A nutrition support team led by general surgeons decreases inappropriate use of total parenteral nutrition on a surgical service

The purpose of this study was to decrease the number of inappropriate orders for total parenteral nutrition (TPN) in surgical patients. From February 1999 through November 2000 and between July 2001 and June 2002, the surgeon-guided adult nutrition support team (NST) at a university hospital monitored new TPN orders for appropriateness and specific indication. In April 1999, the NST was given authority to discontinue inappropriate TPN orders. Indications, based on the American Society for Parenteral and Enteral Nutrition (ASPEN) standards, included short gut, severe pancreatitis, severe malnutrition/catabolism with inability to enterally feed > or =5 days, inability to enterally feed >50 per cent of nutritional needs > or =9 days, enterocutaneous fistula, intra-abdominal leak, bowel obstruction, chylothorax, ischemic bowel, hemodynamic instability, massive gastrointestinal bleed, and lack of abdominal wall integrity. The number of inappropriate TPN orders declined from 62/194 (32.0%) in the first 11 months of the study to 22/168 (13.1%) in the second 11 months (P < 0.0001). This number further declined to 17/215 (7.9%) in the final 12 months of data collection, but compared to the second 11 months, this decrease was not statistically significant (P = 0.1347). The involvement of a surgical NST was associated with a reduction in inappropriate TPN orders without a change in overall use.

Outpatient laparoscopic cholecystectomy: patient outcomes after implementation of a clinical pathway

OBJECTIVE

To determine the success of a clinical pathway for outpatient laparoscopic cholecystectomy (LC) in an academic health center, and to assess the impact of pathway implementation on same-day discharge rates, safety, patient satisfaction, and resource utilization.

SUMMARY BACKGROUND DATA

Laparoscopic cholecystectomy is reported to be safe for patients and acceptable as an outpatient procedure. Whether this experience can be translated to an academic health center or larger hospital is uncertain. Clinical pathways guide the care of specific patient populations with the goal of enhancing patient care while optimizing resource utilization. The effectiveness of these pathways in achieving their goals is not well studied.

METHODS

During a 12-month period beginning April 1, 1999, all patients eligible for an elective LC (n = 177) participated in a clinical pathway developed to transition LC to an outpatient procedure. These were compared with all patients undergoing elective LC (n = 208) in the 15 months immediately before pathway implementation. Successful same-day discharges, reasons for postoperative admission, readmission rates, complications, deaths, and patient satisfaction were compared. Average length of stay and total hospital costs were calculated and compared.

RESULTS

After pathway implementation, the proportion of same-day discharges increased significantly, from 21% to 72%. Unplanned postoperative admissions decreased as experience with the pathway increased. Patient characteristics, need for readmission, complications, and deaths were not different between the groups. Patients surveyed were highly satisfied with their care. Resource utilization declined, resulting in more available inpatient beds and substantial cost savings.

CONCLUSIONS

Implementation of a clinical pathway for outpatient LC was successful, safe, and satisfying for patients. Converting LC to an outpatient procedure resulted in a significant reduction in medical resource use, including a decreased length of stay and total cost of care.

Laparoscopic bariatric surgery

At present, laparoscopic bariatric surgery is a controversial topic among bariatric and laparoscopic surgeons. Although difficult to perform, the traditional procedures used for treating severe obesity are now being performed successfully using a laparoscopic approach. In addition, a new procedure, adjustable gastric banding, has been used in Europe; however, it lacks US Food and Drug Administration approval and adequate scientific follow-up data about long-term effectiveness. Appropriate patient selection and adherence to the principles that have been learned by experience through open bariatric surgery remain of paramount importance for the success of laparoscopic bariatric operations.

Postprandial response of gastric slow waves: correlation of serosal recordings with the electrogastrogram

Controversial interpretations have been given to the postprandial increase in the dominant power (amplitude) of the electrogastrogram (EGG). The aim of this study was to find an appropriate interpretation of the postprandial EGG power changes. Simultaneous serosal and cutaneous recordings of gastric myoelectrical activity were made in 11 patients with gastroparesis in the fasting state and after the ingestion of 8 oz of water. The dominant frequency and corresponding power of the recording before and after water were computed using the power spectral analysis method. It was found that the dominant frequency of the EGG was the same as that of the serosal recording in 10 patients. One patient showed a substantial amount of dysrhythmia and no obvious dominant frequency was noted. A decrease in the dominant frequency was found in these 10 patients after the ingestion of water. Tachygastria of higher than 4 cycles/min was observed in one of 11 patients both in the prewater and postwater states. Consistent changes in amplitude after a drink of water were noted in both serosal recording and EGG. Statistical analysis demonstrated that the dominant power change after water computed from the EGG was correlated with that observed in the serosal recording (r = 0.757, P = 0.007). In conclusion, exogenous stimulation, such as ingestion of water, may change the amplitude of the gastric slow wave and this change is reflected in the EGG, suggesting that the change of the slow-wave amplitude is an important contributing factor to the postprandial change in the EGG dominant power.

Evidence for a peripheral mechanism of esophagocrural diaphragm inhibitory reflex in cats

The esophagogastric junction (EGJ) is guarded by two sphincters, a smooth muscle lower esophageal sphincter (LES) and a skeletal muscle crural diaphragm. These two sphincters relax simultaneously under certain physiological conditions, i.e., swallowing, belching, vomiting, transient LES relaxation, and esophageal distension. Esophageal distension-induced crural diaphragm relaxation is mediated through vagal afferents that are thought to exert inhibitory influence on the central mechanism (brain stem) of crural diaphragm contraction. We conducted studies in 10 cats to determine whether a mechanism of crural diaphragm relaxation was located at the level of the neuromuscular junction and/or muscle. Stimulation of the crural diaphragm neuromuscular junction through 1) the electrodes implanted in the muscle and 2) the bilateral phrenic nerve resulted in an increase in EGJ pressure. Nicotinic receptor blockade (pancuronium, 0.2 mg/kg) abolished the EGJ pressure increase caused by electrical stimulation of the neuromuscular junction. Esophageal distension and bolus-induced secondary esophageal peristalsis caused relaxation of the EGJ during the stimulation of the neuromuscular junction. Bilateral phrenicotomy and vagotomy had no influence on this relaxation. These data suggest the existence of a peripheral mechanism of crural diaphragm inhibition. This peripheral inhibitory mechanism may reside at the level of either the neuromuscular junction or the skeletal muscle.

