An official website of the United States government
NCBI Bookshelf. A service of the National Library of Medicine, National Institutes of Health.
Gallstone Disease: Diagnosis and Management of Cholelithiasis, Cholecystitis and Choledocholithiasis. London: National Institute for Health and Care Excellence (NICE); 2014 Oct. (NICE Clinical Guidelines, No. 188.)
What signs, symptoms, and risk factors should prompt a clinician to suspect symptomatic gallstone disease in adults presenting to healthcare services?
The aim of this question was to identify the specific signs, symptoms, and risk factors that can predict gallstone disease in adults who present at healthcare services. This question did not aim to identify signs, symptoms and risk factors for gallstone disease in the general population. This is because the majority of people with gallstone disease in the general population are asymptomatic, and the potential signs, symptoms and risk factors identified at a population level may be different to the signs, symptoms and risk factors that cause people to seek medical attention.
A systematic search was conducted (see appendix D.1), which identified 7802 references. After removing duplicates the references were screened by their titles and abstracts. This led to 74 references being obtained and reviewed against the inclusion and exclusion criteria as described in the review protocol (appendix C.1).
Primary research of any study design was eligible for inclusion if it satisfied the following criteria:
Overall, 73 studies were excluded as they did not meet the eligibility criteria. A list of excluded studies and reasons for their exclusion is provided in appendix F.
One study met the eligibility criteria and was included. Data were extracted into detailed evidence tables (see appendix G.1) and are summarised in Table 1 below.
Summary of included studies for review question 1.
The GRADE framework was modified for this review. As prospective studies were considered to be the highest quality evidence, these were rated initially as high quality while retrospective studies were downgraded to start as low quality. The evidence for the outcomes was then assessed in the normal GRADE framework by downgrading or upgrading on the basis of inconsistency, imprecision and indirectness. The modified GRADE profiles are in appendix I.1. After applying the modified GRADE framework the evidence was judged to be very low in quality. Full GRADE profiles are in appendix I.1
A literature search was conducted jointly for review questions 1 and 2 by applying standard health economic filters to the clinical search strategies (see Appendix D). For review questions 1 and 2, 914 references were retrieved, of which 16 were retained after title and abstract screening. No health economic studies were found for question 1. Health economic modelling was not prioritised for this review question.
Very low quality evidence from 1 retrospective cohort study did not identify any factors that could predict gallstone disease in those presenting with upper abdominal pain lasting less than one week.
| Relative value of different outcomes | There was insufficient evidence relating to any outcome from this review question as the single included study only included people presenting with upper abdominal pain lasting less than one week. |
|---|---|
| Trade off between benefits and harms | Some people present to healthcare services with typical signs, symptoms and risk factors for gallstone disease (such as sudden severe abdominal pain, nausea, jaundice etc.) as in the included study. The GDG agreed that most medical professionals would suspect gallstone disease in people presenting with these typical symptoms, as these features are well recognised and medical students are routinely educated about them. There was insufficient evidence available to support or refute these typical signs, symptoms and risk factors as the single study identified only included people presenting with upper abdominal pain lasting less than a week. The GDG considered putting typical signs and symptoms in a recommendation based on their combined knowledge and experience but felt that that may become a barrier to people who present with atypical signs and symptoms. The GDG also acknowledged that some people with gallstone disease present with symptoms that can be vague and easily misattributed to other conditions (such as indigestion or general abdominal discomfort) by both the patient and their healthcare professional. This can mean that gallstone disease is not immediately considered, and patients may be investigated or treated for a condition that they do not have. This ultimately affects patient quality of life, and the use of NHS resources as patients will continue to have unresolved symptoms that may get worse, resulting in inappropriate treatments and investigations being offered to the patient. However, there was insufficient evidence available to enable this group of patients to be identified. Thus the GDG decided not to make a recommendation based on typical signs and symptoms or risk factors that should prompt a clinician to suspect symptomatic gallstone disease in adults presenting to healthcare services. |
| Consideration of Health Benefits and Resource Use | No health economic evidence was found. |
| Quality of evidence | The single study available for this review was of very low quality, and insufficient evidence was provided to support decision making. |
| Other considerations | The GDG acknowledged the lack of research that was available around identifying signs, symptoms and risk factors for gallstone disease, but felt that good quality research in this area would have limited value as the benefits would be small. The identification of specific signs, symptoms and risk factors would ultimately refine the number of people who were offered ultrasound and liver function tests (these tests were reviewed and recommended in section 4.2 of this guideline). Since these tests are relatively low cost and easy to perform, are low risk and minimally invasive to patients, and are widely used for a range of conditions, refining the number of these tests that are performed would not lead to major cost savings or improvements in the quality of NHS care. Therefore, the GDG did not feel that a research recommendation would be useful. |
No recommendations were made in relation to this review question.
No research recommendations were made in relation to this review question.
What is the most accurate strategy for diagnosing gallstone disease in adults suspected of the condition?
The aim of this question was to assess all available methods for diagnosing gallstone disease and establish which methods are the most accurate.
A systematic search was conducted (appendix D.2), which identified 6312 references. After removing duplicates and screening the references based on their titles and abstracts, 310 references were obtained. These were reviewed against the inclusion and exclusion criteria as detailed in the review protocols (appendix C.2).
