1116
laparoscopic versus mini cholecystectomy
Randomised controlled trial of
Laparoscopic cholecystectomy (LC) has gained wide acceptance for treatment of cholelithiasis in preference to open cholecystectomy, though it has been not formally compared with mini cholecystectomy (MC). We have compared these two techniques in a randomised trial. 70 patients with ultrasound-proven cholelithiasis were randomly allocated LC (38) or MC (32); 37 and 25, respectively, underwent the assigned procedure. The mean hospital stay (including 1 preoperative day) was significantly shorter in the LC than the MC group (median 3 [interquartile range 2-3] vs 4 [3-5], p=0·001) as was duration of convalescence (mean 11·9 [SD 9·1 ] vs 20·2 [16·5] days, p=0·04). The rate of return to normal activities was 1·77 times greater in the LC group than in the MC group (95% confidence interval 1·01-3·11, p=0·03). In regression analysis, the type of cholecystectomy done was the only variable significantly associated with the duration of convalescence. Although there was significant postoperative improvement in all of three quality of life scores in both groups, LC patients improved more quickly than did MC patients. This randomised trial shows the superior effectiveness of LC over MC in treating cholelithiasis. Lancet 1992; 340: 1116-19.
Introduction
Laparoscopic cholecystectomy (LC) has largely replaced surgery in the elective treatment of cholelithiasis. 1-5 The enthusiasm for this symptomatic from stems the results of many personal approach and case series and its appealing modem experiences Mini cholecystectomy (MC) uses a technology.6-9 conventional open approach, but the incision is kept as small as possible. Proponents of this technique have claimed results similar to those achieved with Le. 10-13 It is not yet clear which method is better because the comparison groups in trials published up to now have been self-selected patients and unmatched concurrent or historical controls.4,6,14,15 There have been calls for more formal controlled comparison of LC and MC,1,2,11,16 although some investigators do not believe that it is feasibleP The introduction of alternatives to conventional
conventional
cholecystectomy in our hospitals gave us the chance to carry out a randomised comparison of LC and MC. Patients and methods From September, 1990, to recruited into this study at four
September, 1991, patients were University Hospitals in Montreal
one in Toronto. Eligible patients had experienced at least one episode of right upper-quadrant or epigastric pain (typical biliary colic), had ultrasound-proven cholelithiasis, were 16-85 years old, and presented for elective cholecystectomy. Reasons for exclusion were unfitness for general anaesthesia, previous upper abdominal surgery, pregnancy, patient refusal of randomisation, and suspicion
and
of a common-bileduct stone, based on historical, biochemical, or ultrasonographic abnormalities.18 Patients who were suitable for lithotripsy, based on standard criteria (three or fewer non-calcified stones measuring 3 cm in a functioning gallbladder") were also excluded from the study because a separate randomised trial of lithotripsy compared with LC was taking place at our hospitals .20 Informed consent was given by all patients before randomisation. The study protocol was approved by the institutional review board at each participating centre. After an initial visit to one of the study clinics, eligible and consenting patients were stratified according to age ( < 55 vs ?: 55 years) and were randomly allocated in blocks of 20 to the treatment groups by means of computer-generated random numbers. Surgeon allocation was random within a centre if the patient was referred through the clinic, but if a participating surgeon referred the patient, the patient was returned to him for the assigned type of surgery. Preoperative and postoperative data were collected by study nurses who did not directly care for the patients but were aware of assigned treatment. All patients received single-dose antibiotic prophylaxis at the time of anaesthesia. Each participating surgeon had done at least 30 LCs before he operated on his first study patient; all eight surgeons were able to do both procedures. LC was done under general anaesthesia, with either European or ’North American trocar positions.7,21 Seven surgeons did MC through a small transverse incision in the right upper quadrant,o one used a small midline incision. Patients were randomly followed up as outpatients by the treating surgeon and in special gallstone clinics by the study nurses for at least 3 months after operation. Preoperative data included patient characteristics, history, and findings of physical examination, biochemistry, and ultrasonographic examination. The severity of co-existing disorders was classified on the American Society of Anaesthesiologists (ASA) scoring system.22 Intraoperative and postoperative information was collected prospectively. Primary endpoints were the length of hospital stay and the time to return to full activity (duration of convalescence). The hospital stay was counted from the day of admission (on the day before surgery in all cases) to the day of discharge. Convalescence was defined as complete when a patient could perform all usual home activities, if unemployed, or full usual duties at the workplace if employed. The ADDRESSES. Departments of Surgery, McGill University, Montreal (J S. Barkun, FRCSC, J S. Sampalls, PhD, G Fried, FRCSC, M J Wexler, FRCSC, Prof J L Meakins, FRCSC) and University of Toronto (Prof B Taylor, FRCSC); Division of Gastroenterology, McGill University (A. N Barkun, FRCPC, Prof C. A Goresky, FRCPC); and Divisions of Epidemiology and Biostatistics, Montreal General Hospital (A N. Barkun) and Royal Victoria Hospital (J S Sampalis ), Montreal, Canada. *Additional members’ Dr John Bailar III, Dr Renaldo N. Battista, Dr Roland Brassard, Dr Patrice M Bret, Dr Edward J Hinchey, Dr Lawrence Joseph, Dr Joseph Mamazza, Dr Harvey Sigman, Dr Larry Stein, and Dr Peter L Szego Correspondence to Prof Jonathan L. Meakins, Department of Surgery, Room S10.34, Royal Victoria Hospital, 687 Pine Avenue West, Montreal, Quebec, Canada H3A 1A1
