Acute cholecystitis has no single clinical or laboratory finding with the level of diagnostic accuracy needed for diagnosis. Instead, the recommended diagnostic technique combines clinical observations with an abdominal ultrasound. The Tokyo Guidelines require one local sign or symptom, one systemic sign, and a confirmatory image test to form a diagnosis. However, these requirements may lead to under-diagnosis in patients presenting fewer symptoms. The Murphy’s sign test is a commonly used diagnostic tool. The physician applies pressure just below the ribs on the right side and asks the patient to inhale. Inhalation will bring the gallbladder into contact with the physician’s fingers, causing pain and an arrest in inspiration if the gallbladder is inflamed. Blood tests showing elevated levels of white blood cells (leukocytosis), elevated C-reactive protein, may be signs of infection and inflammation. Show
Imaging techniques are used to directly observe gallstones, gallbladder wall thickness, or cystic duct obstruction. The gallbladder wall is pathologically thickened if it is >3mm or wider. The two main imaging techniques used for cholecystitis are abdominal ultrasound and hepatobiliary scintigraphy (HIDA scan). Abdominal ultrasound is often the first test due to its widespread availability, lack of invasiveness, lack of ionizing radiation, and high accuracy in detecting gallbladder stones. Ultrasound can show the presence of stones, wall thickening and pericholecystic fluid. A hepatobiliary iminodiacetic acid (HIDA) scan tracks the production and flow of bile from the liver to the small intestine and shows blockage. This test involves the intravenous injection of HIDA, a radioactively labelled compound, secreted into bile. A specialized camera can detect the radioactivity, allowing it to trace the movement of this bile. If the gallbladder does not fill within an hour, the cystic duct is likely obstructed. Although it is the most sensitive and specific diagnostic tool, it is limited in its use because of limited availability, long testing time, and ionizing radiation exposure. It can also be inaccurate if bilirubin is elevated which is indicative of decreased ability of the liver to secrete compounds such as HIDA into bile. Cholangiography and computed tomography (CT) may also be used to identify cholecystitis, although their diagnostic accuracy is unknown. During CT scanning, a computer and x-rays are used to create a film showing cross-sectional images of certain tissue structures. During cholangiography, a contrast dye is injected into the bloodstream, which enables x-rays to create an image of the bile ducts. Additional imaging techniques include MRI. Diagnosis is often delayed in elderly patients, as the only symptoms may be an alteration of mental state or decreased food intake. Mild (grade 1): Moderate (grade 2): ● Elevated white cell count ● Palpable, tender mass in upper right quadrant ● Duration longer than 72 hours ● Local inflammation (could be biliary peritonitis, pericholecystic abscess, hepatic abscess, gangrenous cholecystitis, or emphysematous cholecystitis) Severe (Grade 3): ● Cardiovascular dysfunction ● Neurological dysfunction ● Respiratory dysfunction ● Renal dysfunction ● Hepatic dysfunction ● Hematologic dysfunction Acute Acalculous Cholecystitis (AAC) Diagnosis of AAC is often much more difficult than that of typical cholecystitis because it is much less common and patients often have severe concomitant medical problems that are the reason for hospital admission. The disease is generally suspected in any critically or chronically ill patient presenting with abdominal pain, fever and unexplained leukocytosis and sepsis. Blood tests may show increases in white blood cells (leukocytosis), transaminases, alkaline phosphatase, bilirubin, and amylase. As in calculous cholecystitis, ultrasound is the primary diagnostic imaging technique for AAC. Ultrasound may show gallbladder wall thickening greater than 5 mm, pericholecystic fluid, biliary sludge, gallbladder distention, gallbladder striation, mucosal peeling, air bubbles (emphysematous cholecystitis), and gallbladder perforation. Presence of at least two of the following is generally used to diagnose AAC: positive ultrasound Murphy’s sign, gallbladder wall thickening, gallbladder distension and pericholecystic fluid, in the absence of gallstones. CT and MRI are used when ultrasounds are inconclusive. An MRI can show gallbladder wall thickening, increased bile density, air bubbles, fluid buildup, and bleeding inside the gallbladder. Hepatobiliary scintigraphy is the best diagnostic tool for AAC because it can detect those without the disease 100% of the time. This test shows improper gallbladder filling in both AAC and calculous cholecystitis.
