Showing papers in "Archives of Surgery in 1983"

Cancer: Principles and Practice of Oncology

Abstract: Anyone who cares for or aspires to care for cancer patients should buy and read this book It will become the standard reference for what now is popularized as "multimodal" cancer therapy (chemotherapy, radiotherapy, and surgery) The arguments for reading this huge tome are straightforward: (1) cancer continues to be an overwhelming health problem; (2) more kinds of cancers are curable now than when most clinicians were trained; and (3) effective cancer therapy is becoming more complicated For those who have difficulty justifying the considerable purchase price, remember that the cost per pound for Cancer is less than that for veal (pointed out first to me by one of the editors!) DeVita, Hellman, and Rosenberg have organized their exposition superbly Initial summary chapters outline the magnitude of the clinical cancer problem, recent changes in the incidence of particular kinds of cancer, and the principles that underlie each of the three

General thoracic surgery

Abstract: This second edition consists of 18 sections with 75 chapters written by 79 contributors. Of these, 26 chapters are new, reflecting advances in thoracic surgery during the past ten years. Diaphragmatic pacing, lung transplantation, and laser endoscopic procedures are examples of this expanding field. The text is arranged in a logical sequence of anatomy, physiology, diagnostic procedures, and preoperative evaluation. Following is more specific consideration of the surgical treatment of trauma and diseases of the chest wall, diaphragm, lungs, esophagus, and mediastinum. The section on anesthetic treatment of patients undergoing thoracic surgery includes discussion of postoperative care and ventilatory support. Considerable attention is devoted to the continuing study of the function and motor disturbances of the esophagus in the sections on physiology and the esophagus. Chapters dealing with radiation therapy, chemotherapy, and immunology discuss the optional or adjuvant therapies currently available. Technical aspects are well presented in the section on

The Influence of Hair-Removal Methods on Wound Infections

Abstract: The influence of preoperative shaving v clipping on wound infection rate was studied in 1,013 patients undergoing elective operations at a single hospital. Patients were prospectively randomized to be either shaved or clipped the night before or the morning of operation. The AM clipper method was associated with significantly fewer infections than were the other methods, both at discharge and at 30-day follow-up. The greatest benefit was in the group with clean wounds. For each 1,000 patients treated, a savings of approximately $270,000 could be realized if the AM clipper method replaced shaving for preoperative hair removal. Preoperative shaving is deleterious, and the practice should be abandoned.

Impact of regionalization. The Orange County experience.

Abstract: • Prior to the designation of a trauma system in Orange County, Calif, 73% of the in-hospital non-CNS deaths secondary to motor vehicular trauma were judged by the autopsy method to have been preventable. In June 1980, a regional system of trauma care with designated trauma centers was established in Orange County. We used the autopsy method to evaluate the first year's experience with this new system and compared the results with previous studies in Orange County for 1974, San Francisco County in 1974, and Orange County in 1978-1979. The results indicate a severe reduction in the number of deaths judged preventable. In addition, a more aggressive approach to the traumatized patient was noted as indicated by an increased percentage of patients who received appropriate surgical intervention. ( Arch Surg 1983;118:740-744)

Reduced Mortality From Bleeding Pseudocysts and Pseudoaneurysms Caused by Pancreatitis

Abstract: • Exigent hemorrhage from pseudocysts and pseudoaneurysms is the most rapidly lethal complication of pancreatitis. Of eight patients with this unusual entity seen by us, all had acute gastrointestinal hemorrhage; two patients had intraperitoneal bleeding as well. Preoperative visceral arteriograms accurately defined the bleeding lesion and greatly aided in planning operative strategy in six patients. Emergency celiotomy and arterial ligation were accomplished in seven patients, and one patient underwent successful transcatheter arterial embolization. Pancreatic resection was not required in any patient for control of hemorrhage, although gastrectomy was necessary in three cases. One elderly patient died of sepsis five weeks after operation. Our mortality of 12.5% compares favorably with the 37% overall mortality from 123 cases reported in the literature. ( Arch Surg 1983;118:45-51)

Prognosis in Generalized Peritonitis: Relation to Cause and Risk Factors

Abstract: • Generalized peritonitis was assessed in 176 patients, 67 (38%) of whom died. Cases were divided into causative groups: (1) appendicitis and perforated duodenal ulcer, (2) intraperitoneal origin other than appendix or duodenum, and (3) postoperative peritonitis. Mortalities were 10%, 50%, and 60%, respectively. Postoperative peritonitis was characterized by lack of influence of age on outcome, late operation, and more frequent organ failure. Delayed surgery carried a worse prognosis. Organ failure was a risk factor with 76% mortality, and was associated with late operation. Early surgery in organ failure improved survival. More sensitive indicators of early organ dysfunction might improve survival. ( Arch Surg 1983;118:285-290)

Barrett's esophagus. Functional assessment, proposed pathogenesis, and surgical therapy.

