Search

Competing Interests: Competing interests: We have read and understood BMJ policy on declaration of interests and have no relevant interests to declare.

Background: The uptake of minimally invasive oesophagectomy (MIO) in the UK has increased dramatically in recent years. Post-oesophagectomy diaphragmatic hernias (PODHs) are rare, but may be influenced by the type of approach to resection. The aim of this study was to compare the incidence of symptomatic PODH following open and MIO in a UK specialist centre., Methods: Consecutive patients undergoing oesophagectomy for malignant disease between 1996 and 2012 were included. A standardised, radical approach to the abdominal phase was employed, irrespective of the type of procedure undertaken. Patient demographics, details of surgery and post-operative complications were collected from patient records and a prospective database., Results: A total of 273 oesophagectomies were performed (205 open; 68 MIO). There were 62 hybrid MIOs (laparoscopic abdomen and thoracotomy) and six total MIOs. Seven patients required conversion and were analysed as part of the open cohort. Nine patients (13.2 %) developed a PODH in the MIO cohort compared with two patients (1.0 %) in the open cohort, (p < 0.001). Five patients developed hernias in the early post-operative period (days 2-10): all following MIO. Both PODHs in the open cohort occurred following transhiatal oesophagectomy. All PODHs were symptomatic and required surgical repair. CT thorax confirmed the diagnosis in 10 patients. Seven hernias were repaired laparoscopically, including two cases in the early post-operative period. PODHs were repaired using the following techniques: suture (n = 6), mesh reinforcement (n = 4) and omentopexy to the anterior abdominal wall without hiatal closure (n = 1). There were two recurrences (18 %)., Conclusions: The incidence of symptomatic PODH may be higher following MIO compared to open surgery. The reasons for this are unclear and may not be completely explained by the reduction in adhesion formation. Strategies such as fixation of the conduit to the diaphragm and omentopexy to the abdominal wall may reduce the incidence of herniation.

Introduction: Centralisation of oesophagogastric (OG) resectional services has been proposed to improve patient outcomes in terms of perioperative mortality and long-term survival. Centralisation of services occurred in Gloucester 5 years ago. The aim of this paper is to assess if local patient outcomes have benefited from centralisation., Methods: All oesophagogastric resections performed in our unit over a 15-year period (10-years pre-centralisation and 5-years post-centralisation) were assessed retrospectively. Patient demographics, pathological details and date of death were identified. Perioperative mortality (30 and 90 day) and estimated Kaplan-Meier survival was compared for cases performed pre- and post-centralisation of services., Results: 456 resections for cancer were performed in the 15-year period; 234 of these were performed pre-centralisation (mean 23.4, range 13-31) and 222 were performed post-centralisation (mean 44.4, range 40-50). Median survival rates for gastric cancer were 1.1 years pre-centralisation and 1.5 years post-centralisation (p = 0.147) and median survival for oesophageal cancer improved from 1.1 years to 2.1 respectively (p = 0.028). Combined OG 30-day mortality rates improved from 10.3% pre-centralisation to 3.6% post-centralisation (p = 0.006, Fisher's exact test)., Discussion: Centralisation of OG services in Gloucester has resulted in twice as many resections being performed locally. Median survival for patients with oesophageal cancer has increased by 1 year and the 30-day mortality rate following resection has reduced by almost two thirds. Although other factors (such as improvements in oncological treatments, staging and critical care management over the 15-year time period) have undoubtedly had roles to play in these improvements, the results of this study support the policy of centralisation of upper GI cancer services., (Copyright © 2012 Surgical Associates Ltd. Published by Elsevier Ltd. All rights reserved.)

