Robert E. Glasgow

A hospital's annual rate of esophagectomy influences the operative mortality rate

The reported operative mortality rate for esophagectomy for malignancy ranges from 2% to 30%. The goal of this retrospective study was to evaluate the relationship between a hospital’s annual rate of esophagectomy for esophageal cancer and the clinical outcome of the operation. Discharge abstracts of 1561 patients who had undergone esophagectomy for malignancy at acute care hospitals in California from 1990 through 1994 were obtained from the Office of Statewide Health Plating and Development. The hospitals were grouped according to the number of esophagectomies performed during the S-year period, and a mortality rate was calculated for each group. Logistic regression analysis was used to determine the relationship between a hospital’s rate of esophagectomy and the mortality rate. Esophageal resections were performed in 273 hospitals. An average of two or fewer resections were performed annually in 88% of hospitals, which accounted for 50% of all patients treated. The mortality rate in hospitals with more than 30 esophagectomies for the S-year period was 4.8%, compared with 16% for hospitals with fewer than 30 esophagectomies. This could not be accounted for by other health variables affecting the patients’ risk for surgery. There was a striking correlation between a hospital’s frequency of esophagectomy and the outcome of this operation. The results support the proposition that high-risk general surgical procedures, such as esophagectomy for malignancy, should be restricted to hospitals that can exceed a yearly minimum experience.

Changing management of gallstone disease during pregnancy

AbstractBackground: Symptomatic gallstones may be problematic during pregnancy. The advisability of laparoscopic cholecystectomy (LC) is uncertain. The objective of this study is to define the natural history of gallstone disease during pregnancy and evaluate the safety of LC during pregnancy. Methods: Review of medical records of all pregnant patients with gallstone disease at the University of California, San Francisco, from 1980 to 1996. Results: Of approximately 29,750 deliveries, 47 (0.16%) patients were treated for gallstone disease, including biliary colic in 33, acute cholecystitis in 12, and pancreatitis in two. Conservative treatment was attempted in all patients but failed in 17 (36%) cases. Two patients required combined preterm Cesarean-section cholecystectomy and 10 required surgery in the early postpartum period for persistent symptoms. Seventeen patients required cholecystectomy during pregnancy for biliary colic (10), acute cholecystitis (six), and pancreatitis (one). Three patients were treated with open cholecystectomy. Fourteen patients underwent LC at a mean gestational age of 18.6 weeks, mean OR time of 74 min, and mean length of stay of 1.2 days. Hasson cannulation was utilized in 11 patients. Reduced-pressure pneumoperitoneum (6–10 mmHg) was used in seven patients. Prophylactic tocolytics were used in seven patients, with transient postoperative preterm labor observed in one. There were no open conversions, preterm deliveries, fetal loss, teratogenicity, or maternal morbidity. Conclusions: In past years, symptomatic gallstones during pregnancy were managed conservatively or with open cholecystectomy. LC is a feasible and safe method for treating severely symptomatic patients.

Safety and Timing of Nonobstetric Abdominal Surgery in Pregnancy

Background/Aims: Abdominal disorders occurring during pregnancy pose special difficulties in diagnosis and management to the obstetrician and surgeon. The advisability of nonobstetric abdominal surgery during pregnancy is uncertain. Our objective was to evaluate the safety and timing of abdominal surgery during pregnancy. Methods: We retrospectively reviewed 77 consecutive gravid patients undergoing nonobstetric abdominal surgery from 1989 to 1996 at an urban academic medical center and a large affiliated community teaching hospital. Medical records were evaluated for clinical presentation, perioperative management, preterm labor, and maternal and fetal morbidity and mortality. Results: The rate of nonobstetric abdominal surgery during pregnancy was 1 in every 527 births. Among the 77 patients, the indications for surgery were adnexal mass (42%), acute appendicitis (21%), gallstone disease (17%) and other (21%). There was no maternal or fetal loss or identifiable neonatal birth defect. Preterm labor occurred in 26% of the second-trimester patients and 82% of the third-trimester patients. Preterm labor was most common in patients with appendicitis and after adnexal surgery. Preterm delivery occurred in 16% of the patients, but appeared to be directly related to the abdominal surgery in only 5%. Conclusion: Surgery during the first or second trimester is not associated with significant preterm labor, fetal loss or risk of teratogenicity. Surgery during the third trimester is associated with preterm labor, but not fetal loss.