Acute cholecystitis is a sudden onset of inflammation of the gallbladder that causes severe abdominal pain. Abdominal pain is often accompanied by fever and abnormally high white blood cell count (leukocytes).1 Acute cholecystitis is usually caused by gallstones obstructing the cystic duct.2 This prevents the normal flow of bile in and out of the gallbladder into the bowel. Increased pressure in the gallbladder due to the obstruction results in inflammation and pain. Up to 14% of acute cholecystitis cases are acalculous.2 In these patients, there is an obstruction but gallstones are not the cause. Acute acalculous cholecystitis usually occurs in patients who are already critically ill from another medical condition. Mortality and morbidity is high in patients with acute acalculous cholecystitis. The initial treatment for acute cholecystitis in the emergency room is usually intravenous antibiotics, hydration, and analgesia. If inflammation of the gallbladder continues, removal of the gallbladder (cholecystectomy) is usually required.3 Complications of acute cholecystitis include gangrenous cholecystitis (gangrene of the gallbladder wall), gallbladder perforation (hole or piercing of the wall of the gallbladder), and emphysematous cholecystitis (acute infection of the gallbladder caused by gas-forming organisms). These complications occur in up to 20% of people with cholecystitis, have high mortality associated with them, and therefore require emergency surgery.2 Population: Patients with suspected acute cholecystitis. Intervention: Cholescintigraphy. Cholescintigraphy, also known as a hepatobiliary iminodiacetic (HIDA) scan, is a nuclear medicine test used to diagnose intrahepatic or extrahepatic obstruction of the bile ducts, gallbladder disease, and bile leaks. Before cholescintigraphy, patients are injected with a radiopharmaceutical tracer (technetium-99m [99mTc]-iminodiacetic acid). Patients need to fast three to four hours before this injection to avoid gallbladder contraction.4 After injection, a gamma camera is used to detect gamma rays emitted by the patient from the injected radiopharmaceuticals. Images are created from the detected gamma rays. If there is no cystic duct blockage, the radiopharmaceutical will enter the gallbladder, which will be visualized in images created by the gamma camera. If a gallstone is obstructing a patient's cystic duct, the radiopharmaceutical will not enter the gallbladder and visualization of the gallbladder cannot occur. Non-visualization of the gallbladder is indicative of acute cholecystitis. If the gallbladder is not seen one hour after injection, images should be retaken three to four hours after injection.5 This delayed imaging increases the specificity of cholescintigraphy for the diagnosis of acute cholecystitis. An alternative to delayed imaging is to inject the patient with a small amount of morphine sulphate (0.02 mcg/kg). Administration of morphine sulphate facilitates the flow of bile toward the cystic duct by causing contraction of the sphincter of Oddi. The injection of morphine sulphate can reduce the time to confirm the diagnosis from three or four hours to 1.5 hours.4 Comparators: For this report, the following diagnostic tests are considered as alternatives to cholescintigraphy:
Outcomes: Eleven outcomes (referred to as criteria) are considered in this report:
Definitions of the criteria are in Appendix 1. MethodsThe literature search was performed by an information specialist using a peer-reviewed search strategy. Published literature was identified by searching the following bibliographic databases: MEDLINE with In-Process records and daily updates via Ovid; The Cochrane Library (2011, Issue 2) via Wiley; and PubMed. The search strategy consisted of both controlled vocabulary, such as the National Library of Medicine's MeSH (Medical Subject Headings), and keywords. The main search concepts were radionuclide imaging and cholecystitis. Methodological filters were applied to limit retrieval to health technology assessments, systematic reviews, meta-analyses, randomized controlled trials, and non-randomized studies, including diagnostic accuracy studies. The search was limited to English language. No date limits were applied for the systematic review search. The primary studies search was limited to documents published between January 1, 1996, and March 2, 2011. Regular alerts were established to update the search until October 2011. Detailed search strategies are located in Appendix 2. Grey literature (literature that is not commercially published) was identified by searching relevant sections of the CADTH Grey Matters checklist. Google was used to search for additional web-based materials. The searches were supplemented by reviewing the bibliographies of key papers. See Appendix 2 for more information on the grey literature search strategy. Targeted searches were done as required for the criteria, using the aforementioned databases and Internet