The accuracy of laparoscopic ultrasound in the detection of colorectal cancer liver metastases

BACKGROUND

The most sensitive and specific method of detecting colorectal cancer hepatic metastases has been shown to be a combination of careful intraoperative palpation and intraoperative ultrasound. Although there has been growing interest in laparoscopic surgical therapy for colorectal cancer, the ability of this technique to adequately evaluate the liver for small metastases has been unknown. This study was undertaken to compare laparoscopic liver ultrasound to the gold standard of open palpation and intraoperative ultrasound in detecting hepatic metastases from colorectal cancer.

METHODS

A preliminary animal model was first performed in adult pigs. Eighteen liver "lesions" were created with chlorhexidine gluconate under laparoscopic guidance. A blinded surgeon then performed laparoscopic liver ultrasound followed by open ultrasound and palpation, comparing the accuracy of these techniques in detecting the lesions. In a second study, 15 patients undergoing laparotomy for colorectal cancer underwent preliminary laparoscopic liver ultrasound followed by open palpation and intraoperative ultrasound to compare these methods of liver evaluation.

RESULTS

Laparoscopic liver ultrasound detected 17 of 18 lesions created in the pig livers, for a sensitivity of 94.4%. There were two false negatives, for a specificity of 77.7%. Laparoscopic liver ultrasound detected 4 of the 5 liver metastases in the human study, for a sensitivity of 80%. There was a single false negative, for a specificity of 90.9%. Several technical difficulties and their solutions are discussed.

CONCLUSIONS

With several technical modifications guided by our initial experience, we believe laparoscopic liver ultrasound can be an effective way of evaluating the liver for metastases during laparoscopic colorectal resection for cancer.

Gastric pacing improves emptying and symptoms in patients with gastroparesis

BACKGROUND & AIMS

No effective treatment is available for patients with gastroparesis refractory to standard medical therapy. The aim of this study was to investigate the effects of gastric pacing on gastric electrical activity, gastric emptying, and symptoms in patients with gastroparesis.

METHODS

Nine patients with gastroparesis participated in this study. Four pairs of cardiac pacing wires were implanted on the serosa of the stomach. The protocol consisted of two portions: a temporary inpatient study period and an outpatient study for a period of 1 month or more.

RESULTS

Gastric pacing entrained the gastric slow wave in all subjects and converted tachygastria in 2 patients into regular 3-cpm slow waves. Gastric emptying was significantly improved after the outpatient treatment with gastric pacing. The gastric retention at 2 hours was reduced from 77.0% +/- 3.3% to 56.6% +/- 8.6% (P < 0.05). Symptoms of gastroparesis were substantially reduced at the end of the outpatient treatment (1.51 +/- 0.46 vs. 2.84 +/- 0.61; P < 0.04). Eight of 9 patients no longer relied on jejunostomy tube feeding, and no adverse events were noted related to the pacing unit.

CONCLUSIONS

Gastric pacing seems to be able to improve symptoms of gastroparesis and to accelerate gastric emptying in patients with gastroparesis. More controlled studies are necessary to further investigate the role of gastric pacing in clinical practice.

Effects of pacing parameters on entrainment of gastric slow waves in patients with gastroparesis

The aim of this study was to investigate the effect of pacing parameters on the entrainment of gastric slow waves in patients with gastroparesis. Four pairs of cardiac pacing wires were placed on the serosal surface of the stomach in 13 patients with gastroparesis. After a baseline recording for 30 min, gastric pacing was performed in a number of sessions with different effective parameters, each lasting for 30 min. The following parameters were found to be effective for the entrainment of the gastric slow wave: a pacing frequency 10% higher than the intrinsic gastric slow wave frequency (IGF), 300 ms pulse width, and 4 mA pacing amplitude. A reduction of pacing amplitude from 4 to 2 mA and 1 mA reduced the percentage of entrainment of the gastric slow wave to 79 +/- 10% and 50 +/- 11%, respectively. Pacing with a pulse width of 30 or 3 ms was not able to entrain the gastric slow wave in any of the patients. An ectopic pacemaker of tachygastria found in three patients was reversed with gastric pacing. It was concluded that gastric pacing at a frequency up to 10% higher than the IGF and with an amplitude of 4 mA and a pulse width of 300 ms is able to completely entrain the gastric slow wave and normalize gastric dysrhythmias in patients with gastroparesis.

Recovery of gastrointestinal tract motility and myoelectric activity change after abdominal surgery

OBJECTIVE

To investigate the relationship between fedstate gastrointestinal tract (GI) function and upper GI myoelectric changes seen after abdominal surgery.

DESIGN

Twenty-one adult female mongrel dogs underwent either an open cholecystectomy, a laparoscopic cholecystectomy alone, or a laparoscopic cholecystectomy with peritoneal injury (n = 7 for each group). Bipolar recording electrodes were placed on the antrum and 3 sites of the proximal small intestine to record fasting myoelectric data each morning postoperatively. Solid-phase, technetium Tc 99m gastric emptying studies were performed on postoperative days 1 and 2. Radiopaque markers were ingested just before operation, and the excreted markers were counted using x-ray films of the feces.

MAIN OUTCOME MEASURES

Postoperative fasting GI myoelectric activity, gastric emptying, and intestinal transit time.

RESULTS

Migrating motor complexes (MMCs) in the small intestine were observed in 33.3% and 75.0% of the dogs on postoperative days 1 and 2, respectively. Gastric dysrhythmias were observed in 23.8% and 45.0% of the dogs on postoperative days 1 and 2, respectively. No relationship between type of surgery and the presence of MMCs or gastric dysrhythmias was noted. Gastric emptying was delayed on postoperative day 1 and was unrelated to the presence of MMCs. Transit time was not significantly delayed in dogs without MMCs on postoperative day 1 compared with that in dogs with MMCs on that day. The presence of gastric dysrhythmias did not affect transit time studies.

CONCLUSION

Fasting GI myoelectric activity, including the return of MMCs and the presence of gastric dysrhythmias, does not accurately predict fed-state gastrointestinal GI function following abdominal surgery.

Laparoscopic percutaneous jejunostomy for long term enteral access

BACKGROUND

Patients suffering from upper gastrointestinal pathology may require jejunal feeding for adequate nutrition. A laparoscopically guided percutaneous jejunostomy offers a minimally invasive means of obtaining such feeding access.

METHODS

Laparoscopic jejunostomy was performed in 32 patients. The most common indications were gastroparesis (n = 16), neurological deficits (n = 7), and proximal obstruction (n = 5). The proximal jejunum was affixed to the abdominal wall using intracorporeal and extracorporeal transabdominal sutures, allowing safe insertion of an 18-Fr Silastic dual-lumen tube.