Background: There is little available evidence on the optimal management of recurrent inguinal hernia, particularly if the original procedure involved the use of mesh. This study was a review of recurrent hernia repair in a district hospital, involving both laparoscopic and open procedures., Methods: The case notes of all patients who had a repair of a recurrent hernia between 1991 and 2000, inclusive, were examined; 171 procedures were included. Where known, the original repair was a nylon darn in 31%, mesh repair in 18%, and laparoscopic repair in 8%., Results: The recurrent hernia was repaired using a Lichtenstein open mesh technique in 63% and by the totally extraperitoneal (TEP) method in 22%. Complication rates were highest after emergency surgery (all had open surgery), where 71% had complications and one patient died. For elective repairs, complication rates were similar after open (13%) and TEP (8%) repairs. The duration of hospital stay was also similar (1.2 vs 1.3 days, respectively), and a single recurrence was seen in each group. Patients with recurrence after primary mesh repair were also managed by both techniques with similar results. Open re-operation for mesh failure was technically straightforward., Conclusions: Most recurrent hernias are still repaired by open techniques. There was no convincing evidence of different outcomes for open and TEP repairs in this review. Even when the original hernia repair involved the use of mesh, further open repair by an experienced surgeon is justified.

Background: With the advent of laparoscopic cholecystectomy, endoscopic retrograde cholangiography (ERC) is frequently employed before operation to detect common bile duct stones. However, this is an invasive technique with recognized complications. This study evaluated the accuracy of magnetic resonance cholangiography (MRC) to detect choledocholithiasis in selected patients before laparoscopic cholecystectomy., Methods: Patients scheduled for elective laparoscopic cholecystectomy, with risk factors for common bile duct stones, underwent MRC followed by ERC or peroperative cholangiography., Results: Over a 2-year period 40 patients were studied. MRC had a sensitivity of 88 per cent, specificity of 93 per cent, positive predictive value of 78 per cent and negative predictive value of 97 per cent for the detection of common bile duct stones., Conclusion: MRC is a simple non-invasive method for preoperative screening for common bile duct stones in at-risk patients. In this study it would have reduced the need for ERC by three-quarters.

The 4th year of the Avon breast screening programme comprises two distinct groups: those called for screening for the first time (prevalent group) and those who were initially screened 3 years earlier (incident group). The cancer detection rate, stage of disease and rate of interval cancers in these patients have been compared. For the prevalent groups of year 1 and year 4 there was no statistically significant difference in the cancer detection rate, proportion of small tumours or node positivity. For the prevalent and incident groups of year 4, there was no statistically significant difference in the cancer detection rate or proportion of small tumours. There were significantly fewer node-positive tumours in the incident group (5/45 vs 8/23; P < 0.05). Fifty-six interval cancers presented in the 3-year period between years 1 and 4 of screening; 28 (50%) after 24 months. The screening programme may result in tumours being detected at an earlier stage, but this may be offset by the high rate of interval cancers. This suggests that the time between screens may need to be reduced to 2 years.

There has been concern about the number of interval cancers which have been detected within the National Breast Screening Programme. A series of 134 women presenting with interval cancers was studied by prospective audit and the rate and radiological classification of the tumours determined. The cancers were classified as true (67), false-negative (22), unclassifiable (28), occult (12), and minimal sign (5). The interval cancer rate did not achieve the new National Guidelines in either the first 2 years or the 3rd interval year. The false-negative cancers presented mainly in the 1st interval year, whereas the true cancers were predominantly confined to the 2nd and 3rd years. These data suggest that alterations to the screening programme may be beneficial. It may be, however, that the programme is still on the learning curve and this should be taken into account when interpreting these data.

The expression of components of the plasminogen activator system was investigated in patients with oesophageal carcinoma. Tumour and normal mucosa were obtained from resected oesophageal carcinomas and antigens were measured by enzyme-linked immunosorbent assay. Median levels of urokinase plasminogen activator (uPA) and the uPA receptor were higher in carcinoma than in matched normal mucosa (squamous cell carcinoma: uPA 4.05 versus 0.66 ng antigen per mg protein, uPA receptor 1.95 versus 0.50 ng/mg, n = 10, P < 0.05; adenocarcinoma: uPA 2.16 versus 0.61 ng/mg, uPA receptor 2.01 versus 0.49 ng/mg, n = 8, P < 0.05). Tissue plasminogen activator (tPA) level was lower than control values in squamous cell carcinoma but not in adenocarcinoma (1.97 versus 4.70 ng/mg, P < 0.05). There was no difference in plasminogen activator inhibitor (PAI) 1 level between carcinoma and normal mucosa. The PAI-2 level was lower than that in normals in adenocarcinoma only (6.0 versus 64.77 ng/mg, P < 0.05). These data support the hypothesis that membrane-bound uPA has a role in the breakdown of extracellular matrix in invasive oesophageal carcinoma.

