1992, The British Journal of Radiology, 65, 581-584
An audit of ultrasound diagnosis of gallbladder calculi By J . Walker, BSc, MRCP, D M R D , *R. T. A. Chalmers, M B , ChB, FRCS and P. L Allan, BSc, FRCPE, FRCR Departments of Radiology and "Surgery, Royal Infirmary, Lauriston Place, Edinburgh EH3 9YW, UK
Based on a paper presented at the BMUS 23rd Annual Meeting in Bournemouth, on 10-12 December 1991 Keywords: Gallstones, Ultrasound, Cholecystectomy
Abstract. Many series suggest that ultrasound is an accurate method for demonstrating cholelithiasis. However, these series were often prospective and the examinations performed by experienced sonographers. This audit addresses whether the accuracy is maintained in daily practice. We reviewed the ultrasound scans of 128 patients who underwent cholecystectomy for cholelithiasis and compared the findings. The operative and ultrasound findings were at variance in eight of the 128 patients (6.2%). Five were reported as having gallstones on ultrasound but none were found at cholecystectomy, a false positive rate for ultrasound of 3.9%. Three had abnormal gallbladders with no gallstones on ultrasound but gallstones were found at operation, a false negative rate of 2.3%. To avoid false positive diagnoses, suboptimal examinations should be repeated and the scan should be repeated immediately pre-operatively if only small calculi are seen. Alternative imaging should be performed if necessary, either cholescintigraphy in the acute situation or elective oral cholecystography. Some false negative examinations may be avoided by performing repeat examinations if the gallbladder is thick-walled and tender. With these provisos we conclude that ultrasound correctly diagnoses cholelithiasis in daily practice.
In the Royal Infirmary of Edinburgh, ultrasound is used as the standard pre-operative investigation in patients presumed to have cholelithiasis. It was suggested to us by one general surgical unit in the Royal Infirmary that on occasion there is some disparity between the ultrasound findings and those of cholecystectomy. We therefore reviewed the published literature and performed an audit of our own results. Cooperberg & Burhenne (1980) first suggested that real-time ultrasound was the diagnostic technique of choice for calculous gallbladder disease. In 313 patients they achieved an accuracy of 96%. They reported five false negative results, one definite false positive result and two others as possible false positives; six examinations were inadequate. Other published series also quote an accuracy of 96% or greater (Crade et al, 1978; Hessler et al, 1981). However, often these figures relate to prospective series where the examinations are performed by a senior, highly motivated sonographer. In our retrospective review, we sought to audit the daily practice of an ultrasound department where sonographers of varying experience are scanning routine and emergency referrals.
radiologist performing the scan and operative details were noted. The ultrasound scanners used were an Acuson 128, a Siemens AC or a Siemens Sonoline SX using 3.5 or 5 MHz transducers. In cases where the ultrasound report was found to be at variance with the findings at cholecystectomy, the report and hard-copy films from the scan were reviewed by a consultant radiologist (P.L.A.). The pathologist's report on the excised gallbladder was also reviewed. Results
A total of 144 cholecystectomy patients were reviewed. Of these, 35 were male and 109 female; seven were emergency operations. 128 had pre-operative ultrasound scans performed (see Table I). Of the others, 14 underwent oral cholecystography and two had calcified gallstones noted on plain radiographs.
Table I. Comparison of the ultrasound detection of gallstones with the findings at cholecystectomy Ultrasound
Methods
Using the Lothian Surgical Audit database, every patient who had undergone cholecystectomy for cholelithiasis in one general surgical unit in the Royal Infirmary of Edinburgh between March 1988 and March 1991 was identified. Individual patient case notes and copies of the pre-operative ultrasound reports were reviewed. The ultrasound findings, the grade of the
Cholecystectomy Stones No stones
115" 5
Total
120
Address correspondence to Dr J. Walker.
" At cholecystectomy, two of these patients had no stones in the gallbladder but stones were found in the common bile duct.
