431
tangible and visible effects, such as reduction in frequency, volume, and duration of diarrhoea, as seen by the users, have been proved beyond doubt. Food-based ORT opens up great possibilities in the treatment of diarrhoea and the
most
child survival. The International Centre for Diarrhoeal Disease Research, Bangladesh, is
supported by Aga Khan Foundation, Arab Gulf Programme, Australia, Bangladesh, Belgium, Canada (Canadian International Development Agency and the International Development Research Centre), the Ford Foundation, Japan, the Norwegian Agency for International Development, Saudi Arabia, the Swedish Agency for Research Cooperation with Developing Countries, Switzerland, the United Kingdom, the United Nations Children’s Fund, the United Nations Development Programme, the United States Agency for International Development, and the World Health Organisation.
Correspondence should be addressed to A. M. M., Aga Khan University, Stadium Road, PO Box 3500, Karachi 74800, Pakistan.
AM, Sarker SA, Hossain M, Molla A, Greenough WB III. Rice powder electrolyte solution as oral therapy in diarrhoea due to Vibrio cholerae and Escherichia coli. Lancet 1982; i: 1317-19. 3. Rahman ASM, Bari A, Molla AM, Greenough WB III. Mothers can prepare and use rice salt oral rehydation solution in rural Bangladesh. Lancet 1985; ii: 539-40. 4. Patra FC, Mahalanabis I, Jalan KN, et al. Is rice electrolyte solution superior to glucose electrolyte solution in infantile diarrhoea. Arch Dis Child 1982; 57: 910-12. 5. Molla AM, Ahmed SM, Greenough WB III. Rice based oral rehydration decreases the stool volume in acute diarrhoea. Bull WHO 1985; 63: 751-56. 6. Mata L. The control and prevention of diarrhoeal disease at a national level. In: ICORT II proceedings: second international conference on oral rehydration therapy. Washington DC, December 10-13, 1985. 7. Kassem AS, Madkour AA, Massoud BZ, et al. Loperamide in childhood diarrhoea: a double blind controlled trial. J Diarrhoeal Dis Res 1983; 1: 10-15. 8. Satha Krishnan BR, Sankar Narayan VS. Rice water solution in diarrhoeal dehydration. Indian JPediatr 1985; 52: 479-82. 9. Population information program, 1985. Oral rehydration therapy (ORT) for childhood diarrhoea. Population reports series L. Population information program. Johns Hopkins University, Baltimore, MD. 2 (i) 41-76. 10. Fabricant S. Distribution and logistics. In: ICORT II proceedings: second international conference on oral rehydration therapy. Washington DC, Dec 10-13, 2. Molla
1985. United Nations Department of International Economic and Social Affairs. Population division. World Population and its age-sex composition by country. 1950-2000, New York: UN, 1984: 188. 12. Kielman AA, Mobarak AB, Hammy MT, et al. Control of deaths from diarrhoeal disease in rural communities. Trop Med Parasitol 1985; 36: 191-98. 13. Molla AM, Molla A, Rohde J, Greenough WB III. Turning of diarrhoea: the role of food and ORS. J Pediatr Gastroenterol Nutr 1989: 8: 81-84. 11.
REFERENCES
Development of an improved formulation of oral rehydration salts (ORS) with anridiarrhoeal and nutritional properties. Geneva: World Health Organisation, 1985: CCD/85.3
1. Mahalanabis D.
Hospital Practice ENDOSCOPIC SPHINCTEROTOMY IN 1000 CONSECUTIVE PATIENTS DINO VAIRA COLIN AINLEY STEPHEN WILLIAMS STUART CAIRNS PAUL SALMON CHRISTOPHER RUSSELL
LUIGI D’ANNA
JOHN DOWSETT JOHN BAILLIE JOHN CROKER PETER COTTON ADRIAN HATFIELD
Departments of Gastroenterology and Surgery, The Middlesex Hospital, London W1N 8AA
Summary
Between
1983
and because many require urgent surgical bileduct clearance with its associated risks.8-11 Since the introduction of endoscopic sphincterotomy, various techniques have been used to improve the results, and include prophylactic antibiotics, which are now used routinely. The success rate of the procedure can be increased by needle-knife sphincterotomy, 12 and by a combined percutaneous and endoscopic technique.13 If clearance of bileduct stones fails, urgent surgery may be avoided by
endoscopic drainage through an endoprosthesis or a nasobiliary drain.14 Have these novel techniques reduced morbidity and mortality rates? We describe the results of endoscopic sphincterotomy in a series of 1000 consecutive patients.
