Inability to manage non-severe complications on an outpatient basis increases non-white patient readmission rates after pancreaticoduodenectomy: a large metropolitan tertiary care center experience

Abstract

Presenter: Jorge G Zarate Rodriguez MD | Washington University, St. Louis Background: Racial disparities play an important role in readmissions after surgery. Pancreaticoduodenectomy (PD) is associated with high rates of morbidity and readmission, but not all readmissions are equal in terms of severity. We hypothesize that, secondary to differences in access to care, white patients are managed more successfully as outpatients for non-severe complications compared to non-white patients. This may be a factor contributing to the increased readmission rate for non-white patients after PD. Methods: Patients undergoing PD were prospectively followed, and their complications graded using the Modified Accordion Grading System (MAGS). Complications were categorized as non-severe if MAGS 1-2 and severe if MAGS ≥3. Readmissions were categorized based on complication severity—a non-severe readmission was defined as a patient readmission for a MAGS non-severe complication, and severe readmission was defined as a readmission for a MAGS severe complication. Univariate and multivariate analysis were used to compare patients with MAGS non-severe and MAGS severe readmission to patients without readmission. Results: 837 patients underwent PD, the 90-day readmission rate was 27.5% and 51.3% of readmissions were for non-severe complications. The overall readmission rate was 25.8% for white patients and 39.3% for non-white (p = 0.004). Non-white patients with readmission were 81.0% Black, 11.9% Asian, 4.8% Latinx, and 2.4% other race. Readmitted patients were more likely to be non-white (18.3% vs 10.7%, p = 0.004), obese (34.4% vs 27.5%, p = 0.047) and have soft glands (37.8% vs 28.8%, p = 0.012), and less likely to have pancreatic adenocarcinoma (49.1% vs 57.0%, p = 0.041). Patients with readmission were more likely to have suffered complications during their initial surgical hospitalization (62.7% vs 48.6%, p <0.001), both MAGS non-severe (53.5% vs 44.3%, p = 0.018) and MAGS severe (26.1% vs 12.0%, p < 0.001). More specifically, patients with readmission were more likely to have suffered bleeding complications (15.7% vs 10.5%, p = 0.042), MAGS severe pancreatic fistula (10.0% vs 4.5%, p = 0.003), MAGS severe delayed gastric emptying (7.0% vs 1.3%, p <0.001), and DVT (5.2% vs 1.7%, p = 0.004). Readmitted patients also had longer initial surgical hospitalizations (p = 0.044). Non-white race was independently associated with readmission (OR 1.83, p = 0.007). The MAGS non-severe readmission rate for white patients was 14.7% and 27.8% for nonwhite (p = 0.002), while the MAGS severe readmission rate for white patients was 14.9% vs 20.7% for non-white (p = 0.172). Non-white race was independently associated with MAGS non-severe readmission (OR 2.13, p = 0.006), but not MAGS severe readmission. Non-white patients were more likely to be readmitted without requiring an intervention that met criteria for a MAGS complication compared to white patients (21.4% vs 9.6%, p = 0.031). Conclusion: Non-white patients are more likely to be readmitted, particularly for non-severe complications. Non-complication associated elements, such as patient, provider, and healthcare system related factors, are likely driving the disproportionate frequency of readmissions in non-white patients after PD. Follow up protocols should be tailored to address racial disparities in the rates of readmission for less severe complications, which could potentially be avoidable.