Complex Pancreatic Surgery: Safety and Feasibility in the Community Setting
- First Online:
- Cite this article as:
- Chamberlain, R.S., Tichauer, M., Klaassen, Z. et al. J Gastrointest Surg (2011) 15: 184. doi:10.1007/s11605-010-1305-9
- 68 Views
Advances in technology, innovative surgical procedures, and enhanced perioperative care have allowed more patients to be considered for complex pancreatic surgery. Published reports on the outcomes of pancreatic surgery performed at high volume tertiary referral centers have yielded excellent results. However, similar outcome and safety data from community hospitals is limited.
Material and Methods
Consecutive complex pancreatic surgery performed by a single surgeon from December 2004 to December 2009 formed the study group. Factors analyzed included patient demographics, operative procedure, operative time, length of hospital stay, pathology, and 30-day morbidity and mortality.
One hundred and nine consecutive patients underwent pancreatic surgery, with a mean patient age of 62.4 ± 15.2 years. Eighty-three patients (76.1%) underwent definitive surgical procedure and 26 patients (23.9%) had palliative bypass after failed palliative biliary stenting. The mean operative time was 229 ± 109 min, the mean length of stay was 8.6 ± 6.5 days and 24 (22.0%) patients had surgical complications.
Complex pancreatic surgery can be performed safely at high-volume tertiary community hospitals with excellent outcomes comparable to tertiary academic centers. In the ongoing debate about the need for mandatory referral of complex surgical procedures, tertiary community hospitals with well-determined outcomes should be included.
KeywordsPancreatic surgeryCommunity hospitalSurgical outcomes
The natural history of pancreatic cancer, including its clinical quiescence and aggressive molecular epidemiology, plays an important role in its lethality. Surgical resection is the only curative option for patients with pancreatic cancer, yet it is associated with a high morbidity rate and disappointing 5-year survival rates of 10–29%.1 Up to 85% of patients with pancreatic cancer are unresectable at the time of diagnosis, and the significant morbidity and limited survival rates associated with complex pancreatic surgery has led some to question the rationale of radical operations for pancreatic cancer.2–8 In view of these facts, and given the complexity of surgical procedures required to treat pancreatic cancer, regionalization of complex pancreatic procedures such as pancreaticoduodenectomy (PD), has been proposed in an effort to optimize patient outcomes.9–15 Previous reports have suggested that hospital procedure volume is associated with superior clinical outcomes in patients undergoing complex operations, such as pancreatectomy and esophagectomy, as well as less complex procedures such as lumpectomy and colectomy.12,16–18 Yet, contrary to these claims, others have reported no association between hospital volume and clinical outcome.19 For example, Enzinger et al.19 examined the relationship between hospital volume and clinical outcomes following gastrectomy among 306 US hospitals, and found no significant differences in 5-year overall survival or disease-free survival (short-term perioperative morbidity and mortality was not addressed) between low, moderate, and high volume hospitals.
Over the last two decades, significant advances in perioperative evaluation and patient selection, improved surgical techniques with combined regional and general anesthesia, and standard perioperative care and management have significantly reduced the mortality associated with pancreatic resection.1 While sporadic reports of excellent clinical outcomes for PD performed at low volume hospitals have emerged, they are few in number.9,19,20 Focusing exclusively on hospital volume, few studies have addressed the question as to whether patient outcomes are more dependent upon the surgeon’s experience/training or the hospital setting in which they are performed. In order to address this question, we sought to analyze our early experience with complex pancreatic surgery performed by a single fellowship-trained hepatobiliary and pancreatic surgeon after the establishment of a community-based hepatobiliary and pancreatic surgery Center of Excellence.
Materials and Methods
Over 350 patients with complex pancreatic problems were evaluated by the surgical oncology service at the Saint Barnabas Medical Center (SBMC), Livingston, New Jersey, from December 2004 and December 2009. Patients with hepatobiliary diseases were excluded from this analysis. One hundred and nine consecutive patients requiring pancreatic surgery formed the study group. All patients who underwent pancreatic resection had preoperative imaging with triple phase contrast-enhanced computed tomography or magnetic resonance imaging, along with positron emission tomography scans when appropriate. Diagnostic laparoscopy was performed in all malignant cases in order to exclude the presence of occult or disseminated intra abdominal disease. Intraoperative ultrasound was used selectively.