RESULTS

Laparoscopic jejunostomy was successfully completed for 28 patients; the procedure was converted to an open operation in four cases. Three of these four were among 14 patients undergoing the procedure who had a history of previous abdominal surgery. Major complications were observed in seven patients, including one reoperation and one death from aspiration pneumonia. Tube feeding was accomplished in all patients; progression to full enteral feeding proceeded without interruption in 20 patients.

CONCLUSION

Laparoscopic jejunostomy can be performed with relative safety. Morbidity, though high, is usually related to preexisting disease. Previous abdominal surgery is not necessarily a contraindication to laparoscopic jejunostomy.

Laparoscopic colon resection

Laparoscopic colectomy, usually performed in a laparoscopy-assisted fashion, is a technically difficult operation not easily mastered by the average surgeon and requiring a skilled team for its successful completion. There is a significant learning curve for the procedure, and conversion to open colectomy has been necessary in about 25% of cases in collected series. As such, its popularity has increased only slowly, and currently it is appropriate for treatment of benign colonic disease and as a palliative approach for unresectable carcinoma. Although the procedure produces an adequate tissue resection, concern about trocar site tumor recurrences has led to the general consensus that the procedure should currently be done only in a prospective investigational protocol setting for the treatment of curable colorectal carcinoma. These studies are expected to yield the data critically needed to assess its role in treating this disease. Experience to date suggests that laparoscopic colectomy can be performed with morbidity and mortality lower than or comparable to those of open colectomy. It likely is associated with less postoperative pain and a shorter hospitalization and has the potential for modestly more rapid recovery of gastro-intestinal function than open colectomy.

Gastrointestinal recovery following laparoscopic vs open colon surgery

BACKGROUND

We prospectively studied the recovery of gastrointestinal motility in patients undergoing laparoscopic (LAP, n = 7) or open (OPEN, n = 7) colon resections.

METHODS

At operation, bipolar recording electrodes were placed on the proximal and distal antrum, the proximal site of the colonic anastomosis, and the rectosigmoid for postoperative myoelectric recordings.

RESULTS

Shorter postoperative hospitalization and earlier resumption of a regular diet of the LAP group just barely failed to achieve significant differences when compared with the OPEN group (p = 0.091, p = 0.050, respectively). There were no differences between groups for slow wave frequency, amplitude, or dysrhythmias in the antrum, nor for return of discrete (DERA) and continuous (CERA) electrical response activity in the colon. Percentage of slow waves with spike activity tended to increase with passage of time postoperatively in both groups. There was a significant difference between POD 3 and 7+ in the LAP group (p < 0.05). However, there were no significant differences in the percentage of slow waves with spike activities between groups on any postoperative day.

CONCLUSIONS

The potential benefits of using a laparoscopic approach to colon resection are not clearly confirmed by these data. While such an approach may possibly result in shorter hospitalization, it appears to offer at best only modest increases in the rapidity of recovery of gastrointestinal function.

Gastric myoelectric activity changes following open abdominal surgery in humans

We examined the postoperative changes in fasting gastric myoelectric activity in 11 patients undergoing nongastric surgery (colon surgery) via celiotomy. Recordings were performed on postoperative days (POD) 1, 2, 3, 5, and 7+ (7-35) for 1-1.5 hr after overnight fasting. Patients had placement of bipolar seromuscular recording electrodes on the proximal (N = 9) and distal (N = 11) antrum at the time of surgery. Data were analyzed visually and analysis of variance or tests of proportion were used for statistical analysis. Although there was a trend of decreasing slow wave frequency from POD 1 to 7+ in the proximal and distal antrum, no significant differences were observed in slow wave amplitude or in the percentage of slow waves with spike activity between postoperative day. In a few of the patients, several types of gastric dysrhythmias were infrequently observed. We conclude that certain parameters of fasting gastric myoelectric activity do not change sufficiently following open abdominal surgery to adequately reflect clinical recovery from postoperative ileus.

Recovery of fasted and fed gastrointestinal motility after open versus laparoscopic cholecystectomy in dogs

OBJECTIVE

The authors investigate the recovery of gastrointestinal motility in the fed and fasted state after laparoscopic and open cholecystectomy.

SUMMARY BACKGROUND DATA

Clinical recovery after laparoscopic cholecystectomy is known to be more rapid than after conventional open cholecystectomy. However, the actual effect of a laparoscopic approach on gastrointestinal motility, particularly fed-state motility, is not well investigated.

METHODS

Laparoscopic (LAP, n=6) or open (OPEN, n=6) cholecystectomy was performed in 12 dogs. Bipolar recording electrodes were placed on the antrum, small intestine, and the transverse and descending colon, and fasting myoelectric data were recorded after operation. Solid meal gastric emptying studies were performed before surgery and on postoperative days 1 and 2. Transit time studies were performed using 10 radiopaque markers.

RESULTS

Gastric emptying was significantly delayed in the OPEN group at 120 minutes on postoperative day 1 compared with pre-operative emptying (p<0.05), but was not delayed on postoperative day 2. Gastric emptying was not delayed in the LAP group after operation. Transit time was the same between groups. Gastric dysrhythmias were more frequent on postoperative day 3 (p<0.05) in the OPEN group. There were no significant differences in the presence, cycle length, or propagation velocity of the migrating motor complex on any postoperative day. Discrete or continuous electrical response activity in the colon was observed by postoperative day 1 in both groups.

CONCLUSIONS

Fed-state motility is the only parameter for which laparoscopic cholecystectomy showed an improvement in postoperative recovery. Recovery of fasted gastrointestinal motility in dogs is equally rapid after either operation.

Recovery of gastrointestinal motility following open versus laparoscopic colon resection in dogs

The recovery of gastrointestinal motility was compared in dogs undergoing either laparoscopic or open sigmoidectomy. During surgery, bipolar recording electrodes were placed on the proximal and distal antrum, mid- and distal colon, and the rectum. Fasting myoelectric data were recorded postoperatively. Scintigraphic gastric emptying studies employing a solid test meal were performed before and after [postoperative day (POD) 2] operation. Ten radiopaque markers were given just before operation and retained markers were counted daily by abdominal x-ray. Gastric emptying on POD 2 was significantly delayed in the open group at 120 min compared with preoperative studies for the open group and compared with the laparoscopic group on POD 2 (P < 0.05 and P < 0.01, respectively). A significant difference in the number of retained markers was observed between the groups on POD 4 (P < 0.05). There were no significant differences in slow-wave frequency, presence of dysrhythmias in the proximal and distal antrum, or presence of either discrete or continuous electrical response activity in the colon and rectum between groups on any days. We conclude that using a laparoscopic approach results in more rapid recovery of fed-state gastrointestinal motility following colon resection. These data also suggest that myoelectric activity alone is not a sensitive enough parameter to detect these differences in recovery in this animal model.

