Minimally invasive pancreaticoduodenectomy outcomes vary with hospital volume

Last Updated: 2017-01-06

By Will Boggs MD

NEW YORK (Reuters Health) - Complication rates associated with minimally invasive pancreaticoduodenectomy (MIPD) are significantly higher for patients treated at low-volume hospitals, according to a new study.

"The most interesting, and perhaps distressing, result of our study is that 83% of patients undergoing minimally invasive pancreaticoduodenectomy in the U.S. today have their surgery at a low-volume hospital, suggesting that, on average, the overwhelming majority of patients are potentially being exposed to significantly elevated risk of sustaining a postoperative complication," said Dr. Julie A. Sosa from Duke University Medical Center in Durham, North Carolina.

"Therefore, there is an opportunity to improve quality of care for many patients if providers, patients, policy-makers, and payers can come together to formulate solutions," she told Reuters Health by email.

Dr. Sosa's team used data from the Health Care Utilization Project National Inpatient Sample database to ascertain whether there is a minimum number of MIPD cases performed by a hospital that is associated with a reduced likelihood of postoperative complications.

Among 229 hospitals that performed MIPD during the study period, the number of MIPD cases ranged from one to 60 per year (median, six), the researchers report in JAMA Surgery, online December 28.

Nearly half the patients (47%) experienced a postoperative complication, and increasing hospital procedural volume was significantly associated with decreasing odds of experience a postoperative complication.

A Markov chain Monte Carlo simulation identified 22 cases per year as the best threshold for hospital procedural volume.

Only 17% of patients underwent surgery at high-volume (>22 cases per year) hospitals.

Patients treated at low-volume hospitals were significantly more likely to experience delayed gastric emptying, fistula formation, cardiopulmonary complications, or any complication, compared with patients treated at high-volume hospitals.

"For referring physicians, it is critical that they know where the high-volume MIPD centers are so that they can help to steer their patients to those hospitals when feasible," Dr. Sosa said. "For surgeons considering adopting MIPD and low-volume surgeons performing MIPD, it is critical that they first consider their experience with open pancreatectomy and also the systems support available to their patients at the hospital level to assure they are adequate to optimize patient safety; concentrating experience among a smaller number of high-volume colleagues at a local (institutional or practice) level might make sense."

Dr. Marco Del Chiaro from Karolinska University Hospital in Stockholm, Sweden, who coauthored an accompanying editorial, told Reuters Health by email that he found two results particularly interesting.

"On one hand, it is amazing that, without any objective advantage compared with open surgery, MIPD is clearly used more and more frequently over time," he said. "On the other hand, it seems that MIPD is (more often) used in low-volume centers than in high-volume ones."

Dr. Del Chiaro said he believes MIPD is clearly a safe procedure, but it shouldn't be used to attract patients. Rather, surgeons should try to identify which patients might benefit from a minimally invasive approach.

Another study published simultaneously in JAMA Surgery takes a step in that direction. Using propensity-score matching, Dr. Charles M. Vollmer Jr. from the University of Pennsylvania Perelman School of Medicine in Philadelphia and colleagues tried to determine whether robotic pancreaticoduodenectomy (RPD) is noninferior to open pancreaticoduodenectomy (OPD) in terms of clinically relevant pancreatic fistula.

In their study of 2,661 OPDs performed by 48 surgeons at 16 high-volume academic centers and 185 RPDs performed by three surgeons at the University of Pittsburgh Medical Center, the robotic procedure was associated with a 60% reduced risk of pancreatic fistula, compared with undergoing OPD (p=0.002).

In the propensity score-matched analysis of 152 pairs of patients, however, there was no significant difference in pancreatic fistula occurrence; the RPD cohort had noninferior outcomes compared with the OPD cohort in terms of complications, readmissions, and mortality.

"It should be emphasized that the focused results for robotic pancreaticoduodenectomy in this study represent those achieved by perhaps the world's most recognized, experienced, and acclaimed group with this technique," Dr. Vollmer told Reuters Health by email.

"These were compared to those achieved by a more diverse and heterogeneous group of surgeons across the globe performing open pancreaticoduodenectomy. It may not accurately reflect the results that would be achieved if robotic PD is disseminated on a more generalized scale, by less experienced or invested robotic surgeons," he said.

Dr. Hong Jin Kim from the University of North Carolina School of Medicine, Chapel Hill, North Carolina, who coauthored an accompanying commentary, told Reuters Health by email, "Although progress has been made at high-volume centers, what determines surgical outcomes (clinically relevant fistula rates, overall complications, length of stay, readmission rates, and mortality) seem less influenced by the surgical approach."

"Once centers have passed the learning curve, RPD appears to offer equivalent surgical outcomes to traditional open approaches," he said. "As more centers offer this option, patients will need to be aware of these outcomes, and where their surgeon/institution lies on the learning curve. In addition, there are many questions that remain unanswered, including cancer outcomes and cost-value analyses."

Coauthor of the editorial Dr. David A. Kooby from Emory University School of Medicine in Atlanta, Georgia, added in an email, "As a proponent of minimally invasive (MIS) pancreatic resection, I think we are closer to proving that MIS distal pancreatectomy has real value over the open approach, but we are not quite there with pancreaticoduodenectomy - at least not for the general public. We may get there in time with more experience and proper dissemination of training. The value question will be a tough one to address."

Dr. Francesco Guerra from Careggi University Hospital, Florence, Italy, who recently reviewed the role of robotics in MIPD, said, "Notwithstanding promising initial results and all the potential advantages connected with laparoscopy over conventional surgery, the intervening years have seen minimally invasive PD failing to obtain wide acceptance in real clinical practice worldwide. What I consider a crucial point is the fact that currently, probably also thanks to the availability of new and more sophisticated technologies (i.e., robotics), MIPD seems gain more substantial diffusion."

"I believe that the development of a dedicated program for MIPD is essential to enhance perioperative and surgical care," Dr. Guerra, who was not involved in the new studies, told Reuters Health by email. "Only the presence of a significant caseload can allow for appropriate experiences and initial learning curve."

SOURCE: http://bit.ly/2hYH4yO, http://bit.ly/2i0Dur5, http://bit.ly/2iQjTLD and http://bit.ly/2hXV0Oe

JAMA Surg 2016.