Vol. 65, No. 775
Stones
No stones
581
J. Walker, R. T. A. Chalmers and P. L. Allan
Figure 1. False positive ultrasound examination. A highly reflective structure casting an acoustic shadow is present in the gallbladder. This is still felt likely to represent a gallstone.
There was a false positive rate for ultrasound of 5.4% (7/128) for calculi in the gallbladder and a false negative rate of 2.3% (3/128). In two of the seven false positives, however, though there were no calculi in the gallbladder, calculi were found in the common bile duct at surgery. It is considered likely that these stones had migrated in the interval between ultrasound and surgery, the interval being 10 days in one case and 21 days in the other. Therefore, excluding these as false positives because of the presence of gallstones, the revised false positive rate was 3.9% (5/128). The overall error rate for this audit is 6.2% (8/128). Of the 128 pre-operative scans performed, 65 were performed by consultants, 48 by senior registrars and 15 by registrars. Of the erroneous scans, three were performed by a consultant radiologist, four by a senior
Figure 2. False positive ultrasound examination. A small reflective structure is present in the gallbladder, which may or may not represent a gallstone.
582
Figure 3. False positive ultrasound examination. Only sludge is seen to be present in the gallbladder; an error may have been made.
registrar and one by a registrar. Respectively, 4.6% (3/65) of the consultants' scans were erroneous, 8.3% (4/48) of the senior registrars' and 7% (1/15) of the registrars'. These results were not statistically significantly different by a x2 test. The hard-copy films of the incorrect scans and the pathology reports on the excised gallbladders of those patients were reviewed. All five of the false positives had chronic cholecystitis. The hard-copy films of three of the five patients were available. In one, an opacity that had definite characteristics of a gallstone was present (Fig. 1). The common duct was dilated at 14 mm and, though the interval between ultrasound and operation was only one day, it was possible that the stone had been passed. In the second case, a small opacity without an acoustic shadow was present, which had been interpreted as a gallstone (Fig. 2). In the third, sludge and calculi were reported but only sludge was present on review (Fig. 3). The three false negative examinations were all in acutely ill patients and, though no gallstones were seen, all three gallbladders were reported as abnormal and dilated. In addition, one was commented on as having a thick and oedematous wall, one as being tender with a thick wall, and one as tender and oedematous. The first two patients were given the diagnoses of acute cholecystitis (Fig. 4) and acalculous cholecystitis, respectively; the third was given no specific diagnosis. Thus, although no gallstones were identified, the site of the patient's pathology was correctly identified and appropriate surgery performed. Discussion Ultrasound is commonly used as the imaging investigation of choice in calculous gallbladder disease (Cooperberg & Burhenne, 1980) and is used routinely in our department. It is convenient and safe (merely requiring an 8 h patient fast and involving neither radiaThe British Journal of Radiology, July 1992
An audit of ultrasound diagnosis of gallbladder calculi
Figure 4. False negative ultrasound examination. At ultrasound no gallstones were identified but the gallbladder was described as being oedematous and thick-walled and the diagnosis of acute cholecystitis was given.