search engines. When no literature was identified that addressed specific criteria, experts were consulted. Search ResultsFourteen articles11-24 were identified through the MA/SR/HTA search; of those, eight13-18,21,24 underwent full text review. One systematic review15 was identified from the full text review that compared the diagnostic accuracy of cholescintigraphy with one of the alternative imaging modalities. A review of primary studies was conducted to identify studies that directly compared the diagnostic accuracy of cholescintigraphy with one of its alternatives. Four primary studies25-28 were found that compared cholescintigraphy with U/S. No primary studies were identified that directly compared cholescintigraphy with CT, with MRCP, or with ERCP. Articles from the grey literature search were used to address criterion 1 (one article)29 and criterion 8 (one article).30 Articles from the primary study search were used to help address criterion 1 (one article),31 criterion 3 (one article),32 criterion 6 (four articles), and criterion 8 (two articles). Literature from targeted searches was used to supplement the articles identified in the primary study search. When no literature was identified addressing specific criteria, experts were consulted. Summary tableTable 1: Summary of Criterion Evidence
Non–radiation-related Risks Cholescintigraphy Risks associated with a cholescintigraphy include allergy to HIDA and pain during CCK injection (causes gallbladder contraction), chills, nausea, and rash.40 CT Some patients may experience an allergic reaction to the contrast agent (if required).41 In addition, patients may experience mild side effects from the contrast agent, such as nausea, vomiting, or hives. A 2009 retrospective review of 456,930 intravascular doses of low-osmolar iodinated and Gd contrast materials administered between 2002 and 2006 found 0.15% of patients experienced side effects, most of which were mild. According to the American College of Radiology Manual on Contrast Media,42 the frequency of severe, life-threatening reactions with Gd is extremely rare (0.001% to 0.01%). MRCP MRCP is an MRI-based test and is contraindicated in patients with metallic implants, including pacemakers.43 MRI is often used in conjunction with the contrast agent Gd. Some patients may experience an allergic reaction to the contrast agent (if required).41 Gd is contraindicated in patients with renal failure or end-stage renal disease, as they are at risk of nephrogenic systemic fibrosis. The frequency of severe, life-threatening reactions with Gd is extremely rare (0.001% to 0.01%).42 U/S There are no reported risks associated with U/S in the literature that was reviewed.Radiation-related Risks Some tests expose patients to radiation. The following table presents the effective radiation dose to which patients are exposed during the various diagnostic tests.
Overall, the risks associated with cholescintigraphy using 99mTc-radiolabelled isotopes is:
As of 2006 in Canada, there were 2,034 diagnostic radiologists, 221 nuclear medicine physicians, 12,255 radiological technologists, 1,781 nuclear medicine technologists, and 2,900 sonographers available across Canada. Yukon, Northwest Territories, and Nunavut do not have the available personnel to perform and interpret tests to detect bile leak. Other jurisdictions (e.g., Prince Edward Island) may offer limited nuclear medicine services. Assuming the necessary equipment is available, if cholescintigraphy using 99mTc-radiolabelled isotopes is not available, it is estimated that:
For the diagnosis of acute cholecystitis, nuclear medicine facilities with gamma cameras (including SPECT) are required. As of January 1, 2007, there was an average of 18.4 nuclear medicine cameras per million people, with none available in the Yukon, Northwest Territories, or Nunavut.47 MRCP No MRI scanners are available in the Yukon, Northwest Territories, or Nunavut.48 According to CIHI's National Survey of Selected Medical Imaging Equipment database, the average number of hours of operation per week for MRI scanners in 2006-2007 ranged from 40 hours in Prince Edward Island to 99 hours in Ontario with a national average of 71 hours.47 In 2010, the average wait time for MR imaging in Canada was 9.8 weeks.49 CT No CT scanners are available in Nunavut.48 For CT scanners, the average weekly use ranged from 40 hours in Prince Edward Island to 69 hours in Ontario, with a national average of 60 hours.47 U/S The median wait time for a U/S in Canada was estimated to be 4.5 weeks in 2010.49 No information was found on the number of U/S machines available in Canada. Assuming the necessary expertise is available, if cholescintigraphy using 99mTc-radiolabelled isotopes is not available, it is estimated that:
According to our estimates, the cost of cholescintigraphy with 99mTc-based radioisotopes is $298.38. CT is minimally more costly and MRCP is moderately more costly. U/S is minimally less costly.