The impact of previous abdominal surgery on outcome following laparoscopic cholecystectomy

The first 1000 patients undergoing laparoscopic cholecystectomy (LC) at our institution were reviewed to investigate the impact of previous abdominal surgery on LC. The 454 patients having no previous abdominal surgery (NS) were compared to the 541 patients who had previous surgery (PS). PS patients were older, more likely to be female, and had a higher ASA risk category. PS patients had a higher incidence of wound infection, but in all other parameters of outcome, including operative duration and completion, length of hospitalization, and morbidity, there were no significant differences between PS and NS. When PS patients with previous upper abdominal surgery (PUAS, n = 59) were separately compared to the remainder of the entire patient group (NUAS, n = 936), the PUAS group was found to be older, to be more likely to be male, and to have a higher ASA risk category. PUAS patients had a longer postoperative hospitalization, and an increased incidence of intraoperative, postoperative, and total complications, readmissions to the hospital, and unrelated deaths. We conclude previous lower abdominal surgery has little impact on the outcome of patients undergoing LC while previous upper abdominal surgery is associated with increased morbidity.

A retrospective analysis of 6,387 cholecystectomies

This is the largest retrospective analysis of biliary tract surgery ever reported involving 6,378 patients operated on during a three year period, 1990-1992. During this time interval, the frequency of laparoscopic procedures has increased dramatically. The use of laparoscopic procedures was associated with a significant decrease in the total length of hospital stay and total charges as compared to the open procedures. Because of laparoscopic surgery's increased acceptance, we propose that the frequency of laparoscopic surgery of the biliary tract should be used as a quality control measure.

Gastric atony and the Roux syndrome

The Roux limb syndrome is a symptom complex characterized by chronic postprandial epigastric pain, fullness, and vomiting observed in approximately one third of patients after gastric reconstructive surgery for reflux gastritis and other conditions in which vagotomy and Roux-en-Y gastroenterostomy have been preformed. The etiology of the symptom complex is controversial, with experimental evidence in animal and human studies suggesting dysfunction of both the gastric remnant and the Roux limb itself. Medical treatment is successful in only about half of cases, but surgical treatment to remove most or all of the gastric remnant is usually successful. These observations suggest that most of the symptoms of the Roux limb syndrome arise from postvagotomy gastric atony.

Serosal and cutaneous recordings of gastric myoelectrical activity in patients with gastroparesis

The aims of this study were to 1) investigate gastric myoelectrical activity in patients with gastroparesis, 2) validate the cutaneous electrogastrogram (EGG) in tracking the frequency change of the gastric slow wave, and 3) investigate the effect of electrical stimulation on gastric myoelectrical activity. Gastric myoelectrical activity was recorded in 12 patients with documented gastroparesis using serosal electrodes for > 200 min in each subject. All recordings were made at least 4 days after surgery. Each session consisted of a 30-min recording in the fasting state and a 30-min recording after a test meal. The test meal (liquid or mixed) was selected according to patient's tolerance. Electrical stimulation was performed in three subjects via the serosal electrodes at a frequency of 3 cycles/min. Gastric myoelectrical activity was recorded using serosal electrodes in each session. The serosal recording showed slow waves of 2.5 to 4.0 cycles/min in all 12 subjects. Absence of spikes was noted in 11 of the 12 subjects. The simultaneous serosal and cutaneous recording of gastric myoelectrical activity showed that the frequency of the EGG was exactly the same as that of the serosal recording. Liquid meals resulted in a significant decrease in slow-wave frequency (Student's t test, P = 0.006), and the EGG accurately reflected this change. Electrical stimulation had no effect on the frequency of the gastric slow wave and did not induce spikes.(ABSTRACT TRUNCATED AT 250 WORDS)

Effect of crural myotomy on the incidence and mechanism of gastroesophageal reflux in cats

BACKGROUND

Esophagogastric junction (EGJ) pressure is the major barrier to gastroesophageal reflux. Recent studies suggest that contraction of the crural diaphragm increases esophagogastric junction pressure. Whether this increase in EGJ pressure is important in the prevention of gastroesophageal reflux is not known. Our aim in this study was to determine the effects of crural myotomy on the occurrence of gastroesophageal reflux.

METHODS

The spontaneous and stress gastroesophageal reflux before and after a surgical crural myotomy in four cats was studied. Spontaneous gastroesophageal reflux was recorded in the awake cats through a pH probe, placed via an esophagostomy, for periods of 12-24 hours. Stress reflux was studied during periods of airway obstruction and abdominal compression in anesthetized animals using the technique of simultaneous esophageal manometry and pH monitoring.

RESULTS

There was a significant increase in the frequency of spontaneous acid reflux after crural myotomy. In anesthetized animals, there was an increase in the EGJ pressure during airway obstruction, which was abolished by a crural myotomy. Abdominal compression caused a reflex contraction at the EGJ that was not affected by crural myotomy. The crural myotomy resulted in a significant increase in the frequency of acid reflux during airway obstruction but not during abdominal compression.

CONCLUSIONS

It is concluded that the crural diaphragm is important in the prevention of gastroesophageal reflux and its dysfunction leads to an increased incidence of gastroesophageal reflux.

Measurement of electrical activity of the human small intestine using surface electrodes

Electrical activity of the human small intestine is usually measured by implanted or intraluminal electrodes. The application of these invasive techniques is, however, very limited. In this paper, a noninvasive technique is introduced to measure electrical activity of the small intestine by placing electrodes on the abdominal skin over the small intestine. Surface recordings were obtained in ten healthy volunteers, three patients with total gastrectomy and five patients with gastroparesis (a slight degree of paralysis of the mucosal coat of the stomach) with implanted electrodes on the serosa of the duodenum. An omnipresent 9-12 cpm electrical activity was observed in all surface recordings. Our findings from the surface electrodes were consistent with those reported in the literature via implanted or intubated electrodes. It is concluded that the technique described in this paper provides a noninvasive way to measure electrical activity of the small intestine. It may have potential application in medical research and clinical diagnosis.