1117
convalescence was measured from the first to the day before return to full activities. This duration was ascertained by the study nurses and reviewed with the study coordinator. Secondary endpoints included postoperative pain and quality of life. Postoperative pain was assessed by administering the McGill pain questionnaire23 within 24 h of surgery, and by measuring the amount of narcotics (as morphine dose equivalents [mg]) used during the hospital stay and the first postoperative week. Quality of life measurements were taken preoperatively and 1 month and 3 months after cholecystectomy. Several weeks into the trial, so as to assess the progress of convalescence, 10-day convalescence data were also collected. The questionnaires were filled out by the patients after they had received standard instructions from the study nurses. We used three different instruments to measure quality of life. The Nottingham Health Profile Questionnaire (NHPQ) is a general quality of life index;’ it includes 38 items covering six domains and a point is given for each positive response and none for a negative response. We also used a German gastrointestinal surgery quality of life index (GGSQLI)25 and a visual analogue scale (VAS) on which patients were asked to rate their quality of life on a linear scale labelled from 0 (poor) to 11 (excellent)?6 Other secondary endpoints included duration of operation, the rate of conversion to open cholecystectomy, and postoperative days to full diet. Morbidity, mortality, and retained stone rates are important variables but, because these rates are low, they were of little use in this trial given the small sample size and the limited time of follow-up. All results were analysed according to the intention-to-treat principle. Results were also analysed according to treatment received. Data are given as mean and standard deviation, except for duration of hospital stay, which showed a skewed distribution. We used Student’s t test and Wilcoxon’s rank sum test for continuous variables and the chi-square test or Fisher’s exact test for categorical variables. We assessed changes from baseline values for quality of life measured with paired tests and between-group differences with non-paired procedures. For the primary study endpoint (duration of convalescence) the Kaplan-Meier method was used to construct life-tables and the non-parametric log-rank test was used to assess between-group differences. Cox’s proportional hazards model was used to evaluate between-group differences for this endpoint. Based on preliminary studies from our own hospital and reasonable estimates of convalescence outcomes, we calculated (before the study) that 50 patients in each group should enable the detection of a difference in duration of convalescence of 18 days, with &agr; = 0.05 and statistical power of 0-8. duration
of
TABLE I-BASELINE PATIENT CHARACTERISTICS
postoperative day
Results
During the 12-month study, consecutive eligible patients take part; 70 consented to do so. 38 were 32 MC. The patient groups were similar in terms of baseline characteristics (table I). The distribution of patients by occupation was also similar; 40% performed most of their activities at home, 15% were involved in manual labour, and less than 10% were professionals. The two groups were similar as regards duration of symptoms, baseline measurements of quality of life (table I), and ASA score (% class 1=84% LC, 91 % were
invited
to
randomly allocated LC and
MC). 37
(97%) patients in the LC group and 25 (78%)
in the
MC group actually underwent the allocated procedure and contributed data to the analysis. 4 patients (1LC, 3 MC) declined to take part after randomisation and refused any follow-up. In the MC group 3 other patients refused to take part after randomisation and opted to have open
cholecystectomy by non-participating surgeons who did not "mini" technique, and 1 other underwent LC at a non-participating hospital. There was no difference in baseline characteristics between patients who withdrew use a
*For
quality of life measures LC n = 35, MC n = 23.
from each group, and they did not obviously differ from the rest of the study patients. We could not obtain follow-up information on the withdrawn patients so they are not included in the analysis of results. Results were the same whether we analysed by intention to treat or by treatment received. The mean duration of surgery was more than 12 min longer for LC than for MC (table II). 1 patient in the LC group had to undergo open cholecystectomy, and a patient in the MC group elected to have LC just before being taken to surgery. Differential analysis of these patients did not affect the results. Intraoperative cholangiography was used in only 1 patient, in the MC group. LC patients had a significantly shorter hospital stay than MC patients (table II). 16 (43%) LC patients were discharged on the first postoperative day compared with 3 (12%) MC patients
(p = 0-02). The
mean
duration
of
convalescence
was
also
significantly longer in the MC than the LC group (p < 0-04, table II). The time at which 50% of patients had returned to full normal activities was 9 days for LC patients and 14 days for MC patients (see figure). Cox’s proportional hazards analysis showed that the rate ratio of return to normal activities for the LC group compared with the MC group was 1-77 (95% confidence interval 1,01-3’11, p=003). In the proportional hazards model, the only variable significantly associated with the speed of convalescence was the type of cholecystectomy used. The other variables in the model (age, sex, Quetelet index, occupation, or interactions of any of the above) were not significant predictors of the duration of convalescence. Nor did inclusion of a variable to represent the surgeon or the centre in the model change the results. The LC patients were able to return to a full diet earlier than patients who had undergone MC (table II). Right shoulder pain was more common in the LC than the MC group, but postoperative use of narcotics during hospital stay and the first postoperative week was significantly greater in the MC group (LC 174 [12’5] vs MC 79-2 [838] Pain on the day after mg morphine equivalents, p < 0 001). TABLE II-POSTOPERATIVE ASSESSMENT
