Readmission and Resource Use After Robotic-Assisted versus Open Pancreaticoduodenectomy: 2010-2017

Abstract

Unplanned rehospitalization is considered an adverse quality of care indicator. Minimally invasive operations carry the potential to reduce resource use while enhancing recovery. Robotic-assisted pancreaticoduodenectomy (RAPD) has been used to improve outcomes of its morbid open counterpart. We sought to identify factors associated with readmission between RAPD and open pancreaticoduodenectomy (OPD). We used the 2010-17 National Readmissions Database to identify adults who underwent RAPD or OPD. The primary outcome was 30-day readmission. Secondary outcomes included readmission diagnosis: index, readmission, and total (index+readmission) length of stay, costs, and mortality. Of an estimated 84,036 patients undergoing pancreaticoduodenectomy, 96.9% survived index hospitalization. Frequency of both RAPD and OPD increased during the study period with similar mortality (2.5% versus 3.2%, P=0.46). Compared with OPD, RAPD was not an independent predictor of 30-day readmission (adjusted odds ratio (AOR): 1.0, P=0.98). Disposition with home health care (AOR: 1.1, P<0.001) or to a skilled nursing facility (AOR: 1.5, P<0.001) was significantly associated with increased 30-day readmission. Readmission after pancreaticoduodenectomy is common, regardless of surgical approach. Although RAPD saves in-patient days on index admission, readmission rates and length of stay are similar between the two modalities. Neither RAPD nor OPD is a risk factor for readmission, highlighting the complexity of pancreaticoduodenectomy, with complications that may result from factors independent of the operative approach.