Laparoscopic versus traditional appendectomy for suspected appendicitis

We compared the results of concurrently performed laparoscopic versus open appendectomy as treatments for suspected acute appendicitis. The 68 laparoscopic procedures resulted in 62 appendectomies, 47 by the laparoscopic (LA) technique and 15 by the open (LO) technique. Another 54 patients underwent open appendectomy (OA). Significantly more females underwent laparoscopy (LA and LO: 52% versus OA: 33%, p = 0.047). Operative duration was shortest for OA (81 +/- 3 minutes), which was shorter than for LO (108 +/- 7 minutes), but not different than LA (86 +/- 6 minutes). The postoperative length of stay was not different for LA (3.5 +/- 0.5 days) compared with OA (5.9 +/- 1.6 days) or LO (4.8 +/- 1.3 days). One death occurred in the OA group. Wound complication rates were not significantly different for LA (4.3%) compared with OA (9.4%) and LO (13.3%). Overall complication rates were lower for LA (10.6%) and OA (18.9%) compared with LO (46.7%, p < 0.01). Median hospital cost for LO ($10,425) was higher (p < 0.02) than for either LA ($5,899) or OA ($5,220). When appendicitis was not present, definitive confirmation of pathology was achieved in 9 of 18 patients undergoing LA versus 4 of 14 patients having OA (p = not significant). We conclude that when laparoscopy and laparoscopic appendectomy can be performed, the procedure is safe and produces results comparable with those of open appendectomy without significant overall cost differences.

Effect of a catheter in the pharynx on the frequency of transient lower esophageal sphincter relaxations

Transient relaxation of the lower esophageal sphincter (LES) is observed fairly frequently during prolonged continuous monitoring of the LES. The aim of this study was to test whether the presence of a catheter in the pharynx through the stimulation of mechanoreceptors may induce transient LES relaxation. LES and esophageal pressure recordings were obtained for 1 hour in six subjects with a manometric catheter placed via a gastrostomy tube. Swallowing was monitored by submental electromyographic recording. Additional recordings were obtained in these subjects with a catheter placed in the pharynx for 1 additional hour. Transient LES relaxations were recorded in both study periods, i.e., with and without a catheter in the pharynx. The frequency of transient LES relaxations was significantly higher in the presence of manometric catheters in the pharynx (6.4 +/- 2.2 vs. 2.0 +/- 1.1 total LES relaxations). The frequency of transient LES relaxation during the first and second hour after placement of the manometric catheter in a group of seven healthy subjects was not significant different. It is concluded that the pharynx is one of the sites that may mediate the induction of transient LES relaxation.

Cost effectiveness of laparoscopic cholecystectomy

An investigation was undertaken to determine whether hospital charges for laparoscopic cholecystectomy are higher than those for traditional open cholecystectomy. Thirty consecutive cases of successfully completed laparoscopic procedures in a single surgeon's experience were compared to 30 open cases performed within the previous calendar year. Patients undergoing open cholecystectomy were excluded if coexisting medical problems or complications prolonged hospitalization beyond 7 days. Mean patient age was comparable (open cholecystectomy = 47.3 +/- 2.9, laparoscopic cholecystectomy = 46.5 +/- 2.7 years), as was the incidence of other significant medical problems. Average duration of hospitalization was significantly longer for open cholecystectomy (3.6 = 0.2 days) than for laparoscopic cholecystectomy (1.0 +/- days, p less than .001). Average hospital charges for open cholecystectomy were $5606 +/- 496 and for laparoscopic cholecystectomy $4726 +/- 98. Hospital charges from operating room and recovery room charges alone were $2684 +/- 131 for laparoscopic cholecystectomy and $2196 +/- 113 for open cholecystectomy. These operating room charges represent a significantly higher percentage of total hospital charges for laparoscopic cholecystectomy than open cholecystectomy patients (laparoscopic cholecystectomy = 56.3 +/- 1.9%, open cholecystectomy = 41.2 +/- 1.5%, p less than .05). Average time for return to work or normal activity was significantly shorter for laparoscopic cholecystectomy 8.6 +/- 9 days) than for open cholecystectomy (32.4 +/- 3.6 days, p less than .001). The authors conclude that laparoscopic cholecystectomy is a cost effective procedure for the treatment of symptomatic cholelithiasis, and that increased operative costs more than offset the significantly decreased length of hospitalization.

Laparoscopic cholecystectomy in the obese patient

The authors' experience with laparoscopic cholecystectomy (LC) in obese (O, n = 96) and morbidly obese (MO, n = 27) patient groups was compared with that in the normal weight (NW, n = 174) group of patients as well as the whole group (WG). There were no operative deaths. There were no significant differences between groups for any of the following: successful intraoperative cholangiography (WG, 52.2%; NW, 52.9%; O, 51.1%; MO, 55.6%), conversion to open cholecystectomy (WG, 9.6%; NW, 9.2%; O, 10.4%; MO, 11.1%), incidence of major complications (WG, 4.1%; NW, 3.4%, O, 5.2%; MO, 0%), incidence of minor complications (WG, 7.4%, NW, 7.5%; O, 6.3%; MO, 3.7%), and length of hospitalization after successful LC (WG, 1.25 days; NW, 1.31 days; O, 1.16 days; MO, 1.13 days). Duration of operation did not differ except LC in the MO group (136.4 +/- 6.9 minutes) was longer when compared with NW patients (123.0 +/- 2.9 minutes, p less than 0.05). The authors conclude LC is a safe and effective treatment for obese patients with symptomatic cholelithiasis.

Acceleration of gastric emptying with electrical stimulation in a canine model of gastroparesis

We investigated the effects of electrical stimulation of the stomach on gastric emptying and the electrical activity of the stomach in 10 dogs. A model of gastroparesis was developed in five dogs using truncal vagotomy combined with injections of glucagon. Glucagon also induced electrical dysrhythmias. Bipolar electrodes were implanted in the stomach and the duodenum for electrical stimulation and for recording electrogastrograms. Gastric emptying of an isotope-labeled solid meal was assessed for 2 h. External electrical stimulation was delivered to the corpus of the stomach at its own physiological frequency to investigate whether it could restore normal gastric emptying. Such stimulation had no significant effect on gastric emptying in intact animals (45 vs. 43%: retention of isotope after 2 h) or when only vagotomy was performed (78 vs. 66%), but it significantly accelerated gastric emptying in animals with vagotomy and glucagon (from 86 to 68%). From this model of delayed gastric emptying, we suggest that electrical stimulation of the stomach at its own intrinsic frequency may recoordinate uncoupled slow wave activity induced by glucagon after vagotomy thus improving the rate of gastric emptying.