1118
shorter than those of the nurses (p < 0-01). Furthermore, the intraclass correlation coefficient measuring the nurses’ and surgeons’ estimates of convalescence was 0-78 (lower 95% CI 064), which indicates moderate to low agreement. =
Discussion The
development and spread of LC in North America spearheaded by consumer demand amid calls for initial, caution and careful comparative assessment.1,2,27 Pooled results of more than 5000 patients show that the laparoscopic approach is safe and effective.5,9 Although previous studies have suggested that LC is better than conventional open cholecystectomy,4,6,14,15 the patient benefits of LC over MC have never been clearly proven. MC is feasible in most patients and is associated with shorter hospital stays and faster recovery than traditional cholecystectomy; it can be done through either a midline or a were
Kaplan-Meier survival analysis of duration of convalescence. LC
patients required significantly shorter patients (p=0.036, Mantel-Cox). surgery
McGill
was
convalescence than MC
greater in the MC patients, as measured by scores, but this difference was not significant
pain (table II). Analysis of postoperative and baseline quality of life scores showed that patients in both groups improved after surgery (table III). With paired analysis, significant improvement in the LC group was detected as early as 10 days after surgery with the VAS (p 0047) and at 1 month with the NHPQ and the GGSQLI (p = 0 0001). The MC group did not show significant improvement until 1 month after surgery with the GGSQLI (p=0004) and the VAS (p 0-002), and until 3 months with the NHPQ (p 0 03). The extent of improvement in all quality of life scores after =
=
=
was similar in both groups. There were 2 patients with complications in the MC group. 1 patient developed a wound infection that was slow to heal and 1 presented 10 days postoperatively with acute abdominal pain. Free intraperitoneal bile was found and drained at laparotomy, but no leak could be seen. 1 patient had a complication in the LC group; she developed a persistent ileus and pain 4 days postoperatively. No abnormality was found at laparotomy and she subsequently improved. She stayed in hospital for 13 days. No deaths occurred and no retained stone has yet been detected in either patient group. We compared duration of convalescence of the LC patients as measured by the study nurses and by the treating surgeons. The surgeon’s estimates were 4 days (15%)
surgery
TABLE III-QUALITY OF LIFE
incision.10-13 Before LC became available, MC had become the procedure of choice for the elective treatment of gallstones for most surgeons in our group. MC requires only slight modifications in technique from conventional cholecystectomy and it does not necessitate sophisticated technology. The introduction of LC gave us the chance to compare MC and LC in a randomised trial. The design of our study was chosen to correct what we thought were important limitations of previous comparative assessments, especially to balance out potential effects of patient motivation, occupation, or personal disposition and expectations, which may strongly affect outcome.28 We used endpoints clearly defined before the start of the trial. These were recorded by surgeons and by independent assessors. To avoid any effect of increasing familarity with LC during the study, participating surgeons had to have done at least 30 LCs previously. Hospital stay, time to full diet, and duration of convalescence in the MC group were similar to previous results with this technique.1O-13 Duration of convalescence in the MC group was shorter than that described in at least one previous comparative trial of LC and MC.14 This finding emphasises the importance of using a suitable control group for comparative analysis. Despite the very good results with MC, patients in the LC group still fared better for all transverse
predefmed outcomes. A statistically and clinically significant difference
in duration of convalescence between the groups was shown by survival analysis. In a comparison of patients who had returned to full activity by 30 days, the slope of each curve up to this time is steeper in the LC than the MC group, the LC patients recovered from surgery 1 -77 times more rapidly than MC patients. All patients who took more than 1 month to return to full activity had suffered complications or exacerbation of co-existing disease. The heterogeneity of the complications and the small number of patients makes comparison of this subset difficult. We used multivariate regression analysis to fmd out which variables could best explain the observed variation in duration of convalescence. The type of operation done was the only variable significantly associated with duration of convalescence even with control for possible confounding variables. Surgeon and centre effects were not found to affect outcome, but such effects may be obscured by the small size of our sample. Mortality and morbidity findings are difficult to interpret, again because of the small sample size, but they were similar to previous rates. 5,9 It is interesting that an MC patient developed a postoperative bile leak; this
1119
complication is more common with LC than with an open technique.9 The LC patient who underwent laparotomy for a persistent ileus might now be managed differently, given our greater experience with LC. Postoperative pain as measured objectively (narcotic doses needed) was significantly greater in the MC group; although subjectively measured pain scores did not differ significantly, the trend confirmed the objective measurements. The quality of life indices were chosen to reflect both general and specific aspects of health status. All were sensitive enough to detect an improvement after operation. Although the change in quality of life scores did not differ significantly between the groups, the LC patients showed faster improvement in all three indices.