CCK = cholecystokinin; CIHI = Canadian Institute for Health Information; CT = computed tomography; Gd = Gadolinium; HIDA = hepatobiliary iminodiacetic acid; MRCP = magnetic resonance cholangiopancreatography; MRI = magnetic resonance imaging; mSv = millisievert; NR = not reported; SPECT = single-photon emission computed tomography; 99mTc = technetium-99m; U/S = ultrasound. Criterion 1: Size of affected population (link to definition) No estimates of point prevalence of acute cholecystitis were found in the literature. An Ontario study that estimated the annual incidence of acute cholecystitis was identified.29 Urbach and Stukel29 sought to find out whether the observed increased rate of elective cholecystectomy resulted in changes in the incidence of severe complications of gallbladder disease, including acute cholecystitis. Cases of severe gallbladder complications occurring from 1988 through 2000 in persons aged 18 years and older in Ontario were identified from hospital admission data from the Canadian Institute of Health Information (CIHI) and the Ontario Health Insurance Plan. Hospital admissions for acute cholecystitis were identified using specific ICD-9-CM codes. The authors estimated the average annual incidence rate of acute cholecystitis in Ontario during the years 1992-2000 to be 88.1 per 100,000 people. This is equivalent to 0.88 people per 1,000 people. No other estimates of the prevalence or incidence of acute cholecystitis were found in the literature search. However, estimates of the prevalence of gallstones, the primary cause of acute cholecystitis, were found. It has been estimated that up to 10% to 20% of residents of the United States have gallstones and that one-third of these patients will suffer from acute cholecystitis at some point in their lives.31 Return to Summary Table Criterion 2: Timeliness and urgency of test results in planning patient management (link to definition) Saskatchewan hospital guidelines indicate that cholescintigraphy for diagnosis of suspected acute cholecystitis should be conducted within 24 hours (Patrick Au, Acute and Emergency Services Branch, Saskatchewan Ministry of Health: unpublished data, 2011). Return to Summary Table Criterion 3: Impact of not performing a diagnostic imaging test on mortality related to the underlying condition (link to definition) If a test for diagnosing acute cholecystitis is not available, treatment might be delayed and complications with associated high mortality rates might be more likely to develop. Complications from acute cholecystitis occur in around 20% of patients. Complicated acute cholecystitis is associated with a mortality rate of around 25%.33 Perforation of the gallbladder, which occurs in 3% to 15% of patients with cholecystitis, has a 60% mortality rate.34 Acute acalculous cholecystitis has a mortality rate of around 30%.35 In an analysis of more than 29,000 elderly Medicare beneficiaries who presented with acute cholecystitis, those who were immediately treated with cholecystectomy had a lower mortality rate than patients not immediately treated with cholecystectomy.34,50 Patients given immediate cholecystectomy had mortality rates of 2.0%, 9.5%, and 15.2% at 30 days, one year, and two years, respectively. Patients not immediately treated with cholecystectomy had mortality rates of 5.0%, 19.4%, and 29.3% at 30 days, one year, and two years, respectively. Return to Summary Table Criterion 4: Impact of not performing a diagnostic imaging test on morbidity or quality of life related to the underlying condition (link to definition) If a test for diagnosing acute cholecystitis is not available, treatment might be delayed and patients may have to suffer symptoms of acute cholecystitis longer than necessary. Additionally, delayed treatment may make patients more susceptible to complications that could affect their survival or their quality of life. Two studies were identified that evaluated the quality of life impact of acute cholecystitis. A 2005 Norwegian study by Vetrhus et al.51 compared the quality of life over a five-year period of 64 patients presenting at their institution with acute cholecystitis. Patients were randomized to one of two treatment groups (all patients treated conservatively with antibiotics): observation, or cholecystectomy. Quality of life was assessed using the Psychological General Well-Being index (PGWB) and the Nottingham Health Profile (NHP) part II. Pain was evaluated using a pain score and a visual analogue pain scale (VAPS). No statistically significant differences between the two treatment groups over time were found in any of the instruments. However, the differences in mean scores in the quality of life and pain instruments at randomization and at five years reflect the morbidity impact of the acute cholecystitis episode. Table 2 presents selected findings of the study. Table 2: Selected Results Reported in Vetrhus et al. 200551
