TIPS in Florida: is its application a result of evidence-based medicine?

BACKGROUND: The typical resident in surgery in the US will not care for a patient with advanced portal hypertension and will not participate in a portacaval shunt. The aim of this study is to compare the number of transjugular intrahepatic portasystemic stent shunts (TIPS) with the number of surgical shunts undertaken in the State of Florida and to assess whether these numbers are consistent with today's evidence-based medicine. METHODS: We examined the database of the Agency for Health Care Administration of the State of Florida from January 1, 2002, through September 30, 2005, for "intraabdominal venous shunt" (ICD-9 code, 39.1). Data collected include "case mix," "case severity," length of stay, total gross charges, and discharge status. Conclusions about longterm survival from a prospective randomized clinical trial comparing TIPS to surgical shunting were applied to this dataset to determine if the relative frequency of TIPS application in Florida was supported by evidence-based medicine. RESULTS: TIPS was undertaken more than 12 times as often as surgical shunting (860 patients versus 70 patients). After TIPS versus surgical shunts, average length of stay and hospital charges were less, but case mix, case severity, and in-hospital mortality (11.4% for each) were not different. Applying survival data from a randomized trial comparing TIPS with surgical shunting to the State of Florida database, 129 more people (p < 0.0001) would be alive at 2 years and 137 more (p < 0.0001) would be alive at 5 years after shunting if surgical shunts had been used in lieu of TIPS. CONCLUSIONS: TIPS leads to shorter hospitalizations and reduced hospital charges and is applied in numbers much greater than surgical shunts, despite evidence that suggests inferior longterm efficacy and survival. Current application of TIPS is not a result of evidence-based medicine, and application of surgical shunting is encouraged.

Stage does not predict survival after resection of hilar cholangiocarcinomas promoting an aggressive operative approach.

INTRODUCTION: Staging systems have been developed to predict survival after resection of hilar cholangiocarcinoma. Notably, they have not been validated nor compared for relative predictive ability. METHODS: Forty-two patients underwent resection of hilar cholangiocarcinoma and have been followed through a prospectively collected database. The tumors were staged using the Bismuth-Corlette, Blumgart, and American Joint Committee on Cancer (AJCC) systems, and a significant relationship with survival was sought. RESULTS: Eleven patients were treated by extrahepatic biliary resection alone, while 31 required extrahepatic biliary resections with in-continuity hepatic resections. All patients underwent adjuvant therapy. To date, 30 patients have died with a mean survival time of 30 months +/- 35.0 (SD). Twelve patients are alive with a mean survival of 90 months +/- 61.8. By regression analysis, none of the staging systems had a significant relationship with survival (Bismuth: P = .64; Blumgart: P = .66; AJCC: P = .31). CONCLUSIONS: Most patients with hilar cholangiocarcinoma require in-continuity hepatic resections. Survival after resection promotes an aggressive approach, with cure in as many as 30%. Staging systems should not impact the decision to operate or postoperative management, as all tumors should be aggressively resected and all patients should receive adjuvant treatment.

H-graft portacaval shunts versus TIPS: ten-year follow-up of a randomized trial with comparison to predicted survivals.

OBJECTIVE: To report long-term outcome of patients undergoing prosthetic 8-mm H-graft portacaval shunts (HGPCS) or TIPS and to compare actual with predicted survival data. METHODS: A randomized trial comparing TIPS to HGPCS for bleeding varices began in 1993. Predicted survival was determined using MELD (Model for End-stage Liver Disease). RESULTS: Patients undergoing TIPS (N = 66) or HGPCS (N = 66) were very similar by Child's class and MELD scores and predicted survival. After TIPS (P = 0.01) and HGPCS (P = 0.001), actual survival was superior to predicted survival. Through 24 months, actual survival after HGPCS was superior to actual survival after TIPS (P = 0.04). Compared with TIPS, survival was superior after HGPCS for patients of Child's class A and B (P = 0.07) and with MELD scores less than 13 (P = 0.04) with follow-up at 5 to 10 years. Shunt failure was less following HGPCS (P < 0.01). CONCLUSIONS: Predicted survival data for patients undergoing TIPS or HGPCS confirms an unbiased randomization. Actual survival following TIPS or HGPCS was superior to predicted survival. Shunt failure favored HGPCS, as did survival after shunting, particularly for the first few years after shunting and for patients of Child's class A or B or with MELD scores less than 13. This trial irrefutably establishes a role for surgical shunting, particularly HGPCS.

