Surg Endosc (1998) 12: 148–150
© Springer-Verlag New York Inc. 1998
A diagnostic score to predict the difficulty of a laparoscopic cholecystectomy from preoperative variables P. Schrenk, R. Woisetschla¨ger, R. Rieger, W. U. Wayand Second Department of Surgery, Krankenhausstraße 9, A-4020 Linz, Austria, and Ludwig Boltzmann Institute for Surgical Laparoscopy, AKH Linz, Krankenhausstraße 9, A-4020 Linz, Austria Received: 9 January 1994/Accepted: 8 May 1997
Abstract Background: Modified logistic regression analysis of 24 variables in 300 patients undergoing laparoscopic cholecystectomy found the following parameters independently predictive for a difficult operation: right upper quadrant pain (p < 0.01), rigidity in right upper abdomen (p < 0.01), previous upper abdominal surgery (p < 0.01), biliary colic within the last 3 weeks (p < 0.05), white blood cell count >10 × 109/l (p < 0.05), thickening of the gallbladder wall (p < 0.05), hydroptic gallbladder (p < 0.05), pericholecystic fluid (p < 0.01), shrunken gallbladder (p < 0.01), and no filling of the gallbladder in preoperative intravenous cholangiography (p < 0.05). Methods: Based on these variables a diagnostic model was developed to predict the difficulty of a laparoscopic cholecystectomy, with scores ranging from 0 (ideal case) to IV (conversion to open cholecystectomy expected) prior to surgery. Results: When the reliability of our model was examined in a second study in 340 consecutive patients undergoing laparoscopic cholecystectomy 80% of the patients were predicted correctly. Conclusions: Our model should help to select patients for either laparoscopic or open cholecystectomy based on the expected difficulties and the experience of the surgeon. Key words: Laparoscopic cholecystectomy — Difficult laparoscopic cholecystectomy — Score model — Conversion to open cholecystectomy
Increasing experience with laparoscopic cholecystectomy (LC) led to an expansion of the indications for LC and more difficult cases are operated laparoscopically [3, 10]. Although these patients will also benefit from laparoscopic
Correspondence to: P. Schrenk
surgery [4], a difficult case has a higher risk for conversion and complication [1, 3, 10, 13]. Several factors have been found to be associated with a difficult case [2, 7, 11], but no reliable criteria are available yet to identify patients with a difficult laparoscopic cholecystectomy from preoperative variables. The purpose of our study was to identify parameters associated with a difficult case. Based on these variables we developed a model to predict the difficulty of a LC prior to surgery. Patients and methods Twenty-four parameters were collected prospectively in 300 patients (187 women, 113 men; mean age 53.7 ± 14.4 years, range 19–75 years) undergoing laparoscopic cholecystectomy from October 1991 to August 1992 and entered in a modified logistic regression analysis [5, 6, 9] to identify variables associated with a difficult LC. The following data were included in the analysis as binary data: sex (male), age > 65 years, previous upper abdominal surgery, previous lower abdominal surgery, gallstone history > 3 years, biliary colic within the last 3 weeks, morbid obesity (Body Mass Index [BMI] > 35%; BMI defined as weight in kilograms divided by the height in meters squared), pain in right upper abdomen, rigidity in right upper abdomen, temperature > 38.0°C, preoperative ERCP, serum amylasis > 120 IU/l, BSR > 20/40, WBC > 10 × 109/l, serum bilirubin > 1.2 mg/dl, alkaline phosphatase > 170 IU/l and g-glutamyl transpeptidase > 25 IU/l, SGOT > 19 IU/l, and SGPT > 25 IU/l, gallstone(s) > 2 cm on ultrasound, thickened gallbladder wall (more than 5 mm on ultrasound), pericholecystic fluid, incarcerated cystic duct stone, hydrops of the gallbladder, shrunken gallbladder, and no visualization of the gallbladder in preoperative intravenous cholangiography. Data on patients in whom LC was ‘‘easy’’ to perform were compared to those for patients in whom difficulties in preparation of ductal or vessel structures or dissection of the gallbladder were encountered. Whether a LC was classified as a difficult or an easy case was due to the judgment of the surgeon and depended on the degree of adhesion to the gallbladder, presence of scarring in the triangle of Calot, acute or chronic inflammatory changes, and duration of surgery. When time necessary for dissection of the gallbladder exceeded 60 min, the LC was rated as difficult. Failure of the laparoscopic equipment and inadequate exposure of the operative field, e.g., in the case of an enlarged left liver lobe, were not taken into account: 197 of 300 laparoscopic cholecystectomies (65.7%) were rated ‘‘not easy.’’ Patients with perforated or gangrenous gallbladder or portal hypertension with severe coagulopathy were considered to be contraindicated to LC. Analyses were performed on an IBM computer using a statistical software package developed at the Institute of Statistics, University of Linz, Austria. A p value of less than 0.05 was regarded as significant.