Incorporation of laparoscopy into a surgical endoscopy training program

The impact of introducing laparoscopy as part of the overall gastrointestinal endoscopy case load performed by residents was reviewed. During 1990, there was a significant increase (56.9%) in the number of flexible diagnostic endoscopic procedures performed compared with 1989. When the total number of laparoscopic procedures was considered, the increase was 117%. Residents participated in the "surgeon's" position in 59% of the therapeutic laparoscopic procedures and as either surgeon or "first assistant" in 86% of all therapeutic laparoscopic procedures and 94% of all diagnostic laparoscopic procedures. Complication rates for diagnostic laparoscopic procedures were low in 1989 (0.03%) and 1990 (0.2%). Complication rates for therapeutic laparoscopic procedures were also low (4%). There was no difference in the complication rate for cases in which residents were in the surgeon's position (4%) versus cases in which they were not (4%). Introduction of laparoscopic procedures into a surgical residency program can be done safely, especially in cases in which an established program in endoscopy exists.

Laparoscopic cholecystectomy. Treatment of choice for symptomatic cholelithiasis

Our initial experience with laparoscopic cholecystectomy (LC) for symptomatic cholelithiasis has involved 152 patients. Patient age ranged from 17 to 83 years; most were female (78%). Their average weight was 170 pounds (range, 75 to 365 lbs.). Twenty-two per cent had a single gallstone, while 9% had two to three stones and 64% had more than three stones. Exclusion criteria initially included upper abdominal scarring, severe acute cholecystitis, choledocholithiasis, and inability to tolerate general anesthesia. The first two of these are now only relative contraindications with increased experience. Thirteen of the one hundred fifty-two procedures (8.5%) required conversion to an open operation. Average time of operation was 138 minutes. Intraoperative cholangiography was attempted in 78% of cases and was completed successfully in 66% of those attempted. There have been no deaths. The complication rate has been low: 4% major, 0% life-threatening, and 7.2% minor complications. Postoperative analgesic requirements are remarkably low: 36% of patients required no narcotics after leaving the recovery room. Eighty-seven per cent of patients successfully undergoing LC were discharged by the first postoperative day. Most patients resumed normal activities within 1 week after discharge. Laparoscopic cholecystectomy offers the majority of patients with symptomatic cholelithiasis an improved treatment option, resulting in significantly less postoperative pain, hospitalization, and recuperation time.

The effect of ambulation on recovery from postoperative ileus

To determine whether ambulation hastens recovery from ileus following laparotomy, 34 patients were studied, 10 of whom followed an ambulatory regimen beginning on postoperative day 1 (group A). The other 24 patients (group C) did not become ambulatory until postoperative day 4. All patients underwent placement of seromuscular bipolar recording electrodes on the Roux limb, if present, stomach, jejunum, and colon at laparotomy. Group A was recorded before and after ambulation so comparisons could be made to determine if ambulation had an acute effect on myoelectric activity. Group A preambulation and group C recordings were compared to judge whether there was an over-all effect of ambulation on myoelectric recovery. No effect on slow wave frequency or percentage of slow waves with associated spike potentials was noted acutely or overall in the stomach, colon, or jejunum in continuity with the duodenal pacemaker. Transient increases in phase II spike activity in patients having a Roux limb and their jejunum distal to the enteroenterostomy were noted on postoperative days 1 to 2, but these differences resolved by postoperative days 3 or 4. The data suggest that ambulation as a means to help resolve postoperative ileus and its accompanying cramps and bloating may be more perceived than real.

Gastrointestinal myoelectric and clinical patterns of recovery after laparotomy

The objective of this study was to define the patterns of myoelectric activity that occur throughout the gastrointestinal tract during normal recovery from laparotomy. Electrodes were placed on the stomach, jejunum, and transverse colon of 44 patients undergoing laparotomy. Basal electric rhythms in all areas showed no changes in frequency after operation (up to 1 month). Gastric spike wave activity showed a gradient of increasing activity from fundus to antrum. Antral spike activity was unchanged during the study. Jejunal spike activity was present in the earliest recordings and occurred in 45.9% +/- 3.5% to 59.9% +/- 5.5% of slow waves. Recovery of normal colon discrete and continuous electric response activity occurred on postoperative day 5.9 +/- 1.5. Bowel sounds returned on day 2.4 +/- 0.5 and passage of flatus and stool occurred on day 5.1 +/- 0.2. The myoelectric parameters measured are not absolutely predictive of uneventful recovery from postoperative ileus but they are, as a group, more informative than any currently available clinical criteria.

Ileostomy using an interposed ileal segment to function as a 'neocolon'

A new type of ileostomy was constructed in rats using an interposed ileal (I-I) segment functioning as a neocolon. The effects of this operation in these rats were compared with a control group undergoing conventional ileostomy (I). Rats in the I group lost weight throughout follow-up, whereas rats in the I-I group regained weight and were significantly heavier four weeks after surgery. Transit time from the stomach to the stoma was significantly prolonged in the I-I group (187 +/- 29 minutes) when compared with the I group (141 +/- 17 minutes, P less than 0.01). Serum analysis revealed no significant differences in total protein, lipids, electrolytes (Na, Cl, CO2), and osmolarity between the groups. Biliary bile salt concentration, normalized bile flow, and normalized bile acid output in the I-I group were not different from those in unoperated control rats, suggesting unimpaired bile acid metabolism. These data suggest this operative procedure has the beneficial effects of slowing intestinal transit and promoting postoperative recovery without impairing distal ileal function.

Current status of proximal gastric vagotomy

Proximal gastric vagotomy is nearing its twentieth year in clinical use as an operation for peptic ulcer disease. No other acid-reducing operation has undergone as much scrutiny or study. At this time, the evidence of such studies and long-term follow-up strongly supports the use of proximal gastric vagotomy as the treatment of choice for chronic duodenal ulcer in patients who have failed medical therapy. Its application in treating the complications of peptic ulcer disease, which recently have come to represent an increasingly greater percentage of all operations done for peptic ulcer disease, is well-tested. However, initial series suggest that it should probably occupy a prominent role in treating some of these complications, particularly in selected patients, in the future. The operation has the well-documented ability to reduce gastric acid production, not inhibit gastric bicarbonate production, and also minimally inhibit gastric motility. The combination of these physiologic results after proximal gastric vagotomy, along with preservation of the normal antropyloroduodenal mechanism of gastrointestinal control, serve to allow patients with proximal gastric vagotomy the improved benefits of significantly fewer severe gastrointestinal side effects than are seen after other operations for peptic ulcer disease.