tion nor drugs), can be extended to visualize the biliary tree and other upper abdominal organs, and is accurate. This accuracy has been determined previously and has been reported as 96% or greater, but these were mostly prospective studies in which the examinations were performed by highly experienced sonographers (Cooperberg & Burhenne, 1980; Crade et al, 1978; Hessler et al, 1981). Our audit of the day-to-day work in a department where both emergency and elective patients are scanned by sonologists of varying experience gave an overall error rate of 6.2% for patients coming to surgery. This compares favourably with the previously reported series and confirms the ability of ultrasound to correctly diagnose cholelithiasis. However, the audit only includes patients coming to surgery and no information is available on those false negative results where gallstones were missed on ultrasound and no surgery has been performed. In our series we diagnosed gallstones in five patients where no gallstones were found at cholecystectomy. These false positive diagnoses must be avoided (Hessler et al, 1981) as unnecessary surgery may be performed. They may arise because gallstones are passed in the interval between ultrasound and surgery. They may also arise because of the mistaken identification of other structures, such as the spiral valve of Heister, polyps or folds in the gallbladdder wall, as calculi. Technically difficult examinations, such as in obese or ill patients, can result in inadequate, erroneous examinations. Thus, an optimal ultrasound examination should be performed when the patient is adequately fasted and can be moved into different positions. It may also be prudent to repeat the ultrasound immediately pre-operatively if there is an interval of several days between the scan and operation, especially if only small calculi are seen initially (Cooperberg & Gibney, 1987). If doubt still persists as to whether highly reflective foci are calculi or Vol. 65, No. 775
not, an alternative imaging investigation should be performed; in an elective situation oral cholecystography is recommended. This has recently been reported as having a detection rate for biliary calculi differing little from ultrasound (Gelfand et al, 1988), although it requires more rigorous patient preparation, the ingestion of contrast medium and the use of radiation. In a patient presenting as an emergency with the suspected diagnosis of acute cholecystitis and in whom the ultrasound is equivocal or suboptimal, cholescintigraphy is the alternative imaging technique of choice. Failure of uptake by the gallbladder confirms cystic duct obstruction and is strongly suggestive of acute cholecystitis (Samuels et al, 1985). In our three patients with false negative diagnosis of calculi, all three gallbladders had been reported at ultrasound as abnormal, with dilatation, thickened, oedematous walls and tenderness. A recent paper (Ekberg & Weiber, 1991) has recognized the clinical importance of thick-walled, tender gallbladders without stones on ultrasound. In their series of 31 patients, follow-up was available in 28. They reported that further imaging (oral cholecystography, intravenous cholangiography or ultrasound) demonstrated calculi in 14 patients. Thus, to avoid some (possibly 50%) of false negative diagnoses, when the gallbladder is abnormal but no stones are seen, either a repeat examination at a later date or an oral cholecystogram will be helpful. We conclude from this audit that ultrasound, as performed daily in our department, correctly diagnoses cholelithiasis. It should remain the diagnostic imaging technique of choice. However, if false positive diagnoses are to be avoided, the examination should be adequate, repeated if necessary and, if doubt persists, oral cholecystography or cholescintigraphy should be performed as indicated. Some false negative diagnoses may be avoided by repeating the scan or performing an oral cholecystogram if, at the initial examination, no stones are demonstrated but the gallbladder is tender and thick-walled. Acknowledgments We thank Mr I. F. McLaren and Mr K. Fearon for their help and advice. References COOPERBERG, P. L. & BURHENNE, H. J., 1980. Real-time ultra-
sonography. Diagnostic technique of choice in calculous gallbladder disease. New England Journal of Medicine, 302, 1277-1279. COOPERBERG, P. L. & GIBNEY, R. G.,
1987. Imaging of the
gallbladder. Radiology, 163, 605-613. CRADE, M., TAYLOR, K. J. W., ROSENFIELD, A. T., D E GRAFF,
C. S. & MINIHAN, P., 1978. Surgical and pathological correlation of cholecystosonography and cholecystography. American Journal of Roentgenology, 131, 227-229. EKBERG, O. & WEIBER, S., 1991. The clinical importance of a
thick walled, tender gallbladder without stones on ultrasonography. Clinical Radiology, 44, 38-41. GELFAND, D. W., WOLFMAN, N. T., OTT, D. J., WATSON, JR, N.
E., CHEN, Y. M. & DALE, W. J., 1988. Oral cholecystography
583
J. Walker, R. T. A. Chalmers and P. L. Allan vs gallbladder sonography: a prospective, blinded reappraisal. American Journal of Roentgenology, 151, 69-72. HESSLER, P. C , HILL, D. S., DETORIE, F. M. & Rocco, A. F., 1981. High accuracy sonographic recognition of gallstones. American Journal of Roentgenology, 136, 517-520.
584
SAMUELS, B. I., FREITAS, J. E. & GROSS, M. D., 1985. A pragmatic review of gallbladder imaging. Seminars in Ultrasound, Computerised Tomography and Magnetic Resonance, 6, 156-171.
