Laparoscopic Versus Open Cholecystectomy: A Prospective Matched-Cohort Study

To compare the results of laparoscopic cholecystectomy (LC) and open cholecystectomy (OC) for symptomatic cholelithiasis in elective surgery we performed a prospective matched-cohort study. Hundred consecutive patients who underwent LC in the period Sept. 1990-June 1992, and 100 patients who were age and sex matched and underwent an elective OC in the foregoing two years (1989-1990) were studied. The median operation time for LC (75, 40-180 min) was significantly longer than for OC (55, 20-155 min; p < 0.001). Postoperative hospitalization was significantly shorter after LC (3, 1-16 days), compared with OC (7, 4-22 days; p<0.001). Conversion of LC to OC occurred in 12 (12%) patients initially scheduled to undergo LC. Complications occurred in 5 patients (5%) after LC and in 5 patients (5%) after OC. The calculated expenses (operation and postoperative hospitalization, 3rd class) were approximately fl. 3740,- for LC (excl. investments for pieces of apparatus) and fl. 6725,- for OC. This study demonstrates that LC can be performed safely with the number of complications comparable to those for OC. Bile duct injury is a serious potential threat. The main advantages ofLC are the minimal trauma, with more rapid recovery. Insurers seem to benefit from reduced postoperative disability and earlier discharge.


INTRODUCTION
The introduction of laparoscopic cholecystectomy (LC) has been a significant milestone in the treatment of gallstone disease. By this method the gallbladder is removed through four small incisions using laparoscopic techniques. Since its introduction in France in 19894 it has rapidly become available in the rest of Europe and the United States. In many centers LC is now the treatment of choice for symptomatic 138 cholelithiasis , 16,18. Although the effectiveness of this technique has been suggested by several studies, the number of clinical trials comparing LC with the "gold standard", the traditional open cholecystectomy (OC), are sparse. The earliest reports have stressed the However, more recent reports warn of the higher risk of common bile duct injuries that may be associated with LC14 '19. To establish the advantages and potential risks of LC, studies comparing this technique with the OC are necessary. However, after the widespread introduction and acceptance of LC among the more educated and demanding patient population it has become practically impossible to perform prospective randomised trials.
In our hospital, data on diagnosis, treatment and complications of each patient are prospectively collected and stored in a computerized data base. We used these prospectively collected data to compare the results of LC over a 2-year period and the results ofOC performed in a foregoing 2-year period. Studied were differences in operation time, postoperative hospitalization, complications and costs.

PATIENTS AND METHODS RESULTS
All operations were performed under general anaesthesia and all patients received prophylactic antibiotics and subcutaneous heparin during the perioperative period.
Laparoscopic Cholecystectomy The first LC was performed in our hospital in September 1990. LC was performed by a standardized procedure11. After the cyctic artery and cystic duct were identified, they were clipped and transsected. The gallbladder was dissected by electrocautery and removed through the small umbilical wound. All wounds were closed with or 2 fascia sutures and some skin sutures or steristrips.
Open Cholecystectomy A control group was constituted of patients who underwent an OC in a 2-year period (1989)(1990), preceding the introduction of LC in our hospital. During this period 238 patients underwent OC. This was on an elective basis in 196 cases and an acute procedure in 42 cases. In the group of 196 elective procedures exploration of the common bile duct or another coincidental (abdominal) surgical procedure was necessary in 10 patients. These patients were excluded from the study. Of the remaining 186 cases, 100 patients were age and sex matched with the LC group.
A traditional OC was performed through a right subcostal incision. All data were prospectively collected and obtained form a computerized data base.

Statistics
Data were subjected to computerized analysis (PATFILE statistical package). Continuous variables are given as median (range) and were analyzed by a nonparametric test for independent variables, the Mann-whitney test. Values for p < 0.05 were considered to be significant.
Since the OC-group was matched for age and sex with the LC-group, these data were comparable for both groups. The median (range) age in the LC-group was 43 (18-88) years and 43 (20-79) years in the OCgroup. The male/female ratio was 19/81% for both groups.
Data of all patients was analyzed according to the intention-to-treat principle, which means that patients who were scheduled to undergo LC but instead underwent OC for any reason, remained in the LC group.
Conversion of Laparoscopic to Open Cholecystectomy In our hospital a very liberal policy regarding the indication for conversion is used. Unnecessary risks are avoided and laparotomy is performed in all cases in which the anatomy is unclear or complications, that cannot be controlled laparoscopically, occur.
Conversion of LC to OC occured in 12 (12%) of the 100 patients initially scheduled to undergo LC. The reasons for conversion to laparotomy included difficult dissection due to adhesions after previous upperabdominal operations or chronic cholecystitis (five patients), a stone in the cystic duct (two patients) and leakage of bile or bleeding (two patients). Two other patients required laparotomy because the gallbladder or stones were lost in the abdominal cavity and could not be found by laparoscopy. The latter is currently no longer considered as an indication for laparotomy since it has become clear that these stone are not harmfull and do not cause problems postoperatively. In one morbidly obese patient laparotomy was needed because the trocars were to short to pass the pannus and enter the abdominal cavity.