Abstract: • Esophageal function was assessed with manometry and 24-hour pH monitoring of the distal esophagus in 22 patients with histologically proven Barrett's esophagus (BE), 31 consecutive patients with endoscopic grade 2 or 3 esophagitis, and 33 normal volunteers. Patients with BE had less lower esophageal sphincter (LES) pressure, but similar length of sphincter exposed to the abdomen, than patients with esophagitis. Both groups had significantly less LES pressure and abdominal length than normal subjects. Patients with BE had statistically more esophageal acid exposure than patients with esophagitis, and both differed markedly from normal subjects. They also had a greater number of reflux episodes lasting longer than five minutes than patients with esophagitis, suggesting that the severity of acid exposure was due to a defect in esophageal clearance. The extent of Barrett's mucosal change was related to the level of LES pressure and the number of reflux episodes that were five minutes or longer in duration. We concluded that BE is related to a mechanical incompetency of the cardia and a decrease in esophageal clearance that requires reconstruction of the cardia for effective therapy. (Arch Surg1983;118:543-549)

Percutaneous Endoscopic Gastrostomy: Review of 150 Cases

Abstract: • Percutaneous endoscopic gastrostomy was performed on 50 children and 100 adults from June 1979 to May 1982 at Case Western Reserve University Hospitals and the Mt Sinai Medical Center in Cleveland. Morbidity was low (10%), and there were no procedure-related deaths. Complications included minor wound infections in seven patients early in the series, extrusion of the tube in three, unnecessary laparotomy in two suspected of having problems with the tube early in the series, a partial separation of the gastrostomy from the abdominal wall in one adult, and gastrocolic fistula in one adult and one child. The last condition disappeared after removal of the gastrostomy tube in both patients. No leakage around the catheter, hemorrhage, peritonitis, or gastric outlet obstruction was encountered. This procedure provided a rapid, safe, and effective method for creating a feeding gastrostomy and did not require general anesthesia and laparotomy. Percutaneous endoscopic gastrostomy should become the method of choice for the creation of a feeding gastrostomy. ( Arch Surg 1983;118:913-914)

Carcinoma in Choledochal Cysts: Age-Related Incidence

Abstract: • We examined a case of malignant degeneration that occurred as a long-term complication of a choledochal cyst. Analysis of the literature shows the incidence of carcinoma varies with age at the initial appearance of symptoms. The child with a choledochal cyst that appears before 10 years of age carries a minimum risk (0.7%) of subsequent malignant degeneration compared with the patient in the second decade (6.8%) and older (14.3%). Data suggest that carcinoma may be readily overlooked at the time of choledochal cyst exploration. In light of the findings, an age-adjusted strategy for management should be used. ( Arch Surg 1983;118:986-988)

Metabolic Abnormalities Associated With Postoperative Organ Failure: A Redox Theory

Abstract: • We studied metabolic abnormalities of postoperative organ failure in 55 patients and classified them into four groups according to the postoperative changes in the arterial blood—ketone body ratio (acetoacetic acid–β-hydroxybutyric acid) reflecting hepatic mitochondrial redox potential: patients in group A had no decrease below 0.7, patients in group B had a transient decrease to 0.4, patients in group C had a progressive decrease to below 0.4, and patients in group D (terminal stage) had a decrease to below 0.25. All group A and B patients tolerated their operations well; the group C and D patients had multiple organ failure. In groups B, C, and D, plasma concentrations of alanine, proline, phenylalanine, and tyrosine were negatively correlated with the blood—ketone body ratio and the molar ratios between the plasma concentrations of branchedchain amino acids and aromatic amino acids were positively correlated with the blood—ketone body ratio. Hepatic energy deficit associated with decreasing blood-ketone body ratio may be the metabolic basis of postoperative organ failure. ( Arch Surg 1983;118:1245-1251)

Determinants of Organ Malfunction or Death in Patients With Intra-abdominal Sepsis: A Discriminant Analysis