The British Journal of Radiology, July 1992
An audit of ultrasound diagnosis of gallbladder calculi By J . Walker, BSc, MRCP, D M R D , *R. T. A. Chalmers, M B , ChB, FRCS and P. L Allan, BSc, FRCPE, FRCR Departments of Radiology and "Surgery, Royal Infirmary, Lauriston Place, Edinburgh EH3 9YW, UK
Based on a paper presented at the BMUS 23rd Annual Meeting in Bournemouth, on 10-12 December 1991 Keywords: Gallstones, Ultrasound, Cholecystectomy
Abstract. Many series suggest that ultrasound is an accurate method for demonstrating cholelithiasis. However, these series were often prospective and the examinations performed by experienced sonographers. This audit addresses whether the accuracy is maintained in daily practice. We reviewed the ultrasound scans of 128 patients who underwent cholecystectomy for cholelithiasis and compared the findings. The operative and ultrasound findings were at variance in eight of the 128 patients (6.2%). Five were reported as having gallstones on ultrasound but none were found at cholecystectomy, a false positive rate for ultrasound of 3.9%. Three had abnormal gallbladders with no gallstones on ultrasound but gallstones were found at operation, a false negative rate of 2.3%. To avoid false positive diagnoses, suboptimal examinations should be repeated and the scan should be repeated immediately pre-operatively if only small calculi are seen. Alternative imaging should be performed if necessary, either cholescintigraphy in the acute situation or elective oral cholecystography. Some false negative examinations may be avoided by performing repeat examinations if the gallbladder is thick-walled and tender. With these provisos we conclude that ultrasound correctly diagnoses cholelithiasis in daily practice.
In the Royal Infirmary of Edinburgh, ultrasound is used as the standard pre-operative investigation in patients presumed to have cholelithiasis. It was suggested to us by one general surgical unit in the Royal Infirmary that on occasion there is some disparity between the ultrasound findings and those of cholecystectomy. We therefore reviewed the published literature and performed an audit of our own results. Cooperberg & Burhenne (1980) first suggested that real-time ultrasound was the diagnostic technique of choice for calculous gallbladder disease. In 313 patients they achieved an accuracy of 96%. They reported five false negative results, one definite false positive result and two others as possible false positives; six examinations were inadequate. Other published series also quote an accuracy of 96% or greater (Crade et al, 1978; Hessler et al, 1981). However, often these figures relate to prospective series where the examinations are performed by a senior, highly motivated sonographer. In our retrospective review, we sought to audit the daily practice of an ultrasound department where sonographers of varying experience are scanning routine and emergency referrals.
radiologist performing the scan and operative details were noted. The ultrasound scanners used were an Acuson 128, a Siemens AC or a Siemens Sonoline SX using 3.5 or 5 MHz transducers. In cases where the ultrasound report was found to be at variance with the findings at cholecystectomy, the report and hard-copy films from the scan were reviewed by a consultant radiologist (P.L.A.). The pathologist's report on the excised gallbladder was also reviewed. Results
A total of 144 cholecystectomy patients were reviewed. Of these, 35 were male and 109 female; seven were emergency operations. 128 had pre-operative ultrasound scans performed (see Table I). Of the others, 14 underwent oral cholecystography and two had calcified gallstones noted on plain radiographs.
Table I. Comparison of the ultrasound detection of gallstones with the findings at cholecystectomy Ultrasound
Methods
Using the Lothian Surgical Audit database, every patient who had undergone cholecystectomy for cholelithiasis in one general surgical unit in the Royal Infirmary of Edinburgh between March 1988 and March 1991 was identified. Individual patient case notes and copies of the pre-operative ultrasound reports were reviewed. The ultrasound findings, the grade of the
Cholecystectomy Stones No stones
115" 5
Total
120
Address correspondence to Dr J. Walker.
" At cholecystectomy, two of these patients had no stones in the gallbladder but stones were found in the common bile duct.
