In-hospital survival after pancreatoduodenectomy is greater in high-volume hospitals versus lower-volume hospitals: a meta-analysis

Approach to postpancreatectomy care Impacts outcomes: Retrospective Validation of the PORSCH trial

BACKGROUND

In the recent PORSCH trial, a three-part postpancreatectomy care algorithm was employed with a near 50 ​% reduction in mortality. We hypothesized that clinical care congruent with this protocol would correlate with better outcomes in our patients.

METHODS

Real-world postoperative care was compared to the pathway described by the PORSCH trial and patients were assigned into groups based on congruence with its recommendations. The primary composite outcome (PCO) consisted of 90-day mortality, organ failure, and interventions for bleeding.

RESULTS

Of 289 patients, care of 12 ​% was entirely congruent with the PORSCH algorithm. The PCO was recorded in 9 ​% of the PORSCH care group, 8 ​% of the Partial-PORSCH care group, and 19 ​% of the Non-PORSCH care group (p ​= ​0.044). Adverse outcomes were highest when pancreaticoduodenectomy patients received care incongruent with the algorithm's CT imaging recommendations.

CONCLUSIONS

These results add external validity to the principles of clinical care underlying the PORSCH algorithm.

Nationwide Outcomes of Pancreaticoduodenectomy for Pancreatic Malignancies: Center Volume Matters

BACKGROUND

Complex surgeries such as pancreaticoduodenectomies (PD) have been shown to have better outcomes when performed at high-volume centers (HVCs) compared to low-volume centers (LVCs). Few studies have compared these factors on a national level. The purpose of this study was to analyze nationwide outcomes for patients undergoing PD across hospital centers with different surgical volumes.

METHODS

The Nationwide Readmissions Database (2010-2014) was queried for all patients who underwent open PD for pancreatic carcinoma. High-volume centers were defined as hospitals where 20 or more PDs were performed per year. Sociodemographic factors, readmission rates, and perioperative outcomes were compared before and after propensity score-matched analysis (PSMA) for 76 covariates including demographics, hospital factors, comorbidities, and additional diagnoses. Results were weighted for national estimates.

RESULTS

A total of 19,810 patients were identified with age 66 ± 11 years. There were 6,840 (35%) cases performed at LVCs, and 12,970 (65%) at HVCs. Patient comorbidities were greater in the LVC cohort, and more PDs were performed at teaching hospitals in the HVC cohort. These discrepancies were controlled for with PSMA. Length of stay (LOS), mortality, invasive procedures, and perioperative complications were greater in LVCs when compared to HVCs before and after PSMA. Additionally, readmission rates at one year (38% vs 34%, P < .001) and readmission complications were greater in the LVC cohort.

CONCLUSIONS

Pancreaticoduodenectomy is more commonly performed at HVCs, which is associated with less complications and improved outcomes compared to LVCs.