SBMC is a 641-bed tertiary-care referral hospital for a network of six community hospitals. The hospital performs more than 29,000 surgical cases per year and has an Accreditation Council for Graduate Medical Education-approved general surgery residency program. SBMC offers a comprehensive cancer program including one of the largest radiation oncology programs in northern New Jersey, treating more than 1,000 patients yearly. Furthermore, the facility operates a highly specialized Gastrointestinal Cancer Program, offering advanced therapeutic alternatives including: selective internal radiation therapy for inoperable liver cancer, radiofrequency and microwave ablation, heated intraperitoneal chemotherapy for advanced tumors of the peritoneal cavity, and robotic surgery. All cases are subject to multi-discipline review and analysis prior to initiation of therapy.
Data from patients who underwent surgical intervention were collected prospectively from medical records, outpatient charts, lab records, and pathology reports and entered into a Microsoft Excel database (Microsoft Corporation™, Redmond, WA, USA). Pre-operative evaluation included physical examination, medical/cardiac clearance, pertinent laboratory and imaging studies, and tumor markers. The operative procedure, operative time, estimated blood loss, length of hospital stay, and 30-day morbidity and mortality were analyzed. Procedures involving pancreatic resection had surgical margin analysis reported as R0 (negative), R1 (microscopically positive), and R2 (grossly positive). The association of continuous variables was statistically analyzed by the Student t test.
Comordities among 109 patients undergoing complex pancreatic procedures
All Patients (N = 109)
PD Patients (N = 40)
Location of the resected pancreatic pathology
Histopathology of pancreatic resections
Focal sclerosing pancreatitis
Pseudo papillary tumor
Mucinous carcinoma peritonei
Ruptured splenic artery aneurysm
Ulcer and stricture
Pancreatic procedures performed in 109 consecutive patients
Palliative bypass and biopsy
Diagnostic laparoscopy and biopsy
Internal Pseudocyst enteric drainage
Pancreas sparing duodenectomy
Postoperative complications in 109 patients undergoing pancreatic surgery
All patients (N = 109)
PD patients (N = 40)
Gastric outlet obstruction
Gastric outlet obstruction
The converse relationship between hospital volume and postoperative mortality among patients undergoing complex surgical procedures, including pancreatic resection, have been extensively examined and documented.9,21,22 The Donabedian Model is a framework for quality-of-care,23 developed to define, measure, and categorize quality in healthcare delivery. This model includes: structure (where the care is delivered), process (evaluating medical practice), and outcome (impact of care on health).23 The outcome measure, which reflects how a unique patient fares following some form of medical intervention is the most difficult to measure. Although hospital volume is simple to measure and may be associated with improved patient outcomes for pancreatic resection and other procedures,21–26 it may not be the sole determinant of outcome.
Many high-volume pancreatic centers throughout the country have reported substantial reduction in hospital mortality over the last 10 years,15,26 which is attributable to a variety of factors. Factors most often cited include the volume of procedures performed, perioperative care and nutrition, and strict adherence to critical care and perioperative pathways.12,22,24 Mukherjee et al.15 evaluated the impact of the UK Cancer Outcome Guidelines (COG) among 140 patients who underwent PD between 1999 and 2006. The COG was introduced in the UK in 1999 and was subsequently implemented in 2003. The institution of these guidelines led to the centralization of cancer services, including upper gastrointestinal cancer services and was restricted to tertiary referral centers.27 In the pre-COG era (1999–2002) there were 41 PD performed compared to 99 performed in the post-COG era (2003–2006).15 The authors reported a trend towards decreased mortality (9.7–5.0%, p < 0.448) and morbidity (41.6–35.3%, p < 0.565), concluding that the COG implementation lead to increased PD volume, higher staffing levels, and a trend towards better outcomes.15
Birkmeyer et al.11 evaluated outcomes based on hospital volume among Medicare patients undergoing PD for pancreatic cancer and reported that more than 50% of these patients received care at hospitals performing fewer than two procedures per year. They evaluated outcomes based on four hospital volume categories, which included very low (<1 case per year), low (1–1.99 cases per year), medium (2–4.99 cases per year), and high volume hospitals (>5 cases per year). These authors reported that in-hospital mortality rates at low and very low-volume hospitals were three- to fourfold higher than at high-volume hospitals (12% and 16%, respectively, vs. 4%, p < 0.001) and concluded that hospital volume is an important factor in surgical outcomes for PD.11