Glucagon lowers canine biliary cholesterol output at physiologic doses

Pharmacologic doses of glucagon affect canine bile secretion by increasing bile flow while simultaneously decreasing biliary cholesterol output. The present study was performed to determine if physiologic doses of glucagon reduce biliary cholesterol output. Awake dogs received both intravenous 1% sodium taurocholate (50 ml/hr) to stabilize bile flow and somatostatin (12 micrograms/kg/hr) to suppress endogenous pancreatic hormone release. Suppression was documented by significant decreases in portal plasma glucagon and insulin levels. During experimental trials, dogs received, in addition, glucagon (5 ng/kg/min) infused via a splenic vein catheter. Bile flow significantly decreased during the initial hour of somatostatin infusion but increased significantly only in experimental trials during subsequent glucagon infusion. Biliary cholesterol output showed no change during control studies (N = 9), but decreased significantly during glucagon infusion studies (N = 11). Biliary phospholipids and bile salts failed to show any changes during glucagon infusion. These data demonstrate that glucagon at physiologic levels influences both the volume and cholesterol content of bile and suggest the mechanism of decreasing cholesterol output must be independent of pathways for influencing bile salt or phospholipid secretion.

Effects of serotonin on canine bile formation

Long-term studies were performed on dogs previously prepared by cholecystectomy, ligation of the lesser pancreatic duct, and insertion of a duodenal cannula. After an overnight fast, bile duct cannulation and stabilization of bile flow with intravenous (IV) sodium taurocholate, serotonin, 10 micrograms/kg/min, or 0.15 N NaCl was infused. In similar experiments, animals were fed a standard meal, and serotonin or 0.15 N NaCl was infused IV beginning simultaneously with or 30 minutes after the meal. Short-term experiments were performed on dogs anesthetized with pentobarbital and prepared by abdominal evisceration, cholecystectomy, and bile duct cannulation. Serotonin caused significant inhibition of fasting bile formation (3.8 +/- 0.3 ml/15 min to 3.2 +/- 0.3 ml/15 min), meal-stimulated choleresis (4.0 +/- 0.3 ml/15 min to 3.5 +/- 0.3 ml/15 min), and bile flow in eviscerated animals (1.6 +/- 0.1 ml/15 min to 1.1 +/- 0.2 ml/15 min). Bile acid output and 14C erythritol clearance were stable while bile bicarbonate output was decreased during serotonin infusion. A similar inhibitory effect was demonstrated with serotonin, 5 micrograms/kg/min, but the inhibition was not statistically significant with 2.5 micrograms/kg/min. These studies demonstrate that serotonin inhibits bile acid-independent bile formation, possibly at the ductular level, and the inhibition occurs independently of endogenous gastrointestinal tract hormone secretion.

Is somatostatin a directly acting cholestatic hormone?

Previous experiments have demonstrated the cholestatic effects of somatostatin administration in several animal species. These effects were confirmed in the rat. Nine pairs of intact awake rats received intravenous sodium taurocholate (23 mg hr-1) to stabilize bile flow, and half were later given somatostatin at doses of 185 micrograms hr-1. After 1 hr of somatostatin the experimental group showed a significant decrease in bile flow compared to the control group. Cholestasis reversed when somatostatin infusion was stopped. An in situ isolated perfused rat liver technique was used to assess the direct effects of somatostatin on biliary flow. In 10 pairs of rat livers, after achieving stable bile flow, half of those perfused (the "experimentals") received continuous (370 micrograms hr-1) somatostatin infusion, while the controls received saline. The percentage change in bile flow from baseline in the somatostatin group was not significantly different from that in controls for any test period. Bile analysis revealed no significant differences between groups for cholesterol, phospholipid, or bile salt concentrations or outputs. These data suggest that somatostatin inhibits bile secretion by some mechanism other than direct inhibition of bile secretory mechanisms.

Hepatic uptake of chylomicrons and triglyceride emulsions in rats fed diets of differing fat content

The hepatic removal of plasma chylomicrons was determined for rats fed the following diets: a) containing no triglyceride, b) regular chow diet with 4.5% of its mass as lipid and, c) a corn oil-supplemented chow with triglyceride accounting for 20% of the mass. The fractional hepatic uptake of either radiolabeled chylomicrons or a triglyceride emulsion was reciprocally related to the amount of lipid in the diet. The animals receiving only carbohydrate and protein calories had the most active hepatic uptake of particulate triglyceride and were observed to have a significant decrease in the plasma concentration of the C apolipoproteins. The addition of either C-I, C-II, or C-III apoproteins to the triglyceride emulsion prior to intravenous injection produced a significantly lower hepatic triglyceride recovery of emulsions containing apoC-III. When the plasma of animals fed a fat-free diet was supplemented with human C-III-1 apolipoprotein, the distribution into the liver of either enterally administered fatty acid or parenteral triglyceride was diminished. The triglyceride content in the liver of the rats fed fat-free or corn oil-supplemented diets was significantly greater than that of the control rats and composition was somewhat similar to that of lymph triglyceride. The studies indicate an important influence of dietary lipid on both the partition of plasma triglyceride into the liver and the steady state hepatic triglyceride content.

Hepatic resection for metastatic cancer

One-year survival is infrequent in patients with metastatic cancer to the liver. This report includes 21 patients who underwent hepatic resection between 1974 and 1981. Operative procedures included one trisegmentectomy, 12 right hepatic lobectomies, two left hepatic lobectomies, two left lateral segmentectomies, and four wedge resections. Operative morbidity and mortality rates were 43% and 5%, respectively. Life-table analysis revealed an overall 7-year survival rate of 34%. The subset of patients (16) with colorectal adenocarcinoma had a 7-year survival rate of 29% after hepatic resection. In three patients with colorectal adenocarcinoma, frequent CEA determinations were made after surgery in order to calculate the serum half-life of CEA. The data fitted a biexponential function yielding two half-lives for CEA disappearance, 0.8 +/- 0.5 days and 25.9 +/- 10.3 days. We conclude that hepatic resection for isolated hepatic metastases can be performed with acceptable morbidity, low mortality, and prolongation of patient survival.

Effects of dibutyryl cyclic AMP and theophylline on biliary cholesterol secretion

Glucagon increases hepatocellular cAMP and decreases biliary cholesterol output. In these experiments, we examined the relation between cAMP and biliary cholesterol secretion. Bile flow and composition were measured in conscious dogs previously prepared by cholecystectomy, ligation of the lesser pancreatic duct, and placement of duodenal and gastric cannulae. Sodium taurocholate (500 mg/hr) was given intravenously to stabilize bile flow. After 2 hr of taurocholate infusion, dibutyryl cyclic AMP (160 mg kg-1 hr-1) or theophylline (20 mg kg-1 hr-1) was administered intravenously. Dibutyryl cAMP caused a decrease in both cholesterol concentration (242 +/- 25 micrograms/ml to 81 +/- 11 micrograms/ml) and cholesterol output (692 +/- 102 micrograms/15 min to 382 +/- 47 micrograms/15 min). Theophylline decreased cholesterol concentration (282 +/- 39 micrograms/ml to 221 +/- 21 micrograms/ml), but there was no significant change in cholesterol output. Bile flow increased significantly with both dibutyryl cAMP (2.8 +/- 0.2 ml/15 min to 4.9 +/- 0.2 ml/15 min) and theophylline (2.6 +/- 0.4 ml/15 min to 4.2 +/- 0.4 ml/15 min). In additional experiments, aminophylline (85% theophylline, 15% ethylenediamine) was administered intravenously (24.7 mg kg-1 hr-1). Aminophylline reduced cholesterol concentration (59 +/- 6 micrograms/ml to 36 +/- 5 micrograms/ml), but cholesterol output was stable. Bile flow increased significantly (3.7 +/- 0.2 ml/15 min to 6.5 +/- 0.4 ml/15 min). The mechanisms of these changes remain unknown. The effect of dibutyryl cAMP on biliary cholesterol secretion supports but does not prove the hypothesis that glucagon decreases biliary cholesterol output via the second messenger, cAMP.