and
1988, endoscopic attempted on 1000 sphincterotomy consecutive patients with a clinical diagnosis of bileduct stones in a centre with a policy to establish immediate bileduct drainage for retained stones. Endoscopic cholangiography was successful in 985 patients, of whom was
782 had visible stones and 203 had a dilated bileduct but no visible stones. Endoscopic sphincterotomy was successful in 975 of these patients, with eventual bileduct clearance in 674 of 772 patients (87·3%) with visible stones; immediate bileduct drainage was achieved in 160 of the 161 patients (99%) in whom bileduct clearance failed at the first attempt. Overall, 771 of 797 patients (96·7%) with visible bileduct stones had successful bileduct clearance or drainage. Complications occurred in 6·9%, with a 30-day mortality rate of 1·2%, but procedure-related mortality was only 0·6%. INTRODUCTION
ENDOSCOPIC sphincterotomy is now a common for patients with bileduct stones, especially in the elderly or unfit. Sphincterotomy is achieved in 90-98% of attempts, with clearance of bileduct stones in 80-95%.1,2 Complication rates of 5-20% and mortality rates of 03-5% have been reported1-3 and, surprisingly, complication rates are not increased in patients with diverticula.4 At greatest risk are patients with retained stones, both from sepsis,3,5-7 treatment
PATIENTS AND METHODS
Between November, 1983, and March, 1988, endoscopic sphincterotomy was attempted on 1000 consecutive patients with a clinical diagnosis of common bileduct stones (mean age 70-6 years, range 20-100). Endoscopic cholangiography showed bileduct in 772 and a dilated bileduct but no visible stones in 203 age 704 and 71 3 years, respectively). The indication for endoscopic sphincterotomy in the two groups was painful jaundice (83% and 70%, respectively; p > 0-05), painless jaundice (7-4% and or pain without jaundice (13-2% and 23-1%; 6,4%; p>0-05), p < 0-01). 714 patients had an intact gallbladder, and 286 had had a cholecystectomy, of whom 70 were referred with a T-tube in the bileduct after a retained stone was found at postoperative stones
(mean
cholangiography. All patients received prophylactic intravenous antibiotics before, and for 24-36 h after, sphincterotomy. The prothrombin ratio was within 3 s of control values, or corrected to that time by use of vitamin K or fresh frozen plasma before the procedure. We used Olympus ’JF-IT 10’ or ’JF-IT 20’ and ’TJF-20’ duodenoscopes (Keymed Ltd, Southend, Essex) and Erlangen pull-type sphincterotomes (Keymed). Bileduct clearance by use of stone baskets, balloon catheters, or both techniques was attempted at the initial procedure. If sphincterotomy failed, access was improved by needle-knife sphincterotomyl2 or, in selected patients, by a combined percutaneous and endoscopic technique." If bileduct clearance failed because of difficult access, size of stones, or a distal common bileduct stricture, immediate drainage was established by insertion of either a nasobiliary drainage catheter or of a pigtail
432
endoprosthesis.14 Endoscopic bileduct clearance was attempted up to
3 times. Results
were prospectively entered into a computer program (’Pedro’, Keymed). Follow-up was by review of inpatient case notes and by postal questionnaire to consultants and general practitioners who referred patients. Data were available for at least 30 days, or longer if a complication occurred, in 950 patients. The X2 test was used for statistical analysis.