PA, Quereshi A, Coleman J, et al. Reduced postoperative hospitalization after laparoscopic cholecystectomy. Br J Surg 1991; 78:
6. Grace
160-62. 7. Dubois F,
Icard P, Berthelot G, Levard H. Coelioscopic cholecystectomy. Ann Surg 1990; 211: 60-62. 8. Peters JH, Ellison EC, Innes JT, et al. Safety and efficacy of laparoscopic cholecystectomy. Ann Surg 1991; 213: 3-12. 9. The Southern Surgeons Club. A prospective analysis of 1518 laparoscopic cholecystectomies. N Engl J Med 1991; 324: 1073-78. 10. Ledet WP Jr. Ambulatory cholecystectomy without disability. Arch Surg 1990; 125: 1434-35. 11. Saltzstein EC, Mercer LC, Peacock JB, et al. Outpatient open cholecystectomy. Surg Gynecol Obstet 1992; 174: 173-75. 12. Armstrong PJ, Burgess RW. Choice of incision and pain following gallbladder surgery. Br J Surg 1990; 77: 746-48. 13. Cheslyn-Curtis S, Russell RCG. New trends in gallstone management. Br J Surg 1991; 78: 143-49. 14. Reddick EJ, Olsen DO. Laparoscopic laser cholecystectomy. Surg Endosc 1989; 3: 131-33. 15. Frazee RC, Roberts JW, Okeson GC, et al. Open versus laparoscopic cholecystectomy. Ann Surg 1991; 213: 651-54. 16. Walsh TN, Russell RCG. Cholecystectomy and gallbladder conservation. Br J Surg 1992; 79: 4-5. 17. Neugebauer E, Troidl H, Spangenberger W, Dietrich A, Lefering R, and The Cholecystectomy Study Group. Conventional versus laparoscopic cholecystectomy and the randomized controlled trial. Br J Surg 1991;
Several features of our trial may limit the scope of its conclusions. The study population is not fully representative of all patients with symptomatic gallstones because lithotripsy candidates and patients with suspected common-bileduct stones were excluded. Since only eight of forty surgeons performing cholecystectomy at McGill University took part in the study, and two-thirds of patients were contributed by three surgeons in different hospitals, we do not know exactly the size of the population from which the study sample is taken. Nevertheless, there is no strong biological reason to assume that the main conclusions cannot be generalised. More patients withdrew from the MC than the LC group; however, all were accounted for by factors that seemed to be similar in both groups. The high rate of withdrawal after randomisation may well reflect the poor acceptance of surgical randomisation by patients. The trial was stopped because patient recruitment had become difficult, and because significant differences in primary endpoints had been reached. We had almost achieved the targeted recruitment based on our pretrial sample size calculations. The surgeons, as in every other reported assessment, were involved in both administering the intervention and measuring the outcome. To investigate the effect of this factor on the assessment of convalescence duration, research nurses also assessed that endpoint; there was a 15 % discrepancy. The surgeons’ estimates are similar to previously published results .14,211 Such observer bias will need to be considered in future unblinded surgical trials where outcomes are measured by the treating surgeon. This trial has shown that LC is superior to MC in patients with symptomatic gallbladder stones not suspected of having common-bileduct stones. Issues requiring further study include the cost-effectiveness of the laparoscopic approach and its place in relation to other treatments for gallstone disease, such as lithotripsy.
78: 150-54. 18. Steinmetz OK, Barkun AN, Barkun JS, and the McGill Gallstone Treatment Group. The role of ERCP in laparoscopic cholecystectomy. Clin Invest Med 1991; 14: A51. 19. Sackmann M, Delius M, Sauerbruch T, et al. Shock wave lithotripsy of gallbladder stones, the first 175 patients. N Engl J Med 1988; 318: 393-97. 20. Barkun AN, Barkun JS, Sampalis JS, and the McGill Gallstone Treatment Group. A randomized clinical trial comparing gallbladder stone shock wave lithotripsy to laparoscopic cholecystectomy. Clin Invest Med 1992; 15: A44. 21. Olsen DO. Laparoscopic cholecystectomy. Am J Surg 1991; 161: 339-44. 22. Owen WD, Felts JA, Spitznagel EL Jr, et al. ASA physical status classification: a study of consistency ratings. Anaesthesiology 1978; 49: 239-43. 23. Melzack R. The McGill pain questionnaire: major properties and scoring methods. Pain 1975; 1: 277-99. 24. Hunt SM, McEwen J, McKenna SP. Measuring health status: a new tool for clinicians and epidemiologists. J R Coll Gen Pract 1985; 35: 185-88. 25. Eypasch E, Troidl H, Wood-Dauphinee S, et al. Quality of life in gastrointestinal surgery: a clinimetric approach to development of a measure. Theoret Surg 1990; 5: 3-10. 26. Streiner DL, Norman GR, eds. Scaling responses. In: Health measurement scales. A practical guide to their development and use. Oxford: Oxford Medical publications, 1989: 23-25. 27. Cameron JL, Gadacz TR. Laparoscopic cholecystectomy. Ann Surg
We thank Ms Diane Thibeault, Ms Christine Wickham, and Ms Becky Sandler for invaluable contributions to the study.
From The Lancet
This study was supported by a Medical Research Council of Canada University-Industry Clinical trial grant in partnership with Ethicon, Limited, and by a grant from the Fonds de la Recherche en Sante du Quebec.
Happy birthday BMA
REFERENCES 1. Paterson-Brown
S,
Garden
OJ,
Carter
DC.
cholecystectomy. Br J Surg 1991; 78: 131-32. 2. Editorial. Cholecystectomy practice transformed.
Laparoscopic
Lancet 1991; 338:
789-90. 3. Schirmer
BD, Edge SB,
Dix
J, Hyser MS, Hanks JB, Jones RS.
Laparoscopic cholecystectomy. Ann Surg 1991; 213: 665-77. 4. Gadacz TR, Talamini MA. Traditional vs laparoscopic cholecystectomy. Am J Surg 1991; 161: 336-38. 5. Larson GM, Vitale GC, Casey J, et al. Multipractice analysis of laparoscopic cholecystectomy in 1983 patients. Am J Surg 1992; 163: 221-26.
1991; 213: 1-2.
GC, Collet D, Larson GM, Cheadle WG, Miller FB, Perinot J. Interruption of professional and home activity after laparoscopic cholecystectomy among French and American patients. Am J Surg
28. Vitale
1991; 161: 396-98.
The anniversary meeting of the British Medical Association is to be held at the British coffee-house, Cockspur-street, on Tuesday evening next.... This Association, from its birth, has kept steadily in the right path. The proceedings have not been characterised by trickery or by shuffling. The Council, and the members generally, have boldly and courageously encountered the abuses of the Colleges, the imperfect state of medical law, and the absence ofa just legal protection for medical practitioners in the exercise of their professional rights, and it has resolutely and perseveringly pointed to the only substantial and effective remedies. Such an association is entitled to the respectful confidence of the profession ....