Abstract: One hundred and six patients found at operation to have intra-abdominal sepsis were prospectively followed up to determine the incidence of organ malfunction and death. These outcomes were correlated with age, preexisting disease, underlying cause of sepsis, shock, nutritional status, and alcoholism. Organ malfunction occurred in 31 patients (29%), 19 (61%) of whom died. Two (3%) of 75 patients without organ malfunction died. Discriminant analysis revealed a significantly increased risk of death in patients with shock at any time, age greater than 65 years, alcoholism, bowel infarction, or malnutrition. A discriminant equation based only on preoperative variables correctly assigned the outcome of death or survival in 97 (92%) of the patients based on probabilities derived from this analysis. At present, this information is primarily of interest for researchers comparing outcomes in groups of patients, but with additional refinements it may become clinically useful for individual patients.

Effectiveness of a Surgical Wound Surveillance Program

Abstract: • A five-year surgical wound surveillance program included the following features: (1) observations were made by a trained nurse-surveyor; (2) all surgical services, without exception, were surveyed; (3) the nurse-surveyor reported directly to the Chief, Surgical Service; (4) all infected wounds and all suspected of harboring an infection were observed daily by the nurse-surveyor; (5) all wounds were inspected on the third and seventh postoperative days, at hospital discharge, and at a follow-up clinic visit; and (6) cultures were obtained from all infected wounds. Data concerning infections for all surgical services were published each month at the mortality-morbidity conference. The number of wounds closed primarily and the number of infected wounds were recorded, together with calculations of wound infection rates by operation class, for each surgical service and for the whole hospital. During the study, the rate of wound infections progressively declined. The overall incidence decreased from 3.5% before the study began to less than 1% at its conclusion. ( Arch Surg 1983;118:303-307)

Mycotic aneurysm. New concepts in therapy.

Abstract: • Fifty patients were treated for 52 mycotic aneurysms secondary to intravenous drug abuse. An initial misdiagnosis of cellulitis or abscess in 17% of the patients was corrected after arteriography or bleeding following operative drainage. There was no ischemia following ligation and excision of aneurysms of the radial, brachial, external iliac, deep femoral, and superficial femoral arteries. Excision of the common femoral artery in four patients and femoral bifurcation in 25 led to marked morbidity in 28 patients without simultaneous revascularization. Ischemia occurred in 53% of these patients; it was mild in 21% with claudication only. Severe, limb-threatening ischemia occurred in 32% and led to amputation in 21%. Six patients underwent artificial bypass, including one for absent back-bleeding at the time of ligation, four for immediate severe ischemia, and one for late ischemia. Two infected grafts were removed; another became thrombotic. Cultures were positive for 73% of aneurysms and blood of 46% of the patients. (Arch Surg1983;118:577-582)

Mycotic Aortic Aneurysms: A Reappraisal

Abstract: Mycotic aneurysms are uncommon but not rare lesions with potential for catastrophic hemorrhage or sepsis. They have been ascribed to bacterial endocarditis and, when present in the aorta, were termed "inevitably fatal" as recently as 1967. A 15-year review of the English-language literature on mycotic aneurysms showed that arterial trauma, concurrent sepsis, and depressed host immunity have become the cardinal "risk factors" in the development of these lesions. Conventional treatment of mycotic aortic aneurysms usually includes aortic ligation, aneurysmal excision, and extra-anatomic bypass grafting. Nevertheless, four of our patients with well-localized mycotic aortic aneurysms survived three to 54 months (mean, 40 months) after aortic excision and in situ prosthetic graft restoration of aortic continuity. This experience suggests that mycotic aortic aneurysms can be successfully treated, frequently by in situ grafting, if diagnosis and treatment are timely and aggressive.

Metastatic patterns in soft-tissue sarcomas.