NHP = Nottingham Health Profile part II; PGWB = Psychological General Well-Being index; VAPS = visual analogue pain scale. Bass et al.52 estimated the quality of life impact of different types and treatments of gallbladder disease. After being presented with descriptions of different diseases and procedures, 40 subjects (without gallstones) provided preference scores by means of either a simple 0 to 100 rating scale (n = 22; score of 0 = immediate death and 100 = perfect health) or standard gamble (n = 18). The relative mean rating score — rated relative to other related conditions — for an episode of acute cholecystitis was 0.36 and 0.77 by standard gamble. Return to Summary Table Criterion 5: Relative impact on health disparities (link to definition) To be scored locally. Return to Summary Table Criterion 6: Relative acceptability of the test to patients (link to definition) Cholescintigraphy CT MRCP U/S Return to Summary Table Criterion 7: Relative diagnostic accuracy of the test (link to definition) One systematic review15 was identified that evaluated the diagnostic accuracy of cholescintigraphy with U/S in patients suspected to have acute cholecystitis. This review was somewhat dated (1994) and the majority of studies included were not head-to-head comparisons of cholescintigraphy and U/S. No systematic reviews were identified that compared cholescintigraphy with CT or MRCP. Therefore, a search for primary diagnostic accuracy studies comparing cholescintigraphy with any of the alternatives (U/S, CT, and MRCP) was conducted. Four primary studies were identifed that compared the diagnostic accuracy of cholescintigraphy to U/S.25-28 Three of the studies reported sensitivity, specificity, or both sensitivity and specificity of the diagnostic test.25,27,28 The other study reported only the correlation of findings between cholescintigraphy and U/S. No primary studies were found that compared cholescintigraphy with CT or MRCP. Cholescintigraphy versus U/S The three primary retrospective studies all found cholescintigraphy to have higher sensitivity than U/S for the diagnosis of acute cholecystitis. Chatziioannou et al.27 found the sensitivity of cholescintigraphy and U/S to be 0.88 and 0.50, respectively. Kalimi et al.28 reported the sensitivity of cholescintigraphy and U/S to be 0.86 and 0.48, respectively, while Alobaidi et al.25 reported the sensitivity of cholescintigraphy and U/S to be 0.91 and 0.62, respectively. In their study, Chatziioannou et al.27 found the specificity of cholescintigraphy to be 0.93 compared with 0.88 for U/S. In all three of these primary studies, findings from the imaging tests were compared with histopathological findings of the same patients suspected of acute cholecystitis. In Chatziioannou et al.,27 all 107 patients in the study underwent both cholescintigraphy and U/S. Tables 3 and 4 present other diagnostic findings from primary studies. Chatziioannou et al.27 found the overall accuracy of cholescintigraphy and U/S to be 0.92 and 0.77, respectively. Blaivas et al.26 found the correlation between the diagnosis of acute cholecystitis with cholescintigraphy and U/S to be 0.74. Table 3: Sensitivity and Specificity of Cholescintigraphy and Ultrasonography
CI = confidence interval; N = number of patients; NR = not reported. Table 4: Other Measures of Diagnostic Accuracy for Cholescintigraphy and Ultrasonography
Acc = accuracy; cor = correlation; N = number of patients; NR = not reported; NVP = negative predictive value; PPV = positive predictive value. Cholescintigraphy versus CT Cholescintigraphy versus MRCP Details of the diagnostic accuracy studies can be found in Appendix 3. Return to Summary Table Criterion 8: Relative risks associated with the test (link to definition) Non–radiation-related Risks Cholescintigraphy CT MRI U/S Radiation-related Risks Among the modalities to diagnose acute cholecystitis, cholescintigraphy, CT, and ERCP expose the patient to ionizing radiation. The average effective dose of radiation delivered with each of these procedures can be found in Table 5. Table 5: Effective Radiation Doses for Various Imaging Tests