(Nov 5, 1842)
laparoscopic versus mini cholecystectomy
Randomised controlled trial of
Laparoscopic cholecystectomy (LC) has gained wide acceptance for treatment of cholelithiasis in preference to open cholecystectomy, though it has been not formally compared with mini cholecystectomy (MC). We have compared these two techniques in a randomised trial. 70 patients with ultrasound-proven cholelithiasis were randomly allocated LC (38) or MC (32); 37 and 25, respectively, underwent the assigned procedure. The mean hospital stay (including 1 preoperative day) was significantly shorter in the LC than the MC group (median 3 [interquartile range 2-3] vs 4 [3-5], p=0·001) as was duration of convalescence (mean 11·9 [SD 9·1 ] vs 20·2 [16·5] days, p=0·04). The rate of return to normal activities was 1·77 times greater in the LC group than in the MC group (95% confidence interval 1·01-3·11, p=0·03). In regression analysis, the type of cholecystectomy done was the only variable significantly associated with the duration of convalescence. Although there was significant postoperative improvement in all of three quality of life scores in both groups, LC patients improved more quickly than did MC patients. This randomised trial shows the superior effectiveness of LC over MC in treating cholelithiasis. Lancet 1992; 340: 1116-19.
Introduction
Laparoscopic cholecystectomy (LC) has largely replaced surgery in the elective treatment of cholelithiasis. 1-5 The enthusiasm for this symptomatic from stems the results of many personal approach and case series and its appealing modem experiences Mini cholecystectomy (MC) uses a technology.6-9 conventional open approach, but the incision is kept as small as possible. Proponents of this technique have claimed results similar to those achieved with Le. 10-13 It is not yet clear which method is better because the comparison groups in trials published up to now have been self-selected patients and unmatched concurrent or historical controls.4,6,14,15 There have been calls for more formal controlled comparison of LC and MC,1,2,11,16 although some investigators do not believe that it is feasibleP The introduction of alternatives to conventional
conventional
cholecystectomy in our hospitals gave us the chance to carry out a randomised comparison of LC and MC. Patients and methods From September, 1990, to recruited into this study at four
September, 1991, patients were University Hospitals in Montreal
one in Toronto. Eligible patients had experienced at least one episode of right upper-quadrant or epigastric pain (typical biliary colic), had ultrasound-proven cholelithiasis, were 16-85 years old, and presented for elective cholecystectomy. Reasons for exclusion were unfitness for general anaesthesia, previous upper abdominal surgery, pregnancy, patient refusal of randomisation, and suspicion
and
of a common-bileduct stone, based on historical, biochemical, or ultrasonographic abnormalities.18 Patients who were suitable for lithotripsy, based on standard criteria (three or fewer non-calcified stones measuring 3 cm in a functioning gallbladder") were also excluded from the study because a separate randomised trial of lithotripsy compared with LC was taking place at our hospitals .20 Informed consent was given by all patients before randomisation. The study protocol was approved by the institutional review board at each participating centre. After an initial visit to one of the study clinics, eligible and consenting patients were stratified according to age ( < 55 vs ?: 55 years) and were randomly allocated in blocks of 20 to the treatment groups by means of computer-generated random numbers. Surgeon allocation was random within a centre if the patient was referred through the clinic, but if a participating surgeon referred the patient, the patient was returned to him for the assigned type of surgery. Preoperative and postoperative data were collected by study nurses who did not directly care for the patients but were aware of assigned treatment. All patients received single-dose antibiotic prophylaxis at the time of anaesthesia. Each participating surgeon had done at least 30 LCs before he operated on his first study patient; all eight surgeons were able to do both procedures. LC was done under general anaesthesia, with either European or ’North American trocar positions.7,21 Seven surgeons did MC through a small transverse incision in the right upper quadrant,o one used a small midline incision. Patients were randomly followed up as outpatients by the treating surgeon and in special gallstone clinics by the study nurses for at least 3 months after operation. Preoperative data included patient characteristics, history, and findings of physical examination, biochemistry, and ultrasonographic examination. The severity of co-existing disorders was classified on the American Society of Anaesthesiologists (ASA) scoring system.22 Intraoperative and postoperative information was collected prospectively. Primary endpoints were the length of hospital stay and the time to return to full activity (duration of convalescence). The hospital stay was counted from the day of admission (on the day before surgery in all cases) to the day of discharge. Convalescence was defined as complete when a patient could perform all usual home activities, if unemployed, or full usual duties at the workplace if employed. The ADDRESSES. Departments of Surgery, McGill University, Montreal (J S. Barkun, FRCSC, J S. Sampalls, PhD, G Fried, FRCSC, M J Wexler, FRCSC, Prof J L Meakins, FRCSC) and University of Toronto (Prof B Taylor, FRCSC); Division of Gastroenterology, McGill University (A. N Barkun, FRCPC, Prof C. A Goresky, FRCPC); and Divisions of Epidemiology and Biostatistics, Montreal General Hospital (A N. Barkun) and Royal Victoria Hospital (J S Sampalis ), Montreal, Canada. *Additional members’ Dr John Bailar III, Dr Renaldo N. Battista, Dr Roland Brassard, Dr Patrice M Bret, Dr Edward J Hinchey, Dr Lawrence Joseph, Dr Joseph Mamazza, Dr Harvey Sigman, Dr Larry Stein, and Dr Peter L Szego Correspondence to Prof Jonathan L. Meakins, Department of Surgery, Room S10.34, Royal Victoria Hospital, 687 Pine Avenue West, Montreal, Quebec, Canada H3A 1A1