In an attempt to identify factors affecting outcomes after complex pancreatic surgery as well as to trace the evolution of a procedure, Cameron et al.26 reported their 30-year experience involving 1,000 consecutive PD at the Johns Hopkins Hospital. They noted that advancements over the last three decades in imaging, intraoperative anesthesia, and peri-operative care were major factors resulting in improved patient outcomes and decreased perioperative morbidity and mortality. These authors also noted a 50% reduction in their operative time and a 30% reduction in estimated operative blood loss over the study period.26
Meguid et al.28 conducted a retrospective analysis of 7,558 patients who underwent pancreatic resection from the Nationwide Inpatient Sample (20% sample of patients in the US from 1998 to 2003). This study reported a median annual institution pancreatic resection volume of 15 cases, mean in-hospital mortality of 7.6% and noted that based on a goodness-of-fit analysis, a minimum of 19 pancreatic resections per year is required to qualify as a high volume center.28 However, they concluded that a volume cutoff for pancreatic surgery was arbitrary, as a difference in perioperative mortality was observed regardless of the volume cutoff used.28 Riall et al.29 reviewed the Texas Hospital Inpatient Discharge Database from 1999 to 2005 and identified 12 high-volume hospitals for pancreatic resection (>11 cases/year). Among these hospitals, there was significant variability in mortality, duration of stay, need for ongoing nursing care, operation within 24 h of admission, and hospital cost per patient visit.29 They concluded that significant variability in outcomes occurred even among high-volume providers and reasoned that individual hospital differences likely accounted for much of the variability not explained by hospital volume.29
In addition to variability in outcomes at high-volume centers, a number of studies from low-volume and community-based hospitals have reported excellent outcomes. Schell et al.25 performed a comparative outcome analysis of 369 patients who underwent PD at the University of California, San Francisco affiliated hospitals between October 1989 and June 2003. They noted that while high-volume centers did attain excellent surgical outcomes, smaller and lower-volume hospitals achieved similar surgical outcomes provided they import expertise and implement care pathways. The low-volume hospital group consisted of community-based hospitals and county general hospitals that performed an average of one PD per year, and a Veterans Affairs Medical Center that performed three PD per year. The high volume tertiary hospitals averaged 23 PD per year. They found no difference in regards to morbidity and complications between the groups (high volume, 58.8% vs. low volume, 60.3%; p < 0.579). Moreover, the perioperative mortality rates for patients undergoing PD were approximately 4% in both groups, with no significant difference in 5-year survival rates (high-volume hospital 19% versus 18.3% for low-volume hospital group, p < 0.096).25
Comparative analysis of major tertiary academic centers and community hospital complex pancreatic surgical series
Chamberlain et al. 
Hoshal et al.30
Mukherjee et al.15
Schell et al.25
Cameron et al.26
Number of Patients
62.4 ± 15.2
242 ± 272
1167 ± 1411
Mean Operative Time (min)
229 ± 109
402 ± 120
Mean LOS (days)
8.6 ± 6.5
16.1 ± 23.5
Although hospital volume is easy to measure, it is not reliable as the sole measure of quality or outcomes after pancreatic surgery. The idea that volume alone can be a proxy to define centers of surgical excellence is an imperfect rationale. Despite emerging reports of excellent surgical outcomes for many complex procedures performed at community based medical centers, a movement towards establishing volume-based referral centers for certain surgical (including pancreatic) procedures continue to be pushed.20,31,32 The Leapfrog Group, comprised of healthcare purchasers and providers representing 33 million patients, is perhaps the most vocal group promoting volume-based referral. In order to concentrate patient care in high volume hospitals, the Leapfrog initiative has set annual hospital volume thresholds for a number of different surgical procedures including: coronary artery bypass graft (450 cases), coronary angioplasties (400 cases), abdominal aortic aneurysm repairs (50 cases), aortic valve replacements (120 cases), esophagectomies (13 cases), pancreatic resections (11 cases), and bariatric surgeries (125 cases).32 Whether there should be regionalization of major hepatobiliary–pancreatic procedures to academic centers of excellence is being similarly debated. Tertiary community based hospitals with excellent results should be included in any proposed mandatory referral system. To date, there remains no optimum combination of number of procedures, years of training, or other factors that assure good outcomes in surgery. Proper patient selection, in combination with a competent surgeon with adequate training, excellent critical care, and interdisciplinary support is the only means of optimizing patient outcome regardless of the hospital setting or procedure volume.
We thank N. Babel MD, R. Singh BSc, and J. Serfin BSc for their efforts with data collection and analysis.
Conflict of Interest
All the authors declare that there are no conflicts of interest and have accepted no financial sponsorship in producing and presenting this manuscript. Each author listed is in agreement with the content of the manuscript.