Preoperative determinants of an esophageal lengthening procedure in laparoscopic antireflux surgery

BACKGROUND In a minority of patients undergoing antireflux surgery, an esophageal lengthening procedure is required to reduce the gastroesophageal junction (GEJ) below the esophageal hiatus. We evaluated risk factors associated with an irreducible GEJ to identify clinical features that were predictive of the need for a Collis gastroplasty in patients undergoing laparoscopic antireflux surgery. METHODS Patients who required a Collis gastroplasty during a laparoscopic antireflux procedure (defined as the inability to reduce the GEJ > 2.5 cm below the esophageal hiatus despite extensive mobilization of the mediastinal esophagus) were compared to a random sample of patients who did not have a Collis gastroplasty. Predictors of the need for an esophageal lengthening procedure were identified using logistic regression modeling. Risks were expressed as odds ratios (OR) and 95% confidence intervals (CI). RESULTS Twenty patients who had a Collis gastroplasty were compared to 133 patients who had adequate esophageal length. The presence of a stricture (OR 3.0; 95% CI 1.0, 9.7), paraesophageal hernia (OR 3.5; 95% CI 1.3, 9.6), Barretts esophagus (OR 3.7, 95% CI 1.3, 10.7), and re-do antireflux surgery (OR 6.4; 95% CI 2.0, 20.7) were associated with the need for gastroplasty. Patients with none of these factors were extremely unlikely to require a gastroplasty (OR 0.08; 95% CI 0.02, 0.34). CONCLUSION Patients undergoing laparoscopic antireflux surgery who are at high risk of needing an esophageal lengthening procedure can be easily identified preoperatively using simple clinical characteristics.

Lessons learned from laparoscopic treatment of gastric and gastroesophageal junction stromal cell tumors

BackgroundStromal cell tumors of the gastric and gastroesophageal junction are rare neoplasms that traditionally have been resected for negative margins using an open approach. This study aimed to evaluate the efficacy laparoscopic resection of gastric and gastroesophageal stromal cell tumors and the lessons learned from experience with this method.MethodsThis retrospective review evaluated all patients who underwent laparoscopic resection of gastric or esophageal stromal cell tumors at a tertiary referral center between December 2002 and March 2005. Medical records were reviewed with regard to patient demographics, preoperative evaluation, operative approach, tumor location and pathology, length of operation, complications, and length of hospital stay.ResultsA total of 12 consecutive patients with a mean age of 55 ± 5.9 years were treated. Preoperative endoscopic ultrasound (EUS) was performed for 11 of 12 patients with a diagnostic accuracy of 100%, whereas EUS-guided fine-needle aspiration was performed for 10 of 12 patients with a diagnostic accuracy of 50%. Four patients with symptomatic gastroesophageal junction leiomyomas were treated with enucleation and Nissen fundoplication. Eight patients were treated with laparoscopic wedge resection of gastric lesions. Complete R0 resection was achieved for all the patients undergoing laparoscopic resection. Intraoperative endoscopy was performed for four patients and resulted in shorter operative times. The average operative time for this entire series was 169 ± 17 min: 199 ± 24 min for the first six cases and 138 ± 19 min for the last six cases. The median hospital length of stay was 2 days. One patient with esophageal leiomyoma had persistent dysphagia at the 12-month follow-up assessment. There were no other complications and no deaths in this series of patients.ConclusionsLaparoscopic resection of gastric and gastroesophageal junction stromal cell tumors may be performed safely with low patient morbidity. This approach can achieve adequate surgical margins and lead to short hospital stays. Improvements in the technique have led to shorter operative times.

Anti-inflammatory effects of PPAR-γ agonists directly correlate with PPAR-γ expression during acute pancreatitis

Peroxisome proliferator-activated receptors (PPARs) are ligand-inducible transcription factors that regulate cellular energy and lipid metabolism. PPAR-γ agonists also have potent anti-inflammatory properties through down-regulation of early inflammatory response genes. The role of PPAR-γ in acute pancreatitis has not been adequately examined. In this study, we determined the effect of PPAR-γ agonists on the severity of pancreatitis and sought to correlate PPAR-γ expression in pancreatic acinar cells and the severity of acute pancreatitis in vivo. Acute pancreatitis was induced in mice by hyperstimulation with the cholecystokinin analog, cerulein. PPAR-γ agonists were administered by intraperitoneal injection 15–30 minutes before induction of pancreatitis (pretreatment) or at various times after induction of pancreatitis (treatment). Pancreata and serum were harvested over the course of 24 hours. Serum amylase activity and glucose levels were measured. Pancreata were used for histological evaluation as well as protein and mRNA analysis. Pretreatment of mice with the PPAR-γ agonists 15-deoxy-Δ12, 14-prostaglandin J2, or troglitazone significantly reduced the severity of pancreatitis in a dose-dependent manner. This reduction was indicated by reduced serum amylase activity and histological damage (leukocyte infiltration, vacuolization, and necrosis). Although cerulein decreased PPAR-γ expression in the pancreas, pretreatment with agonists maintained PPAR-γ expression early in acute pancreatitis. The expression of PPAR-γ inversely correlated with pancreatitis severity and expression of the proinflammatory cytokines, interleukin-6, and tumor necrosis factor-α. Treatment with troglitazone after the induction of pancreatitis reduced serum amylase activity. The results suggest that PPAR-γ plays a direct role in the inflammatory cascade during the early events of acute pancreatitis. Our data are the first to demonstrate that PPAR-γ agonists represent a promising therapeutic strategy for acute pancreatitis.