1117
convalescence was measured from the first to the day before return to full activities. This duration was ascertained by the study nurses and reviewed with the study coordinator. Secondary endpoints included postoperative pain and quality of life. Postoperative pain was assessed by administering the McGill pain questionnaire23 within 24 h of surgery, and by measuring the amount of narcotics (as morphine dose equivalents [mg]) used during the hospital stay and the first postoperative week. Quality of life measurements were taken preoperatively and 1 month and 3 months after cholecystectomy. Several weeks into the trial, so as to assess the progress of convalescence, 10-day convalescence data were also collected. The questionnaires were filled out by the patients after they had received standard instructions from the study nurses. We used three different instruments to measure quality of life. The Nottingham Health Profile Questionnaire (NHPQ) is a general quality of life index;’ it includes 38 items covering six domains and a point is given for each positive response and none for a negative response. We also used a German gastrointestinal surgery quality of life index (GGSQLI)25 and a visual analogue scale (VAS) on which patients were asked to rate their quality of life on a linear scale labelled from 0 (poor) to 11 (excellent)?6 Other secondary endpoints included duration of operation, the rate of conversion to open cholecystectomy, and postoperative days to full diet. Morbidity, mortality, and retained stone rates are important variables but, because these rates are low, they were of little use in this trial given the small sample size and the limited time of follow-up. All results were analysed according to the intention-to-treat principle. Results were also analysed according to treatment received. Data are given as mean and standard deviation, except for duration of hospital stay, which showed a skewed distribution. We used Student’s t test and Wilcoxon’s rank sum test for continuous variables and the chi-square test or Fisher’s exact test for categorical variables. We assessed changes from baseline values for quality of life measured with paired tests and between-group differences with non-paired procedures. For the primary study endpoint (duration of convalescence) the Kaplan-Meier method was used to construct life-tables and the non-parametric log-rank test was used to assess between-group differences. Cox’s proportional hazards model was used to evaluate between-group differences for this endpoint. Based on preliminary studies from our own hospital and reasonable estimates of convalescence outcomes, we calculated (before the study) that 50 patients in each group should enable the detection of a difference in duration of convalescence of 18 days, with &agr; = 0.05 and statistical power of 0-8. duration
of
TABLE I-BASELINE PATIENT CHARACTERISTICS
postoperative day
Results
During the 12-month study, consecutive eligible patients take part; 70 consented to do so. 38 were 32 MC. The patient groups were similar in terms of baseline characteristics (table I). The distribution of patients by occupation was also similar; 40% performed most of their activities at home, 15% were involved in manual labour, and less than 10% were professionals. The two groups were similar as regards duration of symptoms, baseline measurements of quality of life (table I), and ASA score (% class 1=84% LC, 91 % were
invited
to
randomly allocated LC and
MC). 37
(97%) patients in the LC group and 25 (78%)
in the
MC group actually underwent the allocated procedure and contributed data to the analysis. 4 patients (1LC, 3 MC) declined to take part after randomisation and refused any follow-up. In the MC group 3 other patients refused to take part after randomisation and opted to have open
cholecystectomy by non-participating surgeons who did not "mini" technique, and 1 other underwent LC at a non-participating hospital. There was no difference in baseline characteristics between patients who withdrew use a
*For
quality of life measures LC n = 35, MC n = 23.
from each group, and they did not obviously differ from the rest of the study patients. We could not obtain follow-up information on the withdrawn patients so they are not included in the analysis of results. Results were the same whether we analysed by intention to treat or by treatment received. The mean duration of surgery was more than 12 min longer for LC than for MC (table II). 1 patient in the LC group had to undergo open cholecystectomy, and a patient in the MC group elected to have LC just before being taken to surgery. Differential analysis of these patients did not affect the results. Intraoperative cholangiography was used in only 1 patient, in the MC group. LC patients had a significantly shorter hospital stay than MC patients (table II). 16 (43%) LC patients were discharged on the first postoperative day compared with 3 (12%) MC patients
(p = 0-02). The
mean
duration
of
convalescence
was
also
significantly longer in the MC than the LC group (p < 0-04, table II). The time at which 50% of patients had returned to full normal activities was 9 days for LC patients and 14 days for MC patients (see figure). Cox’s proportional hazards analysis showed that the rate ratio of return to normal activities for the LC group compared with the MC group was 1-77 (95% confidence interval 1,01-3’11, p=003). In the proportional hazards model, the only variable significantly associated with the speed of convalescence was the type of cholecystectomy used. The other variables in the model (age, sex, Quetelet index, occupation, or interactions of any of the above) were not significant predictors of the duration of convalescence. Nor did inclusion of a variable to represent the surgeon or the centre in the model change the results. The LC patients were able to return to a full diet earlier than patients who had undergone MC (table II). Right shoulder pain was more common in the LC than the MC group, but postoperative use of narcotics during hospital stay and the first postoperative week was significantly greater in the MC group (LC 174 [12’5] vs MC 79-2 [838] Pain on the day after mg morphine equivalents, p < 0 001). TABLE II-POSTOPERATIVE ASSESSMENT