Primary research utilising a randomised controlled trial, cohort, or cross sectional design was eligible for inclusion. Studies also had to have utilised a valid reference standard. Currently there is no accepted reference standard for confirming the presence or absence of gallstone disease, so studies were only included if they met the following criteria:
These procedures can accurately confirm the presence of gallstone disease by extracting the gallstone(s). Other tests such as endoscopic ultrasound, magnetic resonance cholangiopancreatographydo do not extract gallstones, so were considered unsuitable as reference standards.
It is much more difficult to confirm the absence of gallstone disease. Gallstones or common bile duct stones may be missed during endoscopic or surgical investigations. Patients can be followed up to establish if signs and symptoms persist, which can indicate that gallstone disease is present but was missed during previous investigations. However, this can be misleading, as the patient's signs and symptoms can be caused by other conditions, or new gallstone disease could have developed since the original investigations. As there is no alternative method for definitively confirming the absence of gallstone disease, surgery and ERCP were accepted as the best available reference standards, and their potential inaccuracies are acknowledged.
During the review, a date restriction was also imposed on all studies that utilised endoscopic, surgical, or radiological methodologies. This was because technological advances have made older studies of limited relevance to clinical practice today. An arbitrary publication of date of 1993 or later was used as it coincides with the approximate introduction of laparoscopic cholecystectomy into clinical practice in the UK. The only exceptions to this date restriction were studies that focused on predicting the presence of gallstone disease using clinical history taking, physical examination and simple blood tests, as these factors are not as dependent on technology, and studies conducted over 20 years ago are likely to still be relevant to clinical practice today.
Overall, 23 studies met the eligibility criteria and were included in the review. Evidence was extracted into detailed evidence tables (see appendix G.2). Diagnostic test accuracy data were provided by 20 studies, and where possible these data were pooled in relevant meta-analyses. Some of these studies compared to see whether diagnostic test accuracy differed depending on who interpreted the test results (for example, radiologist compared with ultrasonographer, or experienced radiologist compared with inexperienced radiologist). This produced different results for the same sets of patients. It would be inappropriate to use both results in the meta-analysis as this would be double counting. Instead the results reported by the interpreter most similar to those intended to use the test in clinical practice were included in the analysis. If both interpreters were intended to use the test, their test results were averaged and this was taken into the meta-analysis.
Data about predictive factors for gallstone disease were provided by 3 studies. It was not possible to pool the data from these studies, as each study investigated different predictive factors.
Data from the included studies were extracted into detailed evidence Table 2 and Table 3 below, and the GRADE framework for diagnostic evidence was used to quality assess the evidence. However, for this review, the GDG took a liberal approach to set the threshold for accuracy of 0.50 for both sensitivity and specificity on the basis that they wanted to identify the test(s) that were better than chance. Any test that did not meet this threshold was not considered clinically useful. Full GRADE profiles are presented in Appendix I.2.
Summary of included studies reporting diagnostic test accuracy.
Studies reporting prognostic data.
A literature search was conducted jointly for review questions 1 and 2 by applying standard health economic filters to the clinical search strategies (see Appendix D). Searches for review questions 1 and 2 retrieved 914 studies, of which 16 were retained after title and abstract screening. For review question 2, 2 health economic studies were found (Howard et al. 2006; Kaltenthaler et al. 2004). An assessment of the quality of the economic evidence is given in Table 4. This review question was not prioritised for original health economic modelling.
Economic Evidence – MRCP versus ERCP for detection of CBDS.
Howard et al (2006) used a decision tree analysis to compare MRCP with ERCP for diagnosing post-cholecystectomy CBDS in patients with abdominal pain and/or abnormal liver function tests. Kaltenhaler et al. (2004) used a decision tree analysis to compare MRCP with diagnostic ERCP for patients with suspected CBDS. The GDG felt that, as it was based on UK data and covered a wider patient population, Kaltenhaler et al (2004) provided slightly higher quality evidence. However, both studies were felt to be of use. Both studies found that MRCP dominated ERCP, but this depended on the prior probability of CBDS. In probabilistic sensitivity analysis, Howard et al. (2006) was found to be 83% cost effective (at Australian $50,000/QALY threshold) and Kaltenhaler et al. (2004) was found to be 100% cost effective (at UK £20,000/QALY threshold). Indeed, Kaltenhaler et al. (2004) was found to be 97% cost saving.
MRCP appeared to be cost effective compared with ERCP for diagnosing CBDS under the majority of assumptions. The GDG considered that the results from the 2 studies reflected what they would expect to occur in standard clinical practice.
While most of the diagnostic tests had sensitivities and specificities above 50%, some had confidence intervals below 50% for either sensitivity or specificity. This means that there is uncertainty in the evidence as some of the tests may actually perform worse than chance alone. The quality of this evidence was mainly moderate to low in quality.
Two partially applicable health economic studies with minor limitations found that, under a variety of assumptions, MRCP appeared cost effective compared with ERCP for diagnosing CBDS.
| Relative value of different outcomes | The various metrics for estimating diagnostic test accuracy were discussed by the GDG, and they agreed to base decisions on sensitivity and specificity. This is because other diagnostic test accuracy measures are less reliable (for example, positive and negative predictive values), or cannot be meta-analysed (for example, likelihood ratios, as robust methods for doing this have not yet been developed). The GDG