Laparoscopic radiofrequency ablation and hepatic artery infusion pump placement in the evolving treatment of colorectal hepatic metastases

Background: Laparoscopic radiofrequency ablation (LRFA) and laparoscopic hepatic artery infusion pump (LHAIP) placement are new treatment options for patients with colorectal liver metastases. This study investigates the selection criteria, safety, efficacy, and preliminary outcomes of patients treated with LRFA and LHAIP placement. Methods: Fourty five patients with colorectal metastases confined to the liver, 37 of whom had failed systemic chemotherapy, were treated with LRFA and/or LHAIP between September 1996 and December 2001. Treatment selection was individualized, based on each patients general health, liver function, and tumor size, number, location, and distribution. Results: Twenty patients (44%) had LRFA alone, 10 (22%) had LHAIP placement alone, and 15 (33%) patients had combined LRFA and LHAIP therapy. The LRFA group had a significantly shorter mean operative time and blood loss (p <0.05), but hospital stays were similar when compared to patients receiving LRFA + LHAIP or LHAIP alone. Tumor characteristics were worse in both LHAIP groups, with a higher incidence of tumors ?4 cm, major vascular involvement, diffuse tumor pattern, bilobar distribution, and involvement of more than three segments. During a mean follow-up period of 11.5 ± 7.8 months (range, 1–38), the actuarial survival was 70%, 67%, and 50% for LRFA, LRFA + LHAIP, and LHAIP, respectively. LHAIP only patients had the shortest estimated mean survival time of the three groups by Kaplan-Meier survival curves (p = 0.001). Conclusion: LRFA and/or LHAIP placement are safe and feasible treatment options for the treatment of colorectal hepatic metastases. The choice of treatment for patients should be based primarily on tumor characteristics. Long-term studies, which will elucidate the role of these evolving treatments, are now under way.

Roux‐en‐Y drainage of a pancreatic fistula for disconnected pancreatic duct syndrome after acute necrotizing pancreatitis

BACKGROUND After acute necrotizing pancreatitis (ANP), a pancreatic fistula may occur from disconnected pancreatic duct syndrome (DPDS) where a segment of the pancreas is no longer in continuity with the main pancreatic duct. AIM To study the outcome of patients treated using Roux-Y pancreatic fistula tract-jejunostomy for DPDS after ANP. METHODS Between 2002 and 2011, patients treated for DPDS in the setting of endoscopic retrograde cholangiopancreatography (ERCP) or magnetic resonance cholangiopanreatography (MRCP) documented main pancreatic duct disruption with Roux-Y pancreatic fistula tract-jejunostomy. RESULTS In all, seven patients with DPDS were treated. The median age was 62 years (range 49-78) and five were men. The cause of ANP was gallstones (2), alcohol (1), ERCP (1) and idiopathic (3). Pancreatic necrosectomy was done in six patients. Time from onset of pancreatitis to fistula drainage was 270 days (164-365). Pancreatic fistulae arose from DPDS in the head/neck (4) and body/tail (3). Patients had a median fistula output of 140 ml (100-200) per day before surgery. The median operative time was 142 min (75-367) and estimated blood loss was 150 ml (25 to 500). Patients began an oral diet on post-operative day 4 (3-6) and were hospitalized for a median of 7 days (5-12). The median follow-up was 264 days (29-740). Subsequently, one patient required a distal pancreatectomy. After surgery, three patients required oral hypoglycaemics. No patient developed pancreatic exocrine insufficiency. CONCLUSION Internal surgical drainage using Roux-en-Y pancreatic fistula tract-jejunostomy is a safe and definitive treatment for patients with DPDS.

The benefits of a dedicated minimally invasive surgery program to academic general surgery practice

In 2001, a dedicated minimally invasive surgery (MIS) program was established at a large university hospital. Changes included improvement and standardization of equipment and instruments, patient care protocols, standardized orders, and staff education. The aim of this study was to evaluate the impact of this program on an academic surgery practice. From January 1999 through October 2003, hospital and departmental databases were reviewed for all records pertaining to general surgery cases. Data trends were analyzed by regression analysis and are expressed as mean +- SEM. In 1999, 15.0 +- 0.1% of all general surgery cases were MIS cases compared with 30.2 +- 0.1% in 2003 (P < 0.0001). During this period, the number of patients requiring conversion from a laparoscopic to an open approach decreased from 14.4% to 4.0% (P < 0.0007). In 1999, 30% of appendectomies were laparoscopic, compared with 92% in 2003 (P < 0.0001). This increase in the rate of laparoscopic appendectomy resulted in a decrease in average length of hospital stay for all patients with acute appendicitis, from 5.5 +- 1.0 days in 1999 to 2.7 +- 0.2 days in 2003 (P < 0.0001), and a decrease in total hospital cost per case, from

Does Travel Distance Influence Length of Stay in Elective Colorectal Surgery

6569 +- 400 in 1999 to