1118
shorter than those of the nurses (p < 0-01). Furthermore, the intraclass correlation coefficient measuring the nurses’ and surgeons’ estimates of convalescence was 0-78 (lower 95% CI 064), which indicates moderate to low agreement. =
Discussion The
development and spread of LC in North America spearheaded by consumer demand amid calls for initial, caution and careful comparative assessment.1,2,27 Pooled results of more than 5000 patients show that the laparoscopic approach is safe and effective.5,9 Although previous studies have suggested that LC is better than conventional open cholecystectomy,4,6,14,15 the patient benefits of LC over MC have never been clearly proven. MC is feasible in most patients and is associated with shorter hospital stays and faster recovery than traditional cholecystectomy; it can be done through either a midline or a were
Kaplan-Meier survival analysis of duration of convalescence. LC
patients required significantly shorter patients (p=0.036, Mantel-Cox). surgery
McGill
was
convalescence than MC
greater in the MC patients, as measured by scores, but this difference was not significant
pain (table II). Analysis of postoperative and baseline quality of life scores showed that patients in both groups improved after surgery (table III). With paired analysis, significant improvement in the LC group was detected as early as 10 days after surgery with the VAS (p 0047) and at 1 month with the NHPQ and the GGSQLI (p = 0 0001). The MC group did not show significant improvement until 1 month after surgery with the GGSQLI (p=0004) and the VAS (p 0-002), and until 3 months with the NHPQ (p 0 03). The extent of improvement in all quality of life scores after =
=
=
was similar in both groups. There were 2 patients with complications in the MC group. 1 patient developed a wound infection that was slow to heal and 1 presented 10 days postoperatively with acute abdominal pain. Free intraperitoneal bile was found and drained at laparotomy, but no leak could be seen. 1 patient had a complication in the LC group; she developed a persistent ileus and pain 4 days postoperatively. No abnormality was found at laparotomy and she subsequently improved. She stayed in hospital for 13 days. No deaths occurred and no retained stone has yet been detected in either patient group. We compared duration of convalescence of the LC patients as measured by the study nurses and by the treating surgeons. The surgeon’s estimates were 4 days (15%)
surgery
TABLE III-QUALITY OF LIFE
incision.10-13 Before LC became available, MC had become the procedure of choice for the elective treatment of gallstones for most surgeons in our group. MC requires only slight modifications in technique from conventional cholecystectomy and it does not necessitate sophisticated technology. The introduction of LC gave us the chance to compare MC and LC in a randomised trial. The design of our study was chosen to correct what we thought were important limitations of previous comparative assessments, especially to balance out potential effects of patient motivation, occupation, or personal disposition and expectations, which may strongly affect outcome.28 We used endpoints clearly defined before the start of the trial. These were recorded by surgeons and by independent assessors. To avoid any effect of increasing familarity with LC during the study, participating surgeons had to have done at least 30 LCs previously. Hospital stay, time to full diet, and duration of convalescence in the MC group were similar to previous results with this technique.1O-13 Duration of convalescence in the MC group was shorter than that described in at least one previous comparative trial of LC and MC.14 This finding emphasises the importance of using a suitable control group for comparative analysis. Despite the very good results with MC, patients in the LC group still fared better for all transverse
predefmed outcomes. A statistically and clinically significant difference
in duration of convalescence between the groups was shown by survival analysis. In a comparison of patients who had returned to full activity by 30 days, the slope of each curve up to this time is steeper in the LC than the MC group, the LC patients recovered from surgery 1 -77 times more rapidly than MC patients. All patients who took more than 1 month to return to full activity had suffered complications or exacerbation of co-existing disease. The heterogeneity of the complications and the small number of patients makes comparison of this subset difficult. We used multivariate regression analysis to fmd out which variables could best explain the observed variation in duration of convalescence. The type of operation done was the only variable significantly associated with duration of convalescence even with control for possible confounding variables. Surgeon and centre effects were not found to affect outcome, but such effects may be obscured by the small size of our sample. Mortality and morbidity findings are difficult to interpret, again because of the small sample size, but they were similar to previous rates. 5,9 It is interesting that an MC patient developed a postoperative bile leak; this
1119
complication is more common with LC than with an open technique.9 The LC patient who underwent laparotomy for a persistent ileus might now be managed differently, given our greater experience with LC. Postoperative pain as measured objectively (narcotic doses needed) was significantly greater in the MC group; although subjectively measured pain scores did not differ significantly, the trend confirmed the objective measurements. The quality of life indices were chosen to reflect both general and specific aspects of health status. All were sensitive enough to detect an improvement after operation. Although the change in quality of life scores did not differ significantly between the groups, the LC patients showed faster improvement in all three indices.
