive pancreatectomies (303 PDs and 152 distal pancreatectomies) was examined. Analysis occurred during 4 eras of practice: learning curve for PD (n = 50); development of the Fistula Risk Score (n = 48); reactive, data-driven adjustments of anastomotic stent use (n = 94); and omission of prophylactic octreotide with adoption of selective drainage (n = 111). Observed to expected ratios of CR-POPF were calculated using a multi-institutional derivation set (5,379 PDs). After adjustment for increasing fistula risk across the 4 eras (p = 0.016), the risk-adjusted CR-POPF rate declined significantly (observed to expected ratio 1.42→1.28→1.01→0.30; p < 0.001). Literature-driven changes in fistula mitigation strategies likewise led to reductions in the overall complication burden (Postoperative Morbidity Index: 0.20→0.24→0.25→0.15; p = 0.015) and resource use (therapeutic antibiotics: p = 0.019; hospital readmission: p = 0.006; postoperative transfusion: p = 0.007). In contrast, the CR-POPF rate after distal pancreatectomy, for which no validated risk-adjustment process exists, did not vary (approximately 12%; p = 0.878). Patient outcomes for PD can be optimized by risk-adjusted evaluation and deliberate modification of practice.