Neural control of gastrointestinal motility: evidence for noncholinergic regulatory influences

Although it is well known that neural control of gastric motility occurs via sympathetic, parasympathetic, noncholinergic, and nonadrenergic fibers contained in the vagus nerve, the central sites of origin of these influences are largely unknown. Recent experiments in our laboratory indicate that noncholinergic neural pathways originating in the posterior hypothalamus can markedly influence gastric motility. At least 2 weeks prior to the experiment, mongrel dogs were surgically prepared with bipolar recording electrodes fixed to the serosal surface of the stomach. This prevented violation of the abdominal cavity on the day of testing. Experiments were performed with the animals under alpha-chloralose anesthesia (100 mg kg-1) in temperature-controlled settings. Under stereotactic guidance, bipolar stimulation of the posterior periventricular hypothalamus produced profound reproducible excitatory or inhibitory effects on gastric myoelectric and motor activity. Changes in the frequency and amplitude of pacesetter potentials (PPs) and in the incidence of action potentials associated with them were observed. Stimulation of various loci in 14 dogs resulted in a 71 +/- 7.5% increase in the incidence of action potentials associated with gastric PPs in "excitatory" areas (n = 7) and a 69.1 +/- 4% decrease in this ratio in "inhibitory" areas (n = 19). In general, more lateral stimulation produced greater inhibitory effects. Responses were frequency dependent, with a threshold greater than 25 Hz in most cases. Excitatory gastric responses to hypothalamic stimulation occurred despite full systemic atropinization (0.1 mg kg-1). The physiologic significance of these noncholinergic excitatory pathways influencing distal gastric motility and the neurotransmitters they employ are as yet unknown.

Somatostatin-induced cholestasis can be independent of portal blood flow

Short-term experiments were performed on adult mongrel dogs (15 to 25 kg) anesthetized with sodium pentobarbital. The operative procedure included cholecystectomy, side-to-side mesocaval shunt with ligation of the portal vein, and cannulation of the common bile duct. Intravenous sodium taurocholate (500 mg/hr) was administered to prevent depletion of bile salts. Somatostatin (125 micrograms over 30 minutes) was given to six dogs after 2 hours of bile salt infusion, while six additional dogs received saline to serve as control. Bile flow decreased significantly during administration of somatostatin (206 +/- 28 to 150 +/- 21 microliters kg-1 15 min-1, P less than 0.001) and was unchanged during administration of saline (216 +/- 45 to 216 +/- 46 microliters kg-1 15 min-1). This decrease persisted for 1/2 hour after cessation of the somatostatin infusion. Bile salt outputs were similar for both groups throughout the experiment. The data demonstrate that somatostatin-induced cholestasis can be independent of portal blood flow.

The direct effect of insulin on canalicular bile flow

Previous experimental studies in dogs have demonstrated the choleretic effects of insulin administered in physiologic doses. In our experiment the isolated perfused rat liver was used as an in situ model for demonstrating the direct effects of insulin on bile salt independent canalicular bile flow. Livers were initially perfused with oxygenated Krebs-Ringer buffer solution of 20% hematocrit after which sodium taurocholate (18.3 mg hr-1) was infused to stabilize bile flow. Insulin (0.5 unit kg-1 hr-1) or saline was then given for a 90-min test period. Control (C) and experimental (E) rats were perfused simulataneously (n = 11). Bile flow was significantly greater in the experimental (21.76 +/- 2.16 microliters g liver-1 15 min-1) than the control (16.87 +/- 2.08 microliters g liver-1 15 min-1) group (P less than .01). In addition, the percentage change in flow in individual rats from baseline values was significantly different with peak effect occurring 45 min after start of hormone infusion (E = +19 +/- 3%, C = -5 +/- 7%) (P less than 0.003). Bile analysis revealed stable bile salt concentrations (E = 7.41 +/- .25, C = 7.8 +/- .24 mM ml-1) and stable bile salt outputs (E = 1.68 +/- .14, C = 1.33 +/- .13 mM 15 min-1) for both groups. Biliary cholesterol outputs did not differ between groups (E = 42.6 +/- 4.8, C = 38.2 +/- 7.4 micrograms 15 min-1), nor did phospholipid outputs (E = 3.56 +/- .63, C = 2.71 +/- .57 mg 15 min-1). These data support the hypothesis that the choleretic action of insulin is a direct one on the hepatocyte, requiring no intermediate step or mediator.

Marginal ulcer. A difficult surgical problem

One hundred sixty-six patients with documented recurrent or marginal ulcers following previous ulcer operations were seen at Duke Medical Center and the Durham VA Hospital from 1950 through 1980. Patients with the diagnosis of gastrinoma were excluded from the series. Evaluation of initial operation for recurrent ulcer showed that the highest recurrence rate occurred following non-acid-reducing operations. Analysis of the symptom-free interval following initial ulcer operation showed a significantly longer interval prior to recurrent ulcer development following gastroenterostomy than other procedures, while resection and Billroth I reanastomosis showed a significantly shorter symptom-free interval than did other procedures. Endoscopy proved 85% sensitive in making the diagnosis of marginal ulcer, while upper GI series was 71% sensitive. Surgical treatment of 132 patients resulted in a 20.4% recurrence rate of second marginal ulcer, with a 2.3% mortality rate and a 10.6% morbidity rate. Second operation for recurrent ulcer in 24 patients yielded no deaths, a 12.5% morbidity rate, and a 29.2% recurrence rate. Average follow-up for the series was 12.3 years, and ultimate outcome of treatment showed, of patients not lost to follow-up, a 58.2% satisfactory to excellent rating, while 42.8% of patients had an unsatisfactory result of treatment.