Operation Time
The median operation time for LC was significantly longer than for OC. The distribution of operation time among both groups is depicted in Figure 1. The median (range) operation time for LC was 75 (40-180) min and 55 (20-155) min for OC (p < 0.001). During the study period operation time for LC showed a tendency to become shorter.

Hospitalization Period
The median postoperative hospitalization period for LC was significantly shorter than for OC. The distribution of postoperative hospitalization time among both groups is presented in Figure 2. The median postoperative hospital stay was 3 (1-16) days after LC and 7 (4-22) days after OC (P < 0.001). We compared the results of LC and OC in patients referred from the same population to one general hospital. Patient groups were formed prospectively. Potential bias by differences in age and sex were excluded by matching both groups for those two variables.
Conversion of LC to OC occurred in 12% of the patients. This is comparable with a series reported by Grace et al. 6, who did 6 laparotomies in a series of 50 patients (12%) who underwent LC for both acute and chronic gallbladder disease. Others have reported conversion rates that are somewhat lower (3%-5.3%) 3,8,15,18. This difference can undoubtedly be ascribed to our liberal policy for conversion and the learning curve with the procedure. In our opinion, conversion should never be seen as a complication or failure. Hesitation to perform a laparotomy when this is indicated and inevitable, may lead to serious and irreversible morbidity. For this reason, it is important that patients are always informed preoperatively about the possibility of conversion.
Complications were seen in 5% of the patients in both groups, which is comparable to the complication rates published by others. In different studies complications were seen in 1.6%-6% of the patient 3,8,15,19. In our series, most complications could be qualified as minor. A serious complication occurred in one patient with an accidental partial occlusion of the common bile duct by a clip. Following the initial enthousiastic reports on LC, later studies have stressed the risk of increased number of complications, especially common bile duct injuries 14,19. It has become clear that common bile duct injuries are a potential threat in LC. The reported incidence of bile duct injuries for OC ranges from 0%-0.5% 5,7 The incidence of common bile duct injury in laparoscopic series ranges from 0-3% 1,3,13. Several investigators have recommended routine operative cholangiography to define more clearly the common bile duct anatomy and thereby reduce the risk of bile duct injury 2,12. Currently, there are no convincing data that operative cholangiography should be performed routinely as part of LC. In fact, we did not use intraoperative cholangiography in any patient. At this point in the development of LC the incidence of bile duct injury appears to correlate more closely with the operative experience of the surgeon than with a particular policy towards operative cholangiography.
The decreased length of hospitalization associated with LC has been considered to be one of its main benefits. In our series we found a median postoperative hospital stay of 3 days compared to 7 days after OC. These data are comparable to those published by others3,6,15. The faster recovery and earlier discharge resulted in a reduction in costs for patients insured by the National Health Insurance. Others have also reported lower costs for patients undergoing LC in Ireland6. This situation may be completely different in other countries. Indeed, in the U.S.A. higher expenses have been reported for LC than for OC15. The true costs for the hospital, however, are also higher in our situation, due to the higher operating room expenses, including longer duration of the operation and use of costly equipment needed for this procedure, such as disposable trocars and clip appliers. This should also be considered when evaluating the financial aspects of this new technique. This study demonstrates that LC can be performed safely with the number of complications comparable to those for conventional, open cholecystectomy. Bile duct injury is a serious potential threat. The main advantages of LC are the minimal trauma, with more rapid recovery. Insurers seem to benefit from reduced postoperative disability and earlier discharge. Although LC is becoming the treatment of choice for symptomatic cholelithiasis, continued study and longterm follow ups are needed to evaluate and further develop this relatively young surgical technique. Besides proper training of unexperienced surgeons, adequate education of endoscopic techniques, both in