Abstract: • In 242 patients with recurrent soft-tissue sarcomas, the most common sites of initial recurrence were the primary site in 47.5% of patients and the lungs in 38% of patients. Further recurrences in the course of the disease concerned the lungs, bones, liver, and brain. Total survival and survival after recurrence were influenced by the histologic type, which also affected the site of recurrence. In the management of local recurrence, a five-year disease-free survival rate of 38% was achieved with surgical treatment, while radiation or chemotherapy alone was ineffective. Local recurrences resulted in significantly higher survival rates than those involving other organs. The disease-free interval was a significant prognostic indicator of subsequent survival in the whole group of patients and among those with local recurrence. ( Arch Surg 1983;118:915-918)

Gauging the Severity of Surgical Sepsis

Abstract: • I developed a method for scoring the severity of a septic process, based on deteriorated functions in seven key organ systems of the body. The scoring system is numeric and recognizes that the risk to a patient rises geometrically as organ-system functions deteriorate step by step. The scoring system was validated by reviewing the clinical course of 30 patients with sepsis. Prognosis and hospital stay correlated well with individual scores. The scoring system offered more accurate comparisons in clinical studies of infected patients and helped follow up a patient with sepsis more accurately. (Arch Surg1983;118:1190-1192)

Antibiotic prophylaxis in surgical procedures. A critical analysis of the literature.

Abstract: The use of prophylactic antibiotics in surgery is widespread and often inappropriate. The lack of well-designed clinical studies partially explains the present confusion regarding the subject. We reviewed the literature in English on antibiotic prophylaxis through June 1982. Antibiotic prophylaxis reduces the incidence of wound infection after colorectal surgery, vaginal hysterectomy, and laryngeal and oropharyngeal resection for carcinoma, and in high-risk patients undergoing gastroduodenal or biliary surgery. In clean operations such as cardiac surgery, vascular procedures, and orthopedic surgery with placement of prostheses, the high morbidity associated with an infection justifies the use of antibiotics even though the risk of infection is small. There are conflicting data on the usefulness of prophylaxis in abdominal hysterectomy, cesarean section, noncardiac thoracic procedures, and urologic surgery. The effectiveness of prophylaxis in neurosurgery cannot be evaluated at the present time.

Preoperative Disseminated Intravascular Coagulation Associated With Aortic Aneurysms: A Prospective Study of 76 Cases

Abstract: A prospective study of 76 preoperative patients with aortic aneurysms was undertaken to determine the true incidence of associated disseminated intravascular coagulation (DIC). Although 39% of the patients showed a notable elevation of the fibrin split products level, only three had thrombocytopenia and a clinical bleeding diathesis, as well. Thus, clinically overt DIC occurred preoperatively in only 4% of the patients. All three patients had extensive aneurysms that involved the thoracoabdominal aorta. Preoperative fibrinogen levels in this series tended to be high-normal or elevated and were not good indicators of underlying excessive fibrinolysis. Hemostatic abnormalities, such as ecchymoses and petechiae, may be the key to the clinical diagnosis of DIC in preoperative patients with aortic aneurysms.

Total Thyroidectomy for Hürthle Cell Neoplasm of the Thyroid

Abstract: • The treatment of Hurthle cell neoplasms of the thyroid is controversial because of a lack of information about their clinical behavior and long-term follow-up. We reevaluated our experience of the past 33 years and compared our early experience with the more aggressive surgical approach used during the past ten years. We treated 62 patients with Hurthle cell neoplasm from 1949 through 1982. Twenty-six patients had benign neoplasms and 35 had malignant neoplasms proven by capsular or vascular invasion or nodal metastasis. Fourteen deaths were directly attributable to recurrent or metastatic disease, three among those with benign lesions. An aggressive surgical approach, involving total thyroidectomy or earlycompletion total thyroidectomy following lobectomy for all tumors histologically malignant or larger than 2cm, resulted in a lower recurrence rate (21%v59%) and fewer operations per patient (1.7v2.9) than in patients having other operations. (Arch Surg1983;118:529-532)

The desmoid tumor. IV. Choice of treatment, results, and complications.

Abstract: We analyzed the results of treatment in 89 cases of desmoid tumor (DT) with a follow-up period of 9.3 +/- 5.9 years. After surgical removal, the frequency of recurrence was equal, regardless of whether a complete removal of the tumor was obtained or not. After combined operation and radiotherapy, recurrence was higher. A higher frequency of recurrences was observed in extraabdominal (45%) v abdominal (10%) DTs. Recurrence was particularly high in the juvenile variety (70%), and low in the DTs of fertile women (11%). If a recurrence followed the first operation, multiple recurrences were the rule. Spontaneous regression was observed in 4% of the cases. The results emphasize that the basic biologic characteristics of this rare neoplasm have a greater impact on the results than the operation itself. A simple excision at the primary operation is recommended.