RESULTS
782 patients had bileduct stones seen on endoscopic cholangiography and 203 patients with a clinical diagnosis of
bileduct stones had a dilated common bileduct but no visible stones. In 15 patients, with strong clinical and radiological evidence of bileduct stones, endoscopic cholangiography was unsuccessful (10 of these patients had duodenal diverticula): all had surgery, which was successful in 13, but 2 patients, aged 85 and 89 years, died during operation. Of the 985 patients in whom cholangiography was possible, endoscopic sphincterotomy was successful in 830 at the first attempt, in 108 of 155 at the second attempt, and in 37 of 47 at the third attempt (overall cumulative success rates of 83-0%, 93-8%, and 97-5%, respectively). Needle-knife diathermy permitted sphincterotomy in 17 of 20 patients and a combined percutaneous and endoscopic technique was successful in 8 of 9 patients. Endoscopic sphincterotomy was unsuccessful in 10 patients in whom the cholangiogram showed bileduct stones, 7 of whom had duodenal diverticula: surgery was successful without complications in 7, and 3 have remained well without intervention. Bileduct clearance was achieved at the first attempt in 611 of 772 patients with bileduct stones in whom endoscopic sphincterotomy was successful, and in a further 63 of 161 at a subsequent attempt after a period of temporary bileduct drainage (32 nasobiliary catheter drainage, 31 temporary pigtail endoprostheses). The overall success rate of bileduct clearance was thus 674 of 797 patients with stones (84-5%). Of the 98 patients in whom bileduct clearance failed, 6 had early surgical bileduct clearance after nasobiliary catheter drainage, 21 had delayed surgery after drainage by an endoprosthesis,1 required urgent surgery because of failure to establish endoscopic drainage, and 70 were unfit for surgical clearance and were managed by drainage with
double-pigtail endoprostheses (fig 1). Endoscopic sphincterotomy was successful in all 203 patients with a dilated bileduct but without visible stones at endoscopic cholangiography. 69 patients had early complications (within 30 days) after endoscopic sphincterotomy. There was clinical evidence of haemorrhage in 39, 6 of whom required transfusion (mean 2-5 units); 5 patients had symptoms and signs of cholangitis and 11 had a pyrexia above 38OC; 9 patients developed acute pancreatitis and duodenal perforation occurred in 5. Surgery was required for haemorrhage in 1 patient and intraperitoneal perforation in another, and was successful in both; all other complications were treated conservatively. Complications occurred in 32 of 611 patients (5-2%) with visible stones and successful bileduct clearance; in 16 of 161 patients (9-9%) with visible stones and failed bileduct clearance managed by endoscopic drainage; and in 16 of 203 patients (7-9%) with a dilated bileduct but no visible stones. These differences are not statistically significant. Agerelated morbidity was similar for patients aged 36-75 years (7-7-9-7%), but was lower in the oldest age-groups: 6-1% in those aged 76-85 years, and 0-6% in those over 85 years permanent
Fig
1—Outcome of attempted
endoscopic sphincterotomy in
1000
patients.