149 Table 1. Score model to predict the difficulty of LCa
Laboratory Preoperative i.v. cholangiography Ultrasonography
Anamnestical
Clinical
Table 2. Results of the score model in 340 patients
Parameters
Points
Preoperative score
WBC > 10 × 109/l No visualization of the gallbladder Either thickened gallbladder wall, hydroptic Gallbladder, or pericholecystic fluid Shrunken gallbladder Previous upper abdominal surgery Biliary colic within the last 3 weeks Right upper quadrant pain Rigidity in right upper abdomen Total
1 1
0 (n 4 118)
I (n 4 97)
II (n 4 64)
III (n 4 30)
IV (n 4 31)
101 (86%) 17
72 (74%) 25
50 (78%) 14
23 (77%) 6
26 (84%) 5
1
1 2 1
Predicted correctly Predicted wrong Underestimated Overestimated Conversion
0
13
14
6
5
17 1
12 0
0 0
0 1
0 26
1 1 9
a
For each parameter a distinctive number of points is allocated, with a total of 9 points available. 0 points 4 score 0 4 easy gallbladder, ideal for LC; 1 point 4 score I 4 little difficulties expected; 2 points 4 score II, more difficulties expected; 3 points 4 score III 4 severe case; 4 points or more 4 score IV 4 conversion to open cholecystectomy expected
The score model Data that were found to be statistically significantly associated with a difficult case were incorporated in a score model with five categories ranging from 0 (ideal for LC) to IV (conversion to open cholecystectomy expected) (Table 1). A distinctive number of points was allocated to each parameter found in the analysis. Previous upper abdominal surgery was allocated 2 points, which was arbitrary. If no preoperative intravenous cholangiography was performed routinely, an incarcerated cystic duct stone found ultrasonographically could be used instead. The points (0 to 9 points possible) were summed to develop the score as follows: 0 points 4 score 0 4 easy gallbladder, ideal for LC; 1 point 4 score I 4 little difficulties expected; 2 points 4 score II 4 more difficulties expected; 3 points 4 score III 4 severe case; 4 points or more 4 score IV 4 conversion to open cholecystectomy expected.
Reliability of the model In a second study the reliability of the model was tested in 340 consecutive patients (221 women, 119 men; mean age 54.2 ± 6.3 years, range 18–74 years) undergoing LC from October 1992 to November 1993. For each patient the score was determined preoperatively and was then compared to the difficulty of the gallbladder found intraoperatively, which was judged by the surgeon on a scale from 0 (ideal for LC) to IV (converted to open cholecystectomy).
Results The study identified 11 of 24 variables as associated with a difficult LC: previous upper abdominal surgery (p < 0.01), biliary colic within the last 3 weeks (p < 0.05), pain (p < 0.01) and rigidity (p < 0.01) in right upper abdomen, WBC > 10 × 109/l (p < 0.05), ultrasonographically found thickened (>5 mm) gallbladder wall (p < 0.05), pericholecystic fluid (p < 0.01), incarcerated cystic duct stone (p < 0.05), hydrops of the gallbladder (p < 0.05), shrunken gallbladder (p < 0.01), and no visualization of the gallbladder in preoperative intravenous cholangiography (p < 0.05). The most reliable factors associated with a difficult LC were variables found with acute cholecystitis and included clinical and laboratory data as well as findings in ultrasonography. Mor-
bid obesity was found to be associated with difficulties in insertion of the trocars or visualization of the triangle of Calot but did not result in a difficult gallbladder. When the model was applied to 340 consecutive patients to validate the accuracy of the model, the overall predictability rate was 80%. Scores of 0 (easy case; 86% predicted correctly) and IV (conversion to open cholecystectomy; 84% predicted correctly) were predicted with the most accuracy (Table 2). Of 340 patients, 28 had to be converted to open cholecystectomy; 26 conversions (93%) were predicted from the model prior to operation; one patient who had to be converted was preoperatively evaluated as a severe case, whereas the other conversion was due to failure of the equipment in an otherwise easy case.