Exchange of Amino Acids by Muscle and Liver in Sepsis: Comparative Studies In Vivo and In Vitro

Abstract: • The amino acid "central fractional clearance rate" (CFCR), the ratio of the rate of amino acid entry into the extracellular pool to the size of the pool, is a measure of amino acid uptake and clearance by liver and other visceral tissues. In nine normal postabsorptive persons, the mean CFCR was 5%, compared with 21% in 31 seriously infected patients. For comparative purposes, biopsy specimens of liver and muscle were obtained for incubation. In infected patients, the rate of hepatic incorporation of tyrosine into protein was three times that in noninfected patients and correlated well with the CFCR. There was no significant difference in hepatic tyrosine oxidation. In muscle from infected patients, net protein degradation was six times that in noninfected patients. Incubated tissues from rats behaved similarly. Thus, accelerated transfer of amino acid from muscle to viscera for protein synthesis occurs in humans with sepsis, as it does in animals. The CFCR demonstrated the importance to survival of visceral amino acid uptake; it was 35% in surviving patients, and only 19% in those who died. (Arch Surg1983;118:167-175)

Femoropopliteal Graft Failures: Clinical Consequences and Success of Secondary Reconstructions

Abstract: • We examined 188 failed femoropopliteal grafts during a 16-year period to evaluate the clinical consequences of graft failure and success of secondary revascularization procedures. In limbs with grafts placed for claudication, conditions of 9% with failed grafts remained improved despite graft occlusion, 67% returned to preoperative status, and 24% showed worsened ischemia. One third of claudicants with failed grafts underwent secondary procedures. Ultimately, only 7% of failed grafts in claudicants resulted in amputation, and overall risk of limb loss in patients undergoing operation for claudication alone was low (2%). After failure of grafts performed for limb-salvage indications, 21% of limbs were still improved, and 79% reverted to limb-threatened status, more than half of the limbs undergoing subsequent reoperation. Overall, secondary attempts at reconstruction appeared justified. Although the five-year patency rate of such grafts was only 31%, limb salvage was achieved in 52%, with relatively low morbidity and mortality. ( Arch Surg 1983;118:1043-1047)

Risk of renal failure after major angiography.

Abstract: • In 400 patients who underwent major aortography, acute renal dysfunction (ARD) occurred in 11.3%. Of the group with normal renal function before the procedure, 8.2% had ARD and 0.8% required dialysis. Patients with prior abnormal renal function had a 41.7% incidence of ARD, and 8.3% required dialysis as a result of angiography. Vigorous intravenous hydration was used in all patients but did not completely prevent renal problems. Two risk factors not previously emphasized were the injection site (higher risk with abdominal aortic studies) and presence of congestive heart failure requiring treatment with digoxin. Other notable risk factors included contrast load and age. These results emphasized that even with modern contrast agents and application of current concepts of treatment, there remains a risk of renal injury with major angiography. ( Arch Surg 1983;118:1417-1420)

Severe Duodenal Injuries: Treatment With Pyloric Exclusion and Gastrojejunostomy

Abstract: • During a 12-year period, 313 patients with duodenal injuries were treated. Primary repair, pyloric exclusion, and gastrojejunostomy were used in 128 patients (41%) with severe duodenal and pancreaticoduodenal injuries, to reduce "duodenal" morbidity and mortality. The duodenal fistula rate was 2.2% overall, and 5.5% in the patients undergoing exclusion. Only two deaths were due to fistulas. Forty-two patients underwent upper gastrointestinal tract examinations after operation. In patients examined 21 days or more after operation, 94% had a patent pylorus. Marginal ulceration was infrequent (four patients), as were complications associated with the procedure (3%). Pyloric exclusion with gastrojejunostomy is a quick and simple procedure that allows return of pyloric patency and is associated with a low incidence of duodenal fistulas. When fistulas do develop, they are usually easily controlled and are associated with a low mortality. We believe pyloric exclusion with gastrojejunostomy to be the procedure of choice in patients with severe duodenal and pancreaticoduodenal trauma. (Arch Surg1983;118:631-635)