CT = computed tomography; ERCP = endoscopic retrograde cholangiopancreatography; GI = gastrointestinal; MRCP = magnetic resonance cholangiopancreatography; NR = not reported; 99mTc-disofenin = technetium-99m disofenin; 99mTc-mebrofenin = technetium-99m mebrofenin; U/S = ultrasound. Return to Summary Table Criterion 9: Relative availability of personnel with expertise and experience required for the test (link to definition) Cholescintigraphy All alternative imaging modalities Service engineers are needed for system installation, calibration, and preventive maintenance of the imaging equipment at regularly scheduled intervals. The service engineer's qualification will be ensured by the corporation responsible for service and the manufacturer of the equipment used at the site. Qualified medical physicists (on-site or contracted part-time) should be available for the installation, testing, and ongoing quality control of CT scanners, MR scanners, and nuclear medicine equipment.56 CT MRCP U/S The availability of expertise to diagnose acute cholecystitis varies across the jurisdictions. Table 6 reports the number of medical imaging professionals nationally and highlights those provinces and territories that lack a specific expertise. Gastroenterologists are not included in this list; however, the number of gastroenterologists in Canada available to perform the procedure is reported to be 1.83 per 100,000 persons.57 Table 6: Medical Imaging Professionals in Canada47
AB = Alberta; BC = British Columbia; MB = Manitoba; NB = New Brunswick; NL = Newfoundland and Labrador; NR = not reported for jurisdictions; NS = Nova Scotia; NT= Northwest Territories; NU = Nunavut; ON = Ontario; PEI = Prince Edward Island; QC = Quebec; YT = Yukon. Return to Summary Table Criterion 10: Accessibility of alternative tests (equipment and wait times) (link to definition) There are notable variations in the availability of medical imaging technologies across Canada. Table 7 provides an overview of the availability of equipment required to diagnose acute cholecystitis. Data for nuclear medicine cameras (including SPECT) are current to January 1, 2007. The number of CT, MRI, and SPECT/CT scanners is current to January 1, 2010. Data were not available for U/S. Table 7: Diagnostic Imaging Equipment in Canada47,48
NT = Northwest Territories; NU = Nunavut; PE = Prince Edward Island; YT = Yukon Cholescintigraphy CT ERCP MRCP U/S Return to Summary Table Criterion 11: Relative cost of the test (link to definition) Fee codes from the Ontario Schedule of Benefits were used to estimate the relative costs of cholescintigraphy and its alternatives. Technical fees are intended to cover costs incurred by the hospital (i.e., radiopharmaceutical costs, medical/surgical supplies, and non-physician salaries). Maintenance fees are not billed to OHIP — estimates here were provided by St. Michael's Hospital in Toronto. Certain procedures (i.e., PET scan, CT scan, MRI scan) are paid for, in part, out of the hospital's global budget — these estimates were provided by The Ottawa Hospital. It is understood that the relative costs of imaging will vary from one institution to the next. According to our estimates (Table 8), the cost of cholescintigraphy with 99mTc-based radioisotopes is $298.38. CT is minimally more costly, MRCP is moderately more costly, and U/S is minimally less costly. An estimate for ERCP could not be obtained; however, actual costs (i.e., excluding professional fees) obtained from one Ontario hospital were reported to be approximately $1900. Therefore, ERCP is a significantly more costly alternative. Table 8: Cost Estimates Based on the Ontario Schedule of Benefits for Physician Services Under the Health Insurance Act (September 2011)58
3-D = three-dimensional; anes = anesthetic; CT = computed tomography; ERCP = endoscopic retrograde cholangiopancreatography; MRCP = magnetic resonance cholangiopancreatography; MRI = magnetic resonance imaging; RNA = radionuclide angiogram; spec = specialist; SPECT = single-photon emission computed tomography; U/S = ultrasound. Return to Summary Table References
Appendix 1: Multi-Criteria Decision Analysis Definitions
Appendix 2: Literature Search Strategy