Discussion Conversion rate and complications associated with laparoscopic cholecystectomies depend on the experience of the surgeon and the difficulty of the individual case [3, 8, 12, 13]. Criteria to evaluate the difficulty of an LC prior to operation would be helpful in planning an LC, for a difficult case may be operated either by a more experienced surgeon or may be selected to convert to open cholecystectomy (OC). Risk factors associated with a difficult LC such as acute cholecystitis and previous upper abdominal surgery have been described before [2, 7, 11]. In our study we identified several variables associated with a difficult case, and based on these parameters we developed a model to predict the difficulty of LC prior to surgery. Although the ideal patient for a laparoscopic cholecystectomy as well as a conversion to open cholecystectomy was predicted correctly, for most of the patients the model contains several disadvantages. The assessment of the intraoperative difficulty of an LC was due to the individual judgment of the surgeon. With growing experience with laparoscopic surgery, a previously difficult case might be assessed as an easy one, but it might still remain difficult for an inexperienced surgeon. There were patients in whom an easy gallbladder was found intraoperatively when a difficult LC was expected from preoperative features; as well, there were patients expected to have an easy gallbladder, which turned out to be a difficult case. We have no explanation for why our model failed to predict these patients correctly, but one possibility is recurrent inflammatory episodes not men-
150
tioned by the patients resulting in dense scarring in the triangle of Calot. However, duration of history of gallbladder disease was not significant for a difficult LC in the analysis. As experience with LC and the model increased we noticed that the clinical findings of an acute inflammation such as pain and/or rigidity in the right upper abdomen proved to be the most important factor predicting a difficult gallbladder, whereas white blood cell count and an incarcerated cystic duct stone became less important. The accuracy of ultrasonographic data, however, was strongly related to the experience of the radiologist performing the examination. Surgeons just beginning with laparoscopic surgery or who have only a little experience with the procedure should only operate on patients with a score of 0 or I. When a score of II or more is expected from preoperative evaluation a more experienced surgeon should perform the operation. Patients with a score of IV are expected to require a conversion to open cholecystectomy 84% of the time. In these cases we inform the patients that there is a higher possibility for conversion, and together with the patient, the surgeon decides whether to perform a scheduled open cholecystectomy or to attempt an LC with a low threshold for conversion. In conclusion, our model provided five categories predicting the difficulty of LC prior to surgery. It should help to avoid an inexperienced laparoscopic surgeon having to operate on a gallbladder which is too difficult for him to perform laparoscopically. As a consequence of the results of our study we preoperatively determine the difficulty score in every patient and, depending on the difficulties expected from preoperative evaluation, elect a surgeon for the operation.
References 1. Deziel DJ, Millikan KW, Economou SG, Doolas A, Ko ST, Airan MC (1993) Complications of laparoscopic cholecystectomy: a national survey of 4292 hospitals and an analysis of 77604 cases. Am J Surg 165: 9–14 2. Fabre JM, Pyda P, de Seguin des Hons C, Lepage B, Balmes M, Baumel H, Domergue J (1992) Evaluation of the laparoscopic cholecystectomy on patients with simple and complicated cholecystolithiasis. World J Surg 16: 113–117 3. Frazee RC, Roberts JW, Symmonds R, Snyder SK, Hendricks J, Smith R, Custer MD (1992) What are the contraindications for laparoscopic cholecystectomy? Am J Surg 164: 491–495 4. Fuchs KH, Freys SM, Heimbucher J, Thiede A (1992) Laparoskopische Cholecystektomie—Lohnt sich die laparoskopische Technik in ‘‘schwierigen’’ Fa¨llen? Chirurg 63: 296–304 5. Haberman SJ (1978) Analysis of quantitative data. Academic Press, New York 6. Immich H (1974) Medizinische Statistik. F.K. Schattauer Verlag, Stuttgart 7. Jorgensen JO, Hunt DR (1993) Laparoscopic cholecystectomy: a prospective analysis of the potential causes of failure. Surg Laparosc Endosc 3: 49–53 8. Lee VS, Chari RS, Cucchiaro G, Meyers WC (1993) Complications of laparoscopic cholecystectomy. Am J Surg 165: 527–532 9. Maddala GS (1983) Limited dependent and qualitative variables in econometrics. Cambridge University Press, Cambridge 10. Peters JH, Miller J, Nichols KE, Ollila D, Avrodopolous D (1993) Laparoscopic cholecystectomy in patients admitted with acute biliary symptoms. Am J Surg 166: 300–303 11. Rattner DW, Ferguson C, Warshaw AL (1993) Factors associated with successful laparoscopic cholecystectomy for acute cholecystitis. Ann Surg 217: 233–236 12. The Southern Surgeons Club (1991) A prospective analysis of 1518 laparoscopic cholecystectomies. N Engl J Med 324: 1073–1078 13. Williams LF, Chapman WC, Bonau RA, McGee EC, Boyd RW, Jacobs JK (1993) Comparison of laparoscopic cholecystectomy with open cholecystectomy in a single center. Am J Surg 165: 459–465