(table and fig 2). In the 28 patients aged less than 35 years, 3-6% had complications. 4 patients died of causes related to endoscopic retrograde cholecystopancreatography (ERCP): 2 patients aged 81 and 87 years, respectively, died from haemorrhage, 1 at surgery; 1 patient, aged 61 years, died from acute pancreatitis 2 days after the procedure; and an 87-year-old patient had a cardiorespiratory arrest during ERCP. 2 patients in whom endoscopic sphincterotomy was unsuccessful died as a result of surgery. 6 further patients died within 30 days of sphincterotomy-3 from carcinomatosis, 2 from acute vascular events, and 1 from acute renal failure: these 6 deaths were considered to be unrelated to endoscopic sphincterotomy. There was no significant association between mortality after ERCP and age. DISCUSSION
Endoscopic sphincterotomy for common bileduct stones is in widespread use and success rates for sphincterotomy of greater than 95 % have been reported by referral centres,1,3-6,15 although success rates may be considerably lower elsewhere. In this series, the 97-5% overall success rate for sphincterotomy compares well with rates reported from other centres, but up to 3 attempts were required in some patients. However, many of the initial unsuccessful attempts were done by trainees, whereas other studies report the results of experienced endoscopists only. Needle-knife diathermy and a combined percutaneous and endoscopic technique allowed successful sphincterotomy in 25 patients, but this figure is an underestimate of present practice because both techniques were introduced during the study and are now used more frequently. Successful sphincterotomy is not an end in itself, but is necessary for bileduct clearance-for which success rates are of greater importance. Bileduct clearance is achieved less often than sphincterotomy, with reported rates of around 85%,s,is and was achieved in 84-5% of our patients. This difference is important because patients in whom bileduct clearance has failed are at increased risk of complications.2,3,5-7 While bileduct clearance remains the ultimate goal, we also tried to establish endoscopic drainage in all patients with failed bileduct clearance: a pigtail endoprosthesis or a nasobiliary drainage catheter was successfully inserted in 97 of these 98 patients, therefore immediate bileduct clearance or drainage was achieved in 96 7% of all patients. This high rate of early bileduct clearance or drainage probably explains the low frequency of complications-
433 COMPLICATION RATES OF ENDOSCOPIC SPHINCTEROTOMY BY AGE I
I
I
I
*Number of patients who required transfusion in
I
I
parentheses.
cholangitis (2-0%) and acute pancreatitis (11%), which are associated with retained bileduct stones, were rare. Although complications were increased in patients having endoscopic drainage for failed bileduct clearance (9-9%) compared with those with successful bileduct clearance (52%), this difference was not statistically significant. In other studies for patients with retained stones, complication rates of30% and 50%3 are much higher than in our patients managed by endoscopic drainage. Only 1 of our patients required emergency surgery for failure to establish drainage; another 27 patients (3-4%) with failed clearance eventually required surgery, but this could be planned electively since immediate drainage had been established with either nasobiliary catheter drainage or a temporary endoprosthesis. Haemorrhage was the most frequent complication, and occurred in 39 patients. Although not proven, it is likely that the risk of haemorrhage is related to the size of sphincterotomy.12 A reduced frequency of haemorrhage in elderly patients was responsible for the surprising finding that the overall complication rate fell in the oldest age groups (table and fig 2), which may have resulted from more cautious sphincterotomy and attempted bileduct clearance in the elderly, in the knowledge that long-term drainage through an endoprosthesis can be the definitive treatment for these patients,14 who would otherwise be at greatest risk. Endoscopic sphincterotomy was done in 203 patients with a dilated bileduct but no visible stones, but in whom other clinical and radiographic evidence made bileduct
stones the most likely diagnosis. It is likely that most of these patients had bileduct stones which passed before ERCP,
and in others stones may have been present but not visible. Indications for sphincterotomy were similar to those in patients with visible stones, except for a significantly reduced incidence of jaundice-which might be expected if patients without visible stones had indeed passed their stones spontaneously. The 100% success of endoscopic sphincterotomy in the patients without visible stones also lends support to this suggestion, because a patulous papilla, through which a stone had recently passed, would be easier to cannulate. Do these patients with a dilated bileduct and no visible stones on cholangiography require a sphincterotomy if most have already passed their stones? Sphincterotomy may not be necessary in younger patients with this finding at ERCP, because it is possible to provide close surveillance for such patients and missed bileduct stones in this age-group are rarely life-threatening. However, for elderly patients-in whom this finding is most common-follow-up may be difficult and missed bileduct stones are associated with increased morbidity and mortality. In such elderly patients, it is our policy to do a sphincterotomy and trawl the bileduct with a balloon catheter or a basket to ensure that a missed stone has not been left untreated. Our results indicate that this procedure is safe, and we believe it is now standard practice in most centres.