Over a 4-year period, a direct-access fibreoptic sigmoidoscopy service was evaluated prospectively. In all, 756 patients were referred (median age 58 years, range 18-91 years). The principal indications were rectal bleeding (45%) or change of bowel habit (28%); both features were present in 13%. Abnormalities were present in 68% of examinations. Major disease was identified in 22% (carcinoma 7.0%, adenoma 6.3%, inflammatory bowel disease 8.3%) and minor disease in 53% (haemorrhoids 36.8%, severe diverticular disease 10.9%, non-adenomatous polyp 3.4%, perianal disease 1.4%). In patients under 40 years of age, major disease was rare (one carcinoma, three adenomas). Of the patients, 21% underwent barium enema for incomplete examination or suspected additional disease. No additional major disease was identified, but one carcinoma found in a patient with stricture. These data show that a direct-access fibreoptic sigmoidoscopy service produces a high diagnostic yield and may be of value to both patients and general practitioners in expediting a clinical colorectal service.

Intra-abdominal adhesions develop in over 90 per cent of patients undergoing laparotomy. Peritoneal fibrinolysis is believed to be important in the pathophysiology of adhesion formation. This study investigated the fibrinolytic response of postoperative peritoneal fluid in 12 patients undergoing elective laparotomy. There was a significant reduction in the plasminogen activating activity to undetectable levels at 24 h, which was sustained at 48 h (P < 0.05). While there was an early reduction in the concentration of tissue plasminogen activator (median 40.0, 28.2, 16.3 and 31.9 ng/ml at 2, 6, 24 and 48 h respectively; P < 0.05), the abolition of functional fibrinolytic activity appeared to be secondary to a marked increase in the concentration of plasminogen activator inhibitor (PAI) 1 (median 86, 196, 800 and 730 ng/ml at 2, 6, 24 and 48 h respectively; P < 0.05) and PAI-2 (median less than 6, 12, 155 and 245 ng/ml at 2, 6, 24 and 48 h respectively; P < 0.05). This reduction in the plasminogen activating activity of peritoneal fluid may favour the formation of permanent fibrous adhesions following surgery.

The deposition of fibrin in the peritoneal cavity leads to fibrous adhesion formation. Recombinant tissue plasminogen activator (rtPA), delivered locally, was investigated as a method of preventing adhesion formation. Six standardised areas of peritoneal ischaemia were formed in each of 36 male Wistar rats randomised to three intraperitoneal treatments: (A) no treatment control; (B) carboxymethylcellulose gel; (C) rtPA-carboxymethylcellulose gel combination. At 1 week all animals underwent relaparotomy and the number of ischaemic sites with an adhesion counted by an independent observer. rtPA-treated animals formed fewer adhesions compared with gel alone or controls (median number of adhesions 1.5 versus 2.5 versus 5, P < 0.001, ANOVA). Intraperitoneal rtPA in a slow-release formulation is able to reduce adhesion formation significantly in an animal model and may prove to have clinical benefit.

Objective: Measurement of the fibrinolytic response of the peritoneum to experimental peritonitis and ischaemia., Design: Controlled study, Setting: Academic surgical unit, UK MATERIAL: Male Wistar rats, Interventions: Peritoneal injuries were caused in four groups of male Wistar rats (n = 35 in each group): (1) control group ("open and close" laparotomy); (2) bacterial peritonitis (mixed faecal flora); (3) chemical peritonitis (10 mg/ml tetracycline) and; (4) ischaemic peritoneum (ligated peritoneal buttons). Peritoneal biopsy specimens were taken from five animals in each group at seven time intervals and plasminogen activating activity (PAA) measured by fibrin plate assay., Results: Compared with the control group the three peritoneal injuries produced a uniform reduction in PAA during the first 6 and 12 hours: at 6 hours the median PAA was 0.029 IU/cm2 for bacterial peritonitis, 0.021 IU/cm2 for chemical peritonitis, and 0.05 IU/cm2 for ischaemic peritoneum compared with 0.112 IU/cm2 for the control group; p < 0.001, ANOVA. At 12 hours the median PAA was 0.024 IU/cm2 for bacterial peritonitis, < or = 0.014 IU/cm2 for chemical peritonitis, and 0.05 IU/cm2 for ischaemic peritoneum compared with 0.112 IU/cm2 for the control group; p < 0.001, ANOVA. There then followed a rebound peak in all groups, maximal at 4-7 days, before a return to baseline values at two weeks., Conclusion: Peritoneal fibrinolysis was appreciably inhibited after three different standardised peritoneal injuries. The data support the hypothesis that there is a single pathophysiological mechanism of adhesion formation.