The following sections of the CADTH grey literature checklist, "Grey matters: a practical search tool for evidence-based medicine" (http://www.cadth.ca/en/resources/grey-matters) were searched:
Appendix 3: Diagnostic Accuracy StudiesShea et al.15 Shea et al.15 conducted a systematic review and meta-analysis in order to estimate the diagnostic accuracy of various tests for biliary tract disease. Acute cholecystitis was one of the biliary diseases that were evaluated. The authors searched for studies published from 1966 to September 1992 on MEDLINE. Bibliographies of selected studies were also reviewed for relevant articles. The search identified articles with MeSH descriptors or either cholelithiasis or cholecystitis AND MeSH descriptors of any of cholecystography, ultrasonography, ultrasonics, tomography, nuclear magnetic resonance, or radionuclide imaging. Titles and abstracts of potential articles were screened by research staff. The full text of articles included after title and abstract screening was screened for inclusion by research staff and by a review committee if the research staff could not determine whether it should be included. Article exclusion criteria included the absence of original study data, sample size less than 20, inability to calculate sensitivity and specificity from data presented, atypical patient population, retrospective review, lack of description of criteria for positive diagnosis, whether the study used an atypical or outdated variant of a diagnostic test, the diagnosis was not confirmed with an acceptable gold standard, or more than 10% of patients were unavailable for follow-up. Sensitivity and specificity for the tests were pooled by what the authors describe as cluster sampling methods for estimating a proportion. The authors corrected their estimate to account for verification bias. The authors state that most diagnostic test studies suffer from verification bias because only a subset of patients have their diagnosis verified with a gold standard. In the case of patients with gallstones, more patients with a positive imaging test result are likely to have the most common gold standard, cholecystectomy. Twenty articles, with a total of 2,466 patients, were included in the cholescintigraphy diagnostic accuracy estimates. Five studies, with a total of 532 patients, were used to estimate diagnostic accuracy of ultrasound (U/S). The authors estimated sensitivity and specificity of cholescintigraphy to be 0.97 (confidence interval [CI], 0.96 to 0.99) and 0.90 (0.86 to 0.95), respectively. No verification bias adjustment was made for cholescintigraphy. The authors estimated verification bias adjusted sensitivity and specificity of U/S to be 0.88 (0.74 to 1.00) and 0.80 (0.62 to 0.98), respectively. The unadjusted sensitivity and specificity of U/S was estimated to be 0.94 (0.92 to 0.96) and 0.78 (0.61 to 0.96), respectively. Chatziioannou et al.27 Chatziioannou et al.27 compared the diagnostic accuracy of cholescintigraphy and U/S for the diagnosis of acute cholecystitis. One hundred and seven consecutive patients presenting to a United States hospital emergency department during 1996 suspected of acute cholecystitis were included in the study. Patients received both cholescintigraphy and U/S at the time of presentation. For patients who went on to surgery (n = 44), pathological findings were used as the gold standard with which imaging findings were compared. For patients who did not go on to surgery (n = 63), the diagnosis made by the primary physician was considered to be the gold standard with which results from imaging tests were compared. For cholescintigraphy, nonvisualization of the gallbladder either three to four hours after radiotracer injection or 30 minutes after radiotracer and morphine sulfate injection was considered consistent with acute cholecystitis. The primary finding from U/S that was considered consistent with acute cholecystitis was sonographic Murphy's sign. In the absence of Murphy's sign, other findings considered consistent with acute cholecystitis were gallstones and gallbladder wall thickness greater than 4 mm, gallstones, and a gallbladder more than 5 cm in length. Acalculous acute cholecystitis was diagnosed with findings of thickened gallbladder wall, edema within the wall, sludge pericholecystic fluid, and sonographic Murphy's sign. The authors presented results separately for all patients, and for patients