Drainage of bileducts with retained stones was established in 97 of 98 such patients in our study and may explain the low complication rate in these patients. However, bileduct clearance is the goal for most patients, and there are now many techniques by which this may be achieved, including surgery, dissolution, and lithotripsy by crushing baskets,16 extracorporeal shock waves,l’ electrohydraulic probes, and lasers.l8 Some of these novel techniques are still under evaluation, and the best management for bileduct calculi after failed endoscopic removal is not yet clear. However, even with these novel techniques, it is unlikely that immediate bileduct clearance will always be achieved after endoscopic sphincterotomy, and temporary and long-term endoscopic bileduct drainage will remain an important therapy. We thank Sister Pat Mitchell and all her staff for their expertise and assistance, and all consultants and general practitioners who referred patients and provided follow-up information.
Correspondence should be addressed to D. V., Department of Gastroenterology, The Middlesex Hospital, Mortimer Street, London WIN 8AA.
REFERENCES 1. Cotton PB.
Fig
2-Age-related
complication
sphincterotomy in 1000 patients.
rates
of
endoscopic
Endoscopic management of bile
duct
atones—apples
and oranges. Gut
1984; 25: 587-97. 2.Dowsett JF, Vaira D, Polydorou A, Russell RCG, Salmon PR. Interventional endoscopy in the pancreatobiliary tree. Am J Gastroenterol 1988; 83: 1328-36. 3. Davidson BR, Neoptolemos JP, Carr-Locke DL. Endoscopic sphincterotomy for common bile duct calculi in patients with gall bladder in situ considered unfit for surgery. Gut 1988; 29: 114-20. 4. Vaira D, Dowsett JF, Hatfield ARW, et al. Is duodenal diverticulum a risk factor for sphincterotomy? Gut 1989; 30: 912-16. 5. Leese T, Neoptolemos JP, Carr-Locke DL. Successes, failures, early complications and their management following endoscopic sphincterotomy. Results in 394 consecutive patients from a single centre. Br J Surg 1985; 72: 215-19. 6.Mee AS, Vallon AG, Croker JR, Cotton PB. Non-operative removal of bile duct stones by duodenoscopic sphincterotomy in the elderly. Br Med J 1981; 283: 521-23. 7. Escourrou J, Cordova JA, Lazorthes F, et al. Early and late complications after endoscopic sphincterotomy for biliary lithiasis with and without the gallbladder in situ. Gut 1984; 25: 598-602. 8. Glenn F. Trends in surgical treatment of calculous disease of the biliary tract. Surg Gynecol Obstet 1975; 140: 877-84.
434
Adverse Reactions VENOUS THROMBOSIS AND RIFAMPICIN NEIL W. WHITE
Respiratory Clinic, Department of Medicine, Groote Schuur Hospital and University of Cape Town, Cape Town, South Africa 7925 In a retrospective analysis of clinically diagnosed and lower limb deep vein thrombosis (DVT) proven by contrast venography, DVT complicated admissions in 46 (3·4%) of 1366 adult patients treated in a tuberculosis hospital during 1986. Analysis of 7542 admissions during 1978-86 showed a relative risk of 4·74 in patients treated with regimens including rifampicin compared with other regimens. DVT was significantly more common in winter months and usually occurred within 2 weeks of treatment being started. This probable association between rifampicin and DVT does not contraindicate use of this drug, but measures to prevent DVT should be taken in inpatients receiving rifampicin.
Summary
Annual percentage frequency of DVT complicating admissions for treatment of tuberculosis (alternate years).
PATIENTS AND METHODS
in 1978-86. The medical superintendent’s admission records in alternate years for this period were reviewed. July, 1981, to June, 1983, was analysed in more detail when it became apparent that the frequency of DVT had increased during this period. Scrutiny ended in 1986 since more widespread awareness of the putative association in our hospital group might have increased the frequency with which DVT was diagnosed. Record-keeping at the hospital had remained unchanged during the review period. There were numerous changes in medical staff but the medical superintendent, who was responsible for recordkeeping, changed only once. All admissions are regularly reviewed by a panel which includes the medical superintendent. Inpatient records of cases of DVT recorded during 1986 were studied to sample the accuracy of the medical superintendent’s records and for detailed study of the epidemiology of DVT in this situation. The records were analysed for numbers of patients over the age of 13 actually treated for tuberculosis, the regimens used, and the occurrence of DVT. Inpatient records for DVT cases occurring during 1986 were analysed for age and sex, the means by which DVT was confirmed, DVT site, occurrence in relation to start of therapy, length of hospital stay, and outcome. An external control group was assembled from computerised records (1980-87) of admissions to adult medical wards of Groote Schuur Hospital of patients with a diagnosis of non-tuberculous lung abscess or necrotising pneumonia (International Classification of Diseases 513.0). This diagnosis was chosen since neither condition is usually associated with an increased frequency of DVT. Since both groups were admitted for the treatment of pulmonary conditions the degree of immobilisation was presumed to be similar.