Human mesothelial cells synthesize plasminogen activator inhibitor (PAI) 1 in inflamed peritoneal tissue. The role of tumour necrosis factor (TNF) in the mediation of this response was studied. Postoperative peritoneal drain fluid contained both TNF and PAI-1. Peak levels of TNF at 4 h (median 271 pg/ml) preceded a rise in PAI-1 concentration, which peaked at 18 h (median 943.1 ng/ml). Thus TNF may mediate increased PAI-1 release in inflamed peritoneum. TNF significantly increased the mean(s.e.m.) release of PAI-1 by human peritoneal mesothelial cells in vitro at 4 h (control 1.84(0.17) ng/micrograms versus TNF 2.37(0.17) ng/micrograms, P < 0.05), 6 h (2.53(0.09) versus 3.88(0.46) ng/micrograms, P < 0.05), 18 h (0.50(0.02) versus 1.04(0.11) ng/micrograms, P < 0.05) and 24 h (0.87(0.05) versus 1.35(0.11) ng/micrograms, P < 0.05). TNF may be an important mediator of PAI-1 production by human mesothelial cells during peritoneal inflammation.

Coagulopathy is a well recognised complication of peritoneovenous shunting for ascites. The relative contributions of primary fibrinolysis and disseminated intravascular coagulation remain controversial. Plasminogen activating activity was significantly lower in malignant ascites (n = 10, median < 0.02 (range < 0.02-1.26) IU/ml) than in alcoholic ascites (n = 10, 1.07 (0.30-1.49) IU/ml) (p < 0.05). Fibrinolytic activity was determined by a balance between tissue plasminogen activator and plasminogen activator inhibitor-1. There was no significant difference between the two groups in the concentration of tissue plasminogen activator (34 (12-64) ng/ml in malignant ascites v 29 (12-43) ng/ml in alcoholic ascites), but the concentration of plasminogen activator inhibitor-1 was significantly higher in malignant ascites (736 (213-1651) ng/ml) than in alcohol ascites (29 (12-43) ng/ml) (p < 0.05). Malignant ascites contained significantly higher concentrations of urokinase (0.7 (< 0.1-1.3) ng/ml v 0.2 (< 0.1-0.6) ng/ml in alcoholic ascites) and plasminogen activator inhibitor-2 (33 (< 6-140) ng/ml v 9 (< 6-28) ng/ml alcoholic ascites). The plasminogen activating activity of alcohol ascites may lead to primary fibrinolysis after peritoneovenous shunting. The considerably lower activity found in malignant ascites may explain why coagulopathy after shunting is less pronounced in this group of patients.

A questionnaire survey of general surgeons was undertaken to estimate the incidence of abdominal adhesions as a clinical problem and to establish current attitudes to the treatment of adhesional bowel obstruction and the prevention of adhesion formation. Replies were received from 362 of 416 surgeons (response rate 87%). In all, 76% (95% confidence limits, 72%-80%) of surgeons operate upon at least two patients each year with adhesional small bowel obstruction and 31% (26-36) operate upon more than five patients. More than five patients are admitted each year with suspected adhesional bowel obstruction by 64% (59-69) of surgeons and 35% (30-40) of surgeons found adhesions to be a problem during a non-adhesion-related laparotomy in more than five patients each year. This represents an estimated annual incidence of between 12,000 and 14,000 adhesion-related clinical problems in the United Kingdom. A number of preventive measures, such as the wearing of starch-free gloves (78% of surgeons; 95% CL 74%-82%), peritoneal lavage (68%; 63-73) and placement of the omentum beneath the wound closure (90%; 87-93) are generally accepted, whereas routine wetting of swabs (39%; 34-44) and the role of non-essential adhesiolysis (49%; 44-54) are controversial. Routine small bowel plication (1%; 0-2) and intubation (2%; 0-5) are rarely used. This survey gives an indication of the large burden on patients and the health services caused by abdominal adhesions, and demonstrates surgeons' wide variety of approaches to both the treatment and prevention of adhesion formation.

