Objective: To determine the nationwide implementation and surgical outcome of minor and major robotic liver surgery (RLS) and assess the first phase of implementation of RLS during the learning curve. Background: RLS may be a valuable alternative to laparoscopic liver surgery. Nationwide population-based studies with data on implementation and outcome of RLS are lacking. Methods: Multicenter retrospective cohort study including consecutive patients who underwent RLS for all indications in 9 Dutch centers (August 2014–March 2021). Data on all liver resections were obtained from the mandatory nationwide Dutch Hepato Biliary Audit (DHBA) including data from all 27 centers for liver surgery in the Netherlands. Outcomes were stratified for minor, technically major, and anatomically major RLS. Learning curve effect was assessed using cumulative sum analysis for blood loss. Results: Of 9437 liver resections, 400 were RLS (4.2%) procedures including 207 minor (52.2%), 141 technically major (35.3%), and 52 anatomically major (13%). The nationwide use of RLS increased from 0.2% in 2014 to 11.9% in 2020. The proportion of RLS among all minimally invasive liver resections increased from 2% to 28%. Median blood loss was 150 mL (interquartile range 50–350 mL] and the conversion rate 6.3% (n=25). The rate of Clavien-Dindo grade ≥III complications was 7.0% (n=27), median length of hospital stay 4 days (interquartile range 2–5) and 30-day/in-hospital mortality 0.8% (n=3). The R0 resection rate was 83.2% (n=263). Cumulative sum analysis for blood loss found a learning curve of at least 33 major RLS procedures. Conclusions: The nationwide use of RLS in the Netherlands has increased rapidly with currently one-tenth of all liver resections and one-fourth of all minimally invasive liver resections being performed robotically. Although surgical outcomes of RLS in selected patient seem favorable, future prospective studies should determine its added value.

Difficulty scoring systems are important for the safe, stepwise implementation of new procedures. We designed a retrospective observational study for building a difficulty score for robotic pancreatoduodenectomy. The difficulty score (PD-ROBOSCORE) aims at predicting severe postoperative complications after robotic pancreatoduodenectomy. The PD-ROBOSCORE was developed in a training cohort of 198 robotic pancreatoduodenectomies and was validated in an international multicenter cohort of 686 robotic pancreatoduodenectomies. Finally, all centers tested the model during the early learning curve (n = 300). Growing difficulty levels (low, intermediate, high) were defined using cut-off values set at the 33rd and 66th percentile (NCT04662346). Factors included in the final multivariate model were a body mass index of ≥25 kg/m² for males and ≥30 kg/m² for females (odds ratio:2.39; P < .0001), borderline resectable tumor (odd ratio:1.98; P < .0001), uncinate process tumor (odds ratio:1.69; P < .0001), pancreatic duct size <4 mm (odds ratio:1.59; P < .0001), American Society of Anesthesiologists class ≥3 (odds ratio:1.59; P < .0001), and hepatic artery originating from the superior mesenteric artery (odds ratio:1.43; P < .0001). In the training cohort, the absolute score value (odds ratio = 1.13; P = .0089) and difficulty groups (odds ratio = 2.35; P = .041) predicted severe postoperative complications. In the multicenter validation cohort, the absolute score value predicted severe postoperative complications (odds ratio = 1.16, P < .001), whereas the difficulty groups did not (odds ratio = 1.94, P = .082). In the learning curve cohort, both absolute score value (odds ratio:1.078, P = .04) and difficulty groups (odds ratio: 2.25, P = .017) predicted severe postoperative complications. Across all cohorts, a PD-ROBOSCORE of ≥12.51 doubled the risk of severe postoperative complications. The PD-ROBOSCORE score also predicted operative time, estimated blood loss, and vein resection. The PD-ROBOSCORE predicted postoperative pancreatic fistula, delayed gastric emptying, postpancreatectomy hemorrhage, and postoperative mortality in the learning curve cohort. The PD-ROBOSCORE predicts severe postoperative complications after robotic pancreatoduodenectomy. The score is readily available via www.pancreascalculator.com.