who went on to surgery and had pathologic confirmation of presence or absence of acute cholecystitis. For all patients, the sensitivity, specificity, positive predictive value, negative predictive and overall accuracy of cholescintigraphy was estimated to be 0.88, 0.93, 0.85, 0.95, and 0.92. For U/S, the sensitivity, specificity, positive predictive value, negative predictive and overall accuracy was estimated to be 0.50, 0.88, 0.64, 0.80, and 0.77. For the 44 patients who went on to surgery, the sensitivity, specificity, positive predictive value, negative predictive and overall accuracy of cholescintigraphy was estimated to be 0.92, 0.89, 0.92, 0.89, and 0.91. For U/S, the sensitivity, specificity, positive predictive value, negative predictive and overall accuracy was estimated to be 0.40, 0.89, 0.83, 0.53, and 0.61. Alobaidi et al.25 Alobaidi et al.25 reviewed data from 117 patients pathologically proven to have acute cholecystitis. Patients were seen in a United States hospital between 1999 and 2002. Patients were stratified into groups depending on which imaging test (U/S, cholescintigraphy) or combination of imaging tests they underwent before surgery. The diagnoses made with each test at the time of exam were used to calculate each test. False-negative U/Ss were reviewed by radiologists, along with 40 true-positive scans from the same group as a control. The review was used to estimate a corrected sensitivity estimate to account for what the authors refer to as limiting factors relating to the date of surgery versus the date of imaging. Criteria used to diagnose acute cholecystitis with U/S included sonographic Murphy's sign, gallbladder wall thickening, pericholecystic fluid, biliary dilatation, and gallbladder hydrops. Diagnosis of acute cholecystitis with cholescintigraphy was based on nonvisualization of the gallbladder three hours after injection of radiotracer or 30 minutes after injection of morphine sulfate. Ninety-seven of the 117 patients had U/S as their initial imaging test. Based on initial diagnosis, the authors reported sensitivity for U/S of 62%. Nine false-negative patients reclassified as true positives upon additional review by radiologists. Based on this reclassification, the sensitivity of U/S was estimated to be 70.4%. The authors estimated the sensitivity of cholescintigraphy to be 90.9%. Kalimi et al.28 Kalimi et al.28 retrospectively reviewed 132 patients admitted to a United States hospital emergency room with upper quadrant pain between 1996 and 2000. These patients were pathologically proven to have acute cholecystitis. At the time of presentation at the emergency room, patients were tested either by means of cholescintigraphy (n = 28), U/S (n = 28), or both cholescintigraphy and U/S (n = 54). Sensitivity for each test or combination of tests was estimated by the number of positive findings at the time of admission. Cholescintigraphy was considered to be positive for acute cholecystitis if there was nonvisualization of the gallbladder despite morphine augmentation. U/S findings considered positive for acute cholecystitis was presence of gallstones along with either wall edema or stone impacted in the gallbladder neck. The authors report the sensitivity of cholescintigraphy and U/S to be 86% (95% CI, 67% to 96%) and 48% (95% CI, 34% to 63%), respectively. The sensitivity for patients undergoing both cholescintigraphy and U/S was 90% (95% CI, 80% to 97%). Blaivas et al.26 Blaivas et al.26 retrospectively compared findings from U/S and cholescintigraphy in patients suspected of acute cholecystitis. A total of 99 patients presenting at a United States hospital emergency department who received both U/S and cholescintigraphy were included in the study. U/S findings that were considered to be indicative of acute cholecystitis included finding of gallstones with a sonographic Murphy's sign, significant wall thickening greater than 5 mm, pericholecystic fluids, or a combination of these. No gold standard was specified in the study and no estimates of sensitivity or specificity were reported. The authors did report an overall correlation of findings between cholescintigraphy and U/S of 0.74. The authors also reported that U/S diagnosed acute cholecystitis in 38% of the 38 patients diagnosed with cholescintigraphy. Of the 25 patients diagnosed with acute cholecystitis using U/S, 80% were diagnosed positive using cholescintigraphy. |