Fibrinolysis in peritoneal tissue may play a role in the development of intra-abdominal adhesions. The plasminogen-activating capacity of human peritoneum results largely from the presence of tissue plasminogen activator (tPA). Inflammation reduces peritoneal plasminogen-activating activity and leads to the appearance of plasminogen activator inhibitor (PAI) type 1. The role of PAI-2 in the inhibition of peritoneal fibrinolysis during inflammation was investigated in this study. The plasminogen-activating activity of peritoneal biopsy homogenates (seven inflamed, seven normal), measured using a fibrin plate technique, was reduced in inflamed compared with normal tissue (median < 0.07 versus 13.9 units/cm2, P < 0.01); tPA antigen levels were not significantly different (median 1.02 versus 1.34 ng/ml). PAI-1 and PAI-2 antigens were not detected in normal human peritoneum but were present in inflamed peritoneum (median concentration 8.8 ng/ml for PAI-1, 26.7 ng/ml for PAI-2). These inhibitors may be important factors in adhesion formation by contributing to the abolition of peritoneal plasminogen-activating activity.

The effect of a management protocol incorporating the selective use of fine catheter peritoneal cytology (FCPC) and laparoscopy on the unnecessary appendicectomy rate was studied in adult patients (> or = 16 years) treated at one district general hospital over an 11-month period. Appendicectomy was performed on 62 adult patients managed according to this protocol, six (10%) of whom had a histologically normal appendix and no other acute condition requiring surgery. A further 57 patients underwent appendicectomy after standard clinical assessment and investigation without the use of FCPC or laparoscopy. Nineteen (33%) of these patients had a histologically normal appendix removed, with no other acute condition requiring surgical treatment. The selective use of FCPC and laparoscopy significantly reduced the unnecessary appendicectomy rate from 33% to 10% (chi 2 = 10.0, P < 0.005). The more widespread use of these techniques in patients with suspected appendicitis is therefore recommended.

We have reviewed the indications for and outcome of surgery in 147 patients who were seropositive for human immunodeficiency virus (HIV), 100 of whom have developed acquired immunodeficiency syndrome (AIDS). There were 256 operations; the commonest indications were anorectal conditions (34%), central venous access (21%), lymph node and soft tissue biopsy (15%) and an important minority underwent laparotomy (4%). Complications occurred in 20% of operations and repeat procedures were required in 35 patients. Both were of equal frequency in the HIV and AIDS populations. Most operations were therefore minor, and achieved satisfactory results with an acceptable morbidity. The possibility of HIV-related infection or malignancy affected the diagnostic and therapeutic approach, particularly in those considered for anorectal surgery or laparotomy. With increasing numbers of HIV-infected patients, knowledge of the types of surgery required and the likely outcome is important to enable provision of a safe and effective surgical service.

One hundred patients undergoing abdominal surgery were randomly allocated to have their skin wound closed by either polypropylene (Prolene) or polydioxanone (PDS) by the subcuticular method. Ninety-one patients completed the study and were assessed at discharge and again 3 months later for wound complications and cosmetic result. No statistically significant differences were found between the two sutures; overall wound infection rate was 5.5% and cosmetic result was equally acceptable for the two suture materials. This study indicates that an absorbable subcuticular suture may be safely used in all classes of wound.

Laparoscopy and fine catheter peritoneal cytology (FCPC) have been advocated as aids in the assessment of acute abdominal pain. In all, 411 patients admitted to a district general hospital during a 10-month period were managed using a standard protocol incorporating these techniques. After initial assessment by a surgical registrar, 151 patients were excluded from further progress through the protocol (age less than 16 years, definite diagnosis made or contraindication to FCPC. The remaining 260 patients were placed in one of four management groups: (A) urgent operation (23 patients); (B) 'look and see' (40 patients); (C) 'wait and see' (59 patients); (D) urgent operation not indicated (138 patients). Eighty-eight of 99 patients (88 per cent) in groups B and C, where the need for operation was uncertain, underwent successful FCPC and 39 patients (39 per cent) underwent laparoscopy. In these patients the initial registrar management decision proved to be incorrect in 33 cases (33 per cent), but by following the protocol the number of management errors actually made was reduced to 13 (13 per cent, P less than 0.001). This would have been reduced to 8 per cent if the protocol had not been violated in five patients. This study demonstrates the effectiveness of a protocol using FCPC and laparoscopy to improve the management of patients with acute abdominal pain.