The use of robot-assisted and laparoscopic pancreatoduodenectomy is increasing, yet large adjusted analyses that can be generalized internationally are lacking. This study aimed to compare outcomes after robot-assisted pancreatoduodenectomy and laparoscopic pancreatoduodenectomy in a pan-European cohort. An international multicenter retrospective study including patients after robot-assisted pancreatoduodenectomy and laparoscopic pancreatoduodenectomy from 50 centers in 12 European countries (2009-2020). Propensity score matching was performed in a 1:1 ratio. The primary outcome was major morbidity (Clavien-Dindo ≥III). Among 2,082 patients undergoing minimally invasive pancreatoduodenectomy, 1,006 underwent robot-assisted pancreatoduodenectomy and 1,076 laparoscopic pancreatoduodenectomy. After matching 812 versus 812 patients, the rates of major morbidity (31.9% vs 29.6%; P = .347) and 30-day/in-hospital mortality (4.3% vs 4.6%; P = .904) did not differ significantly between robot-assisted pancreatoduodenectomy and laparoscopic pancreatoduodenectomy, respectively. Robot-assisted pancreatoduodenectomy was associated with a lower conversion rate (6.7% vs 18.0%; P < .001) and higher lymph node retrieval (16 vs 14; P = .003). Laparoscopic pancreatoduodenectomy was associated with shorter operation time (446 minutes versus 400 minutes; P < .001), and lower rates of postoperative pancreatic fistula grade B/C (19.0% vs 11.7%; P < .001), delayed gastric emptying grade B/C (21.4% vs 7.4%; P < .001), and a higher R0-resection rate (73.2% vs 84.4%; P < .001). This European multicenter study found no differences in overall major morbidity and 30-day/in-hospital mortality after robot-assisted pancreatoduodenectomy compared with laparoscopic pancreatoduodenectomy. Further, laparoscopic pancreatoduodenectomy was associated with a lower rate of postoperative pancreatic fistula, delayed gastric emptying, wound infection, shorter length of stay, and a higher R0 resection rate than robot-assisted pancreatoduodenectomy. In contrast, robot-assisted pancreatoduodenectomy was associated with a lower conversion rate and a higher number of retrieved lymph nodes as compared with laparoscopic pancreatoduodenectomy.

Differences in patient demographics and disease burden can influence comparison of hospital performances. This study aimed to provide a case-mix model to compare short-term postoperative outcomes for patients undergoing liver resection for colorectal liver metastases (CRLM). This retrospective, population-based study included all patients who underwent liver resection for CRLM between 2014 and 2018 in the Netherlands. Variation in case-mix variables between hospitals and influence on postoperative outcomes was assessed using multivariable logistic regression. Primary outcomes were 30-day major morbidity and 30-day mortality. Validation of results was performed on the data from 2019. In total, 4639 patients were included in 28 hospitals. Major morbidity was 6.2% and mortality was 1.4%. Uncorrected major morbidity ranged from 3.3% to 13.7% and mortality ranged from 0.0% to 5.0%. between hospitals. Significant differences between hospitals were observed for age higher than 80 (0.0%-17.1%, p < 0.001), ASA 3 or higher (3.3%-36.3%, p < 0.001), histopathological parenchymal liver disease (0.0%-47.1%, p < 0.001), history of liver resection (8.1%-36.3%, p < 0.001), major liver resection (6.7%-38.0%, p < 0.001) and synchronous metastases (35.5%-62.1%, p < 0.001). Expected 30-day major morbidity between hospitals ranged from 6.4% to 11.9% and expected 30-day mortality ranged from 0.6% to 2.9%. After case-mix correction no significant outliers concerning major morbidity and mortality remained. Validation on patients who underwent liver resection for CRLM in 2019 affirmed these outcomes. Case-mix adjustment is a prerequisite to allow for institutional comparison of short-term postoperative outcomes after liver resection for CRLM.

Liver resection is high-risk surgery in particular in elderly patients. The aim of this study was to explore postoperative outcomes after liver resection in elderly patients. In this nationwide study, all patients who underwent liver resection for primary and secondary liver tumours in the Netherlands between 2014 and 2019 were included. Age groups were composed as younger than 70 (70-), between 70 and 80 (septuagenarians), and 80 years or older (octogenarians). Proportion of liver resections per age group and 30-day major morbidity and 30-day mortality were assessed. In total, 6587 patients were included of whom 4023 (58.9%) were younger than 70, 2135 (32.4%) were septuagenarians and 429 (6.5%) were octogenarians. The proportion of septuagenarians increased during the study period (aOR:1.06, CI:1.02-1.09, p < 0.001). Thirty-day major morbidity was higher in septuagenarians (11%) and octogenarians (12%) compared to younger patients (9%, p = 0.049). Thirty-day mortality was higher in septuagenarians (4%) and octogenarians (4%) compared to younger patients (2%, p < 0.001). Cardiopulmonary complications occurred more frequently with higher age, liver-specific complications did not. Higher age was associated with higher 30-day morbidity and 30-day mortality in multivariable logistic regression. Thirty-day major morbidity and 30-day mortality are higher after liver resection in elderly patients, attributed mainly to non-surgical cardiopulmonary complications.