The Natural History of Carotid Plaque Development

Abstract: • Fifty-seven carotid artery plaques in 54 patients were harvested at surgery and studied. Preoperative noninvasive findings, arteriograms, clinical symptoms, surgical findings, and light microscopic findings were compared. Thirty-four patients had symptoms, 28 with TIAs and six with a previous stroke. Thirty-three of the 34 had intraplaque hemorrhage, and 28 had a connection between the hemorrhage and the arterial lumen. Eleven of the 21 asymptomatic patients also had an intraplaque hemorrhage, and one had a connection. In eight of the 11, noninvasive studies showed progression of disease. The development of an intraplaque hemorrhage appears to be an important factor in an innocent plaque's becoming a clinically relevant one. If a connection develops between the intraplaque hemorrhage and the arterial lumen, the patient may have TIAs. ( Arch Surg 1983;118:1048-1052)

Surgical Therapy for Small-Bowel Carcinoid Tumors

Abstract: Eighty-two symptomatic patients with carcinoid tumors of the small intestine were examined and treated over a 20-year period. Common clinical features included weight loss, diarrhea, and symptoms of intermittent bowel obstruction; 25 patients (30%) exhibited the carcinoid syndrome. Multiple carcinoid tumors occurred in 23 patients (28%), and hepatic metastases were present in 30 (37%). All patients underwent operation. The overall mortality was 7%, and the cumulative five-year survival rate was 59%. Two factors influenced prognosis after operation: hepatic metastases and incomplete resection. Other variables, including the sex and age of the patient and the size of the primary tumor, were of no additional prognostic value. Wide resection of the tumor, including regional lymph nodes, is indicated, regardless of the size of the primary tumor.

Biologic Behavior of Thin Malignant Melanomas With Regressive Changes

Abstract: In Reply. —The purpose of our short article was not to stir any controversy. The questions raised by Wolinsky and Silvers are well taken. We are aware of the pitfalls and drawbacks of retrospective studies. Our report was based on a study of all thin melanomas that were less than 0.76 mm in maximal thickness over a period of 25 years. Most of the patients were followed up for long periods, with a median follow-up of 76 months and an average follow-up of 102 months. Twenty-one of these patients were studied retrospectively and only 15 patients were studied prospectively. All primary melanomas prospectively studied were step sectioned and only 13 of 21 melanomas retrospectively studied were step sectioned. In eight patients the paraffin blocks were unavailable and we had to depend on two or three sections stained with hematoxylin-eosin for this review. It would be presumptuous to say that these

Straight Ileoanal Anastomosis v Ileal Pouch-Anal Anastomosis After Colectomy and Mucosal Proctectomy

Abstract: • The postoperative results of 50 patients who underwent straight ileoanal anastomosis after total colectomy and mucosal proctectomy were compared with those of 74 patients who underwent ileal pouch—anal anastomosis. No deaths occurred. Of the straight ileoanal anastomoses, 32% failed because of sepsis or diarrhea and necessitated abdominal ileostomy; only 1.3% failed in the pouch-anal group (P (Arch Surg1983;118:696-701)

Carcinoma of the bifurcation of the hepatic ducts.

Abstract: • Thirty-one patients with carcinoma of the bifurcation of the hepatic ducts were treated during a period of ten years. Sixteen of the 31 patients underwent resection. Left, right, and extended right lobectomy with the resection of the bilateral hepatic ducts, common hepatic and bile ducts were performed on 11, 1, and 3 patients, respectively. In a cirrhotic patient, resection of the left medial segment and bile ducts was carried out. The biliary tract was reconstructed using hepaticojejunostomy or hepatojejunostomy. Four of the 16 patients are still alive for six months, 1½ years, one year and 11 months, and two years and ten months, respectively. The remaining 12 patients died of recurrence at periods ranging from six months to three years and eight months. The treatment is now focused on increasing long-term survival. (Arch Surg1983;118:1147-1151)

Efficacy of Oral and Systemic Antibiotic Prophylaxis in Colorectal Operations

Abstract: A cooperative Veterans Administration study of the septic complication rate during large-bowel surgery was undertaken in two groups of patients. The first group received oral neomycin and erythromycin base plus parenteral placebo; the second, the oral antibiotics plus parenteral cephalothin sodium. During a five-year period, 1,128 patients were studied. The overall septic complication rate was 7.8% in patients receiving only oral antibiotics, and 5.7% in patients receiving both oral and parenteral antibiotics. This difference was not significant. The only significant finding was a greater incidence of fever of unknown origin in patients receiving only oral antibiotics. None of those patients were treated with additional antibiotics, and all fevers cleared spontaneously. There seems to be no discernible benefit from adding parenteral antibiotic prophylaxis when performing elective colon surgery if appropriate mechanical cleansing and oral neomycin and erythromycin therapy are employed.