The prevalence of human immunodeficiency virus (HIV) infection is increasing. Risks of hospital transmission, though small, will consequently also increase. Precautions taken during invasive procedures, particularly surgery, and in the handling of specimens from patients diminish this risk. As HIV antibody screening is not permissible, we have assessed the possible efficacy of a routine questionnaire to identify high-risk groups. Of 525 questionnaires given to routine surgical patients, 506 (96%) were completed. Twenty-eight (5.5%) indicated that they were in a high-risk group. High-risk patients were more likely to present with anorectal disease, to be male and to be young compared with patients from the remainder of the sample. The questionnaire was acceptable to 97% of patients although to a significantly smaller number (82%) of high-risk patients. Routine HIV antibody testing would be acceptable to 88% of low-risk patients but to only 60% of high-risk patients. Questionnaire assessment is an acceptable and perhaps more reliable method of assessing HIV risk status than the other currently available options. It has to be accepted that it will never result in complete identification but we recommend this method as one approach to the problem faced by the surgical team.

The prevalence of HIV in the UK has been estimated to be 1 in 1000 of the population. Surgeons are at particular risk of occupational transmission from infected blood. To determine the effect of HIV on surgical practice we sent a questionnaire to 681 general surgeons in England and Wales; 450 replied (66%). Of those who replied, 42% were aware of having operated on an HIV-infected patient at least once, and 28 had recognised self-injury in such circumstances; 79% attempted to identify HIV-infected patients preoperatively, though many depended on clinical suspicion alone, which is known to be unreliable. Of those who had operated on a seropositive patient, 90% reported taking special precautions to avoid blood contact and minimise sharps injuries for such cases. The majority wore double gloves, eye protection and fluid-resistant gowns, but only a minority reported changes in surgical technique. Half had made no changes in procedures or technique when operating on patients not identified as being at risk of HIV infection. Among a wide variety of comments made by the surgeons, the commonest was a call for facilitation of HIV testing prior to surgery. This survey indicates that surgery on HIV-infected patients is not restricted to specialist centres. We review the means of identifying HIV-infected patients, the precautions that can be taken to minimise HIV transmission during surgery, and the possible influences of HIV status on surgical decisions. We conclude that the prevalence of HIV among surgical patients is being underestimated at present, that several simple changes in surgical technique should be adopted generally, and that there is limited value in preoperative HIV testing, though this may become more useful in the foreseeable future.

The mechanisms leading to reduction of peritoneal fibrinolytic activity in conditions that are associated with the formation of intra-abdominal adhesions were studied. Tissue plasminogen activator was found, by antibody inhibition techniques, to be the activator of fibrinolysis in homogenates of control peritoneum (n = 6). Homogenates of control (n = 10) and inflamed peritoneum (n = 10) were analysed. Plasminogen activating activity was much lower in inflamed peritoneum (median 0.07 IU/cm2) than in control tissue (median 12.0 IU/cm2) (p less than 0.001). Levels of tissue plasminogen activator and alpha 2-antiplasmin were similar in both control and inflamed tissue. Plasminogen activator inhibitor-1, not detectable in control peritoneum, was present in inflamed tissue and might be the reason for the reduction in functional fibrinolytic activity.