Brooklyn Chest Hospital is the only tuberculosis hospital servicing metropolitan Cape Town and surrounding areas. About a
RESULTS
fifth of notified tuberculosis cases are admitted here every year. Patients are generally admitted because they are too sick or live too far away to attend an outpatient clinic. Occasionally patients are admitted because of poor social circumstances or previous noncompliance. 7542 adults were admitted and treated for tuberculosis
In July, 1982, rifampicin (usual daily adult dose 450 or 600 mg) was introduced as standard therapy in combination with pyrazinamide, isoniazid, and streptomycin or ethambutol. Up to that time, because of cost, usual adult
INTRODUCTION
CLINICIANS in our hospital group recognise an association between lower limb deep vein thrombosis (DVT) and admission for the treatment of tuberculosis, a recent observation. However, review of published studies was unrevealing;1.2 most reports deal with unusual and isolated cases of haemorrhagic disorders.3,4 The use of warfarin as an anticoagulant in patients receiving rifampicin is frequently complicated by interaction.5 Large doses of warfarin are often required and prothrombin time must be regularly monitored. Abrupt cessation of rifampicin while continuing warfarin is potentially hazardous. The policy of Brooklyn Chest Hospital in Cape Town is now to prolong the stay of patients with DVT to allow 3 months of oral anticoagulant therapy concomitant with treatment for tuberculosis. To investigate a possible association between therapy of tuberculosis and DVT, I have retrospectively reviewed hospital records.
9. 10.
Doyle PJ, Ward-McQuaid JN, McEwen-Smith A. The value of routine preoperative cholangiography—a report of 4000 cholecystectomies. Br J Surg 1982; 69: 617-19. Spohn K, Fux HD, Mehnert U, Muller-Kluge M, Tewes G. Cholecystektomie und choledochotomie—taktik und techniken. Langenbecks Arch Chir 1973; 334: 249-54.
16.
11. Sullivan DM, Ruffin-Hood T, Griffen WO. 12. 13.
Biliary tract surgery in the elderly. Am J
Surg 1982; 143: 218-20. Dowsett JF, Polydorou AA, Vaira D, et al. Needle knife papillotomy: how safe and how effective? Gut (in press). Dowsett JF, Vaira D, Hatfield ARW, et al. Endoscopic biliary therapy using the combined percutaneous and endoscopic technique. Gastroenterology 1989; 96:
1180-86. 14. Cairns SR, Dias
15.
LM, Cotton PB, Salmon PR, Russell RCG. Additional endoscopic
procedures instead of urgent surgery for retained common bile duct stones. Gut 1989; 30: 535-40. Martin DF, Tweedle DEF. Endoscopic management of common duct stones without cholecystectomy. Br J Surg 1987; 74: 209-11. Demlmg L, Seuberth K, Rieman JF. A mechanical lithotripter. Endoscopy 1982, 14:
100. 17. Sauerbruch
T, Delius M, Paumgartner G, et al. Fragmentation of gallstones by extracorporeal shock waves. N Engl J Med 1986; 314: 818-22. 18. Koch H, Stolte M, Walz V. Endoscopic lithotripsy in the common bile duct Endoscopy 1977; 9: 192-93. 19. Lux G, Hochenberger J, Muller D, et al. The first successful endoscopic retrograde laser lithotripsy of common bile duct stone in man using a pulsed neodynium-YAG laser. Endoscopy 1986; 18: 144-45.