TIPS versus peritoneovenous shunt in the treatment of medically intractable ascites: a prospective randomized trial.

OBJECTIVE: We undertook a prospective randomized clinical trial comparing TIPS to peritoneovenous (PV) shunts in the treatment of medically intractable ascites to establish relative efficacy and morbidity, and thereby superiority, between these shunts. METHODS: Thirty-two patients were prospectively randomized to undergo TIPS or peritoneovenous (Denver) shunts. All patients had failed medical therapy. RESULTS: After TIPS versus peritoneovenous shunts, median (mean +/- SD) duration of shunt patency was similar: 4.4 months (6 +/- 6.6 months) versus 4.0 months (5 +/- 4.6 months). Assisted shunt patency was longer after TIPS: 31.1 months (41 +/- 25.9 months) versus 13.1 months (19 +/- 17.3 months) (P < 0.01, Wilcoxon test). Ultimately, after TIPS 19% of patients had irreversible shunt occlusion versus 38% of patients after peritoneovenous shunts. Survival after TIPS was 28.7 months (41 +/- 28.7 months) versus 16.1 months (28 +/- 29.7 months) after peritoneovenous shunts. Control of ascites was achieved sooner after peritoneovenous shunts than after TIPS (73% vs. 46% after 1 month), but longer-term efficacy favored TIPS (eg, 85% vs. 40% at 3 years). CONCLUSION: TIPS and peritoneovenous shunts treat medically intractable ascites. Absence of ascites after either is uncommon. PV shunts control ascites sooner, although TIPS provides better long-term efficacy. After either shunt, numerous interventions are required to assist patency. Assisted shunt patency is better after TIPS. Treating medically refractory ascites with TIPS risks early shunt-related mortality for prospects of longer survival with ascites control. This study promotes the application of TIPS for medically intractable ascites if patients undergoing TIPS have prospects beyond short-term survival.

Portal blood flow, effective hepatic blood flow, and outcome after partial portal decompression.

What a great time to practice Surgery! Things we only dreamed about 20 years ago are finding their way into clinical care and greater advances are around the corner. As practicing surgeons, we have prospered by being raised on the shoulders of giants in Surgery that trained and inspired us. Dr. David B. Skinner was such a giant.

Reoperative fundoplications are effective treatment for dysphagia and recurrent gastroesophageal reflux.

With wide application of antireflux surgery, reoperations for failed fundoplications are increasingly seen. This study was undertaken to document outcomes after reoperative fundoplications. Sixty-four patients, 26 men and 38 women, of average age 55 years+/-15.6 (SD), underwent reoperative antireflux surgery between 1992 and 2003. Fundoplication prior to reoperation had been undertaken via celiotomy in 27 and laparoscopically in 37. Both before and after reoperative antireflux surgery, patients scored their reflux and dysphagia on a Likert Scale (0 = none, 10 = continuous). Reoperation was undertaken because of dysphagia in 16 per cent, recurrent reflux in 52 per cent (median DeMeester Score 52), or both in 27 per cent. Failure leading to reoperation was due to hiatal failure in 28 per cent, wrap failure in 19 per cent, both in 33 per cent, and slipped Nissen fundoplication in 20 per cent. Laparoscopic reoperations were completed in 49 of 54 patients (91%); 15 had reoperations undertaken via celiotomy. Eighty-eight per cent of reoperations were Nissen fundoplications. With reoperation, Dysphagia Scores improved from 9.5+/-0.7 to 2.6+/-2.8, and Reflux Scores improved from 9.1+/-1.4 to 1.8+/-2.7. Seventy-nine per cent of patients with reflux prior to reoperation, 100 per cent with dysphagia, and 74 per cent with both noted excellent or good outcomes after reoperation. We conclude that failure after fundoplication occurs. Reoperations reduce the severity of dysphagia and reflux, thus salvaging excellent and good outcomes in most. Laparoscopic reoperations are generally possible. Reoperative fundoplications are effective treatment for dysphagia and recurrent gastroesophageal reflux, and their application is encouraged.