We have conducted three in vitro experiments (using cadaver internal mammary artery as a model for crural artery), which establish the advantage of the Miller collar and Taylor patch over direct PTFE-artery anastomoses: (1) A new method of simultaneously measuring longitudinal and circumferential elasticity in human saphenous vein (n = 16) has demonstrated the latter to be anisotropic (i.e.: greater longitudinal than transverse compliance, P less than 0.002). This suggests that benefit may be gained by aligning the saphenous vein's longitudinal axis with that of the arteriotomy--a feature of both vein interposition anastomoses. (2) We have compared maximal pulsatile and constant flow across each type of anastomosis (n = 10 of each) using a flow model incorporating standard pressure, viscosity, graft and outflow vessel length, and anastomotic angle. This experiment has demonstrated loss of anastomotic reliability only in direct PTFE-artery anastomoses (i.e. no correlation between flow and vessel diameter, r = 0.04). (3) Casts of the internal anatomy of the anastomoses (n = 10 of each) have demonstrated a greater degree of oval distortion of the toe of the outflow vessel of direct PTFE-artery anastomoses than either of the other techniques (P less than 0.01). This experimental evidence suggests that both of these vein interposition techniques optimise the mechanical properties of saphenous vein and protect small arteries from anastomotic distortion. These benefits do not occur with direct PTFE-artery anastomoses.

A 67-year-old woman developed three separate pelvic malignancies forty years after undergoing radiotherapy for a squamous cell carcinoma of the cervix. Although previous exposure to ionising radiation is associated with certain malignancies, review of the literature indicates that there is no strong evidence to support such an association with urothelial or colorectal malignancy.

Fine catheter aspiration cytology of the peritoneal cavity was performed successfully in 61 patients admitted with acute abdominal pain. Aspirates were examined microscopically and the percentage of neutrophils in the specimen counted. In patients in whom the clinical need for operation was certain (n = 25) all patients required surgery and the peritoneal neutrophil count was greater than 50 per cent. In patients in whom the clinical need for operation was uncertain (n = 36) 19 patients required operation: the peritoneal neutrophil count was greater than 50 per cent in 18 and in one patient with ectopic pregnancy fresh blood was aspirated. Of the 17 patients not requiring operation the peritoneal neutrophil count was less than 10 per cent in 15 and greater than 50 per cent in two patients (both had acute pelvic inflammatory disease). This study confirms peritoneal cytology as a useful adjunct to decision making in those patients with acute abdominal pain in whom the decision to operate is in doubt.

Acute abdominal pain continues to provide not only a large workload for the general surgeon but also many diagnostic and management problems. Many different techniques have been introduced over the past two decades to help in the management of the acute abdomen and this review considers their relative claims to become incorporated into the process of clinical decision-making. The evidence in support of formally structured patient interview pathways with or without computer-aided diagnostic programs is now overwhelming and should become routine. Both laparoscopy and peritoneal cytology have an important role to play in the management of patients in whom the decision to operate is in doubt, and a combination of the two would be complementary. Ultrasonography has become increasingly popular for investigating the acute abdomen, and results from specialist centres are impressive. However, the problems of operator variation and the difficulties in providing a 24-h service will probably prevent it from becoming a first-line investigation in most hospitals. Although plain radiography has been available for many years, its routine use in the management of the acute abdomen remains controversial. Recent studies have confirmed that contrast radiography is an important adjunct to decision-making, particularly in the management of large bowel obstruction, and there is increasing evidence to support its use in suspected small bowel obstruction, perforated peptic ulcer and acute diverticular disease.

A prospective study has been undertaken of 321 patients with acute abdominal pain admitted to hospital under one surgical firm over a period of 21 months. During the first 10 months patients were classified on admission according to the perceived need for operation, with laparoscopy being performed on all those in whom the need for operation was uncertain. In addition, all women with suspected appendicitis underwent laparoscopy because previous studies by us and others have demonstrated a high error rate in this group. During the second 11 months of the study a similar system of classification and procedure was used but the patient's initial assessment was entered on a structured data sheet. After the patient had been discharged home this information was entered into a computer-aided diagnosis program. Hypothetical retrospective computer-aided decisions were then made about patient management. The final management error rate (correct decision to operate or not) was compared with the actual error rate using the clinical system. The final overall error rate in the first 10 months was 11 out of 163 patients and this was improved to 3 out of 158 in the second 11 months of the study by the addition of the structured data sheet to selective laparoscopy. A management policy based entirely on diagnostic probabilities taken from the computer-aided diagnosis program would have produced an error rate of 26 out of 158. We conclude that in the management of the acute abdomen a policy based on clinical decision combined with selective laparoscopy may be superior to one based on diagnostic probabilities alone. Further improvement in results follows the introduction of a structured data sheet for initial data collection.