Vol. 65, No. 775
Stones
No stones
581
J. Walker, R. T. A. Chalmers and P. L. Allan
Figure 1. False positive ultrasound examination. A highly reflective structure casting an acoustic shadow is present in the gallbladder. This is still felt likely to represent a gallstone.
There was a false positive rate for ultrasound of 5.4% (7/128) for calculi in the gallbladder and a false negative rate of 2.3% (3/128). In two of the seven false positives, however, though there were no calculi in the gallbladder, calculi were found in the common bile duct at surgery. It is considered likely that these stones had migrated in the interval between ultrasound and surgery, the interval being 10 days in one case and 21 days in the other. Therefore, excluding these as false positives because of the presence of gallstones, the revised false positive rate was 3.9% (5/128). The overall error rate for this audit is 6.2% (8/128). Of the 128 pre-operative scans performed, 65 were performed by consultants, 48 by senior registrars and 15 by registrars. Of the erroneous scans, three were performed by a consultant radiologist, four by a senior
Figure 2. False positive ultrasound examination. A small reflective structure is present in the gallbladder, which may or may not represent a gallstone.
582
Figure 3. False positive ultrasound examination. Only sludge is seen to be present in the gallbladder; an error may have been made.
registrar and one by a registrar. Respectively, 4.6% (3/65) of the consultants' scans were erroneous, 8.3% (4/48) of the senior registrars' and 7% (1/15) of the registrars'. These results were not statistically significantly different by a x2 test. The hard-copy films of the incorrect scans and the pathology reports on the excised gallbladders of those patients were reviewed. All five of the false positives had chronic cholecystitis. The hard-copy films of three of the five patients were available. In one, an opacity that had definite characteristics of a gallstone was present (Fig. 1). The common duct was dilated at 14 mm and, though the interval between ultrasound and operation was only one day, it was possible that the stone had been passed. In the second case, a small opacity without an acoustic shadow was present, which had been interpreted as a gallstone (Fig. 2). In the third, sludge and calculi were reported but only sludge was present on review (Fig. 3). The three false negative examinations were all in acutely ill patients and, though no gallstones were seen, all three gallbladders were reported as abnormal and dilated. In addition, one was commented on as having a thick and oedematous wall, one as being tender with a thick wall, and one as tender and oedematous. The first two patients were given the diagnoses of acute cholecystitis (Fig. 4) and acalculous cholecystitis, respectively; the third was given no specific diagnosis. Thus, although no gallstones were identified, the site of the patient's pathology was correctly identified and appropriate surgery performed. Discussion Ultrasound is commonly used as the imaging investigation of choice in calculous gallbladder disease (Cooperberg & Burhenne, 1980) and is used routinely in our department. It is convenient and safe (merely requiring an 8 h patient fast and involving neither radiaThe British Journal of Radiology, July 1992
An audit of ultrasound diagnosis of gallbladder calculi
Figure 4. False negative ultrasound examination. At ultrasound no gallstones were identified but the gallbladder was described as being oedematous and thick-walled and the diagnosis of acute cholecystitis was given.