PA, Quereshi A, Coleman J, et al. Reduced postoperative hospitalization after laparoscopic cholecystectomy. Br J Surg 1991; 78:
6. Grace
160-62. 7. Dubois F,
Icard P, Berthelot G, Levard H. Coelioscopic cholecystectomy. Ann Surg 1990; 211: 60-62. 8. Peters JH, Ellison EC, Innes JT, et al. Safety and efficacy of laparoscopic cholecystectomy. Ann Surg 1991; 213: 3-12. 9. The Southern Surgeons Club. A prospective analysis of 1518 laparoscopic cholecystectomies. N Engl J Med 1991; 324: 1073-78. 10. Ledet WP Jr. Ambulatory cholecystectomy without disability. Arch Surg 1990; 125: 1434-35. 11. Saltzstein EC, Mercer LC, Peacock JB, et al. Outpatient open cholecystectomy. Surg Gynecol Obstet 1992; 174: 173-75. 12. Armstrong PJ, Burgess RW. Choice of incision and pain following gallbladder surgery. Br J Surg 1990; 77: 746-48. 13. Cheslyn-Curtis S, Russell RCG. New trends in gallstone management. Br J Surg 1991; 78: 143-49. 14. Reddick EJ, Olsen DO. Laparoscopic laser cholecystectomy. Surg Endosc 1989; 3: 131-33. 15. Frazee RC, Roberts JW, Okeson GC, et al. Open versus laparoscopic cholecystectomy. Ann Surg 1991; 213: 651-54. 16. Walsh TN, Russell RCG. Cholecystectomy and gallbladder conservation. Br J Surg 1992; 79: 4-5. 17. Neugebauer E, Troidl H, Spangenberger W, Dietrich A, Lefering R, and The Cholecystectomy Study Group. Conventional versus laparoscopic cholecystectomy and the randomized controlled trial. Br J Surg 1991;
Several features of our trial may limit the scope of its conclusions. The study population is not fully representative of all patients with symptomatic gallstones because lithotripsy candidates and patients with suspected common-bileduct stones were excluded. Since only eight of forty surgeons performing cholecystectomy at McGill University took part in the study, and two-thirds of patients were contributed by three surgeons in different hospitals, we do not know exactly the size of the population from which the study sample is taken. Nevertheless, there is no strong biological reason to assume that the main conclusions cannot be generalised. More patients withdrew from the MC than the LC group; however, all were accounted for by factors that seemed to be similar in both groups. The high rate of withdrawal after randomisation may well reflect the poor acceptance of surgical randomisation by patients. The trial was stopped because patient recruitment had become difficult, and because significant differences in primary endpoints had been reached. We had almost achieved the targeted recruitment based on our pretrial sample size calculations. The surgeons, as in every other reported assessment, were involved in both administering the intervention and measuring the outcome. To investigate the effect of this factor on the assessment of convalescence duration, research nurses also assessed that endpoint; there was a 15 % discrepancy. The surgeons’ estimates are similar to previously published results .14,211 Such observer bias will need to be considered in future unblinded surgical trials where outcomes are measured by the treating surgeon. This trial has shown that LC is superior to MC in patients with symptomatic gallbladder stones not suspected of having common-bileduct stones. Issues requiring further study include the cost-effectiveness of the laparoscopic approach and its place in relation to other treatments for gallstone disease, such as lithotripsy.
78: 150-54. 18. Steinmetz OK, Barkun AN, Barkun JS, and the McGill Gallstone Treatment Group. The role of ERCP in laparoscopic cholecystectomy. Clin Invest Med 1991; 14: A51. 19. Sackmann M, Delius M, Sauerbruch T, et al. Shock wave lithotripsy of gallbladder stones, the first 175 patients. N Engl J Med 1988; 318: 393-97. 20. Barkun AN, Barkun JS, Sampalis JS, and the McGill Gallstone Treatment Group. A randomized clinical trial comparing gallbladder stone shock wave lithotripsy to laparoscopic cholecystectomy. Clin Invest Med 1992; 15: A44. 21. Olsen DO. Laparoscopic cholecystectomy. Am J Surg 1991; 161: 339-44. 22. Owen WD, Felts JA, Spitznagel EL Jr, et al. ASA physical status classification: a study of consistency ratings. Anaesthesiology 1978; 49: 239-43. 23. Melzack R. The McGill pain questionnaire: major properties and scoring methods. Pain 1975; 1: 277-99. 24. Hunt SM, McEwen J, McKenna SP. Measuring health status: a new tool for clinicians and epidemiologists. J R Coll Gen Pract 1985; 35: 185-88. 25. Eypasch E, Troidl H, Wood-Dauphinee S, et al. Quality of life in gastrointestinal surgery: a clinimetric approach to development of a measure. Theoret Surg 1990; 5: 3-10. 26. Streiner DL, Norman GR, eds. Scaling responses. In: Health measurement scales. A practical guide to their development and use. Oxford: Oxford Medical publications, 1989: 23-25. 27. Cameron JL, Gadacz TR. Laparoscopic cholecystectomy. Ann Surg
We thank Ms Diane Thibeault, Ms Christine Wickham, and Ms Becky Sandler for invaluable contributions to the study.
From The Lancet
This study was supported by a Medical Research Council of Canada University-Industry Clinical trial grant in partnership with Ethicon, Limited, and by a grant from the Fonds de la Recherche en Sante du Quebec.
Happy birthday BMA
REFERENCES 1. Paterson-Brown
S,
Garden
OJ,
Carter
DC.
cholecystectomy. Br J Surg 1991; 78: 131-32. 2. Editorial. Cholecystectomy practice transformed.
Laparoscopic
Lancet 1991; 338:
789-90. 3. Schirmer
BD, Edge SB,
Dix
J, Hyser MS, Hanks JB, Jones RS.
Laparoscopic cholecystectomy. Ann Surg 1991; 213: 665-77. 4. Gadacz TR, Talamini MA. Traditional vs laparoscopic cholecystectomy. Am J Surg 1991; 161: 336-38. 5. Larson GM, Vitale GC, Casey J, et al. Multipractice analysis of laparoscopic cholecystectomy in 1983 patients. Am J Surg 1992; 163: 221-26.
1991; 213: 1-2.
GC, Collet D, Larson GM, Cheadle WG, Miller FB, Perinot J. Interruption of professional and home activity after laparoscopic cholecystectomy among French and American patients. Am J Surg
28. Vitale
1991; 161: 396-98.
The anniversary meeting of the British Medical Association is to be held at the British coffee-house, Cockspur-street, on Tuesday evening next.... This Association, from its birth, has kept steadily in the right path. The proceedings have not been characterised by trickery or by shuffling. The Council, and the members generally, have boldly and courageously encountered the abuses of the Colleges, the imperfect state of medical law, and the absence ofa just legal protection for medical practitioners in the exercise of their professional rights, and it has resolutely and perseveringly pointed to the only substantial and effective remedies. Such an association is entitled to the respectful confidence of the profession ....
(Nov 5, 1842)