tion nor drugs), can be extended to visualize the biliary tree and other upper abdominal organs, and is accurate. This accuracy has been determined previously and has been reported as 96% or greater, but these were mostly prospective studies in which the examinations were performed by highly experienced sonographers (Cooperberg & Burhenne, 1980; Crade et al, 1978; Hessler et al, 1981). Our audit of the day-to-day work in a department where both emergency and elective patients are scanned by sonologists of varying experience gave an overall error rate of 6.2% for patients coming to surgery. This compares favourably with the previously reported series and confirms the ability of ultrasound to correctly diagnose cholelithiasis. However, the audit only includes patients coming to surgery and no information is available on those false negative results where gallstones were missed on ultrasound and no surgery has been performed. In our series we diagnosed gallstones in five patients where no gallstones were found at cholecystectomy. These false positive diagnoses must be avoided (Hessler et al, 1981) as unnecessary surgery may be performed. They may arise because gallstones are passed in the interval between ultrasound and surgery. They may also arise because of the mistaken identification of other structures, such as the spiral valve of Heister, polyps or folds in the gallbladdder wall, as calculi. Technically difficult examinations, such as in obese or ill patients, can result in inadequate, erroneous examinations. Thus, an optimal ultrasound examination should be performed when the patient is adequately fasted and can be moved into different positions. It may also be prudent to repeat the ultrasound immediately pre-operatively if there is an interval of several days between the scan and operation, especially if only small calculi are seen initially (Cooperberg & Gibney, 1987). If doubt still persists as to whether highly reflective foci are calculi or Vol. 65, No. 775
not, an alternative imaging investigation should be performed; in an elective situation oral cholecystography is recommended. This has recently been reported as having a detection rate for biliary calculi differing little from ultrasound (Gelfand et al, 1988), although it requires more rigorous patient preparation, the ingestion of contrast medium and the use of radiation. In a patient presenting as an emergency with the suspected diagnosis of acute cholecystitis and in whom the ultrasound is equivocal or suboptimal, cholescintigraphy is the alternative imaging technique of choice. Failure of uptake by the gallbladder confirms cystic duct obstruction and is strongly suggestive of acute cholecystitis (Samuels et al, 1985). In our three patients with false negative diagnosis of calculi, all three gallbladders had been reported at ultrasound as abnormal, with dilatation, thickened, oedematous walls and tenderness. A recent paper (Ekberg & Weiber, 1991) has recognized the clinical importance of thick-walled, tender gallbladders without stones on ultrasound. In their series of 31 patients, follow-up was available in 28. They reported that further imaging (oral cholecystography, intravenous cholangiography or ultrasound) demonstrated calculi in 14 patients. Thus, to avoid some (possibly 50%) of false negative diagnoses, when the gallbladder is abnormal but no stones are seen, either a repeat examination at a later date or an oral cholecystogram will be helpful. We conclude from this audit that ultrasound, as performed daily in our department, correctly diagnoses cholelithiasis. It should remain the diagnostic imaging technique of choice. However, if false positive diagnoses are to be avoided, the examination should be adequate, repeated if necessary and, if doubt persists, oral cholecystography or cholescintigraphy should be performed as indicated. Some false negative diagnoses may be avoided by repeating the scan or performing an oral cholecystogram if, at the initial examination, no stones are demonstrated but the gallbladder is tender and thick-walled. Acknowledgments We thank Mr I. F. McLaren and Mr K. Fearon for their help and advice. References COOPERBERG, P. L. & BURHENNE, H. J., 1980. Real-time ultra-
sonography. Diagnostic technique of choice in calculous gallbladder disease. New England Journal of Medicine, 302, 1277-1279. COOPERBERG, P. L. & GIBNEY, R. G.,
1987. Imaging of the
gallbladder. Radiology, 163, 605-613. CRADE, M., TAYLOR, K. J. W., ROSENFIELD, A. T., D E GRAFF,
C. S. & MINIHAN, P., 1978. Surgical and pathological correlation of cholecystosonography and cholecystography. American Journal of Roentgenology, 131, 227-229. EKBERG, O. & WEIBER, S., 1991. The clinical importance of a
thick walled, tender gallbladder without stones on ultrasonography. Clinical Radiology, 44, 38-41. GELFAND, D. W., WOLFMAN, N. T., OTT, D. J., WATSON, JR, N.
E., CHEN, Y. M. & DALE, W. J., 1988. Oral cholecystography
583
J. Walker, R. T. A. Chalmers and P. L. Allan vs gallbladder sonography: a prospective, blinded reappraisal. American Journal of Roentgenology, 151, 69-72. HESSLER, P. C , HILL, D. S., DETORIE, F. M. & Rocco, A. F., 1981. High accuracy sonographic recognition of gallstones. American Journal of Roentgenology, 136, 517-520.
584
SAMUELS, B. I., FREITAS, J. E. & GROSS, M. D., 1985. A pragmatic review of gallbladder imaging. Seminars in Ultrasound, Computerised Tomography and Magnetic Resonance, 6, 156-171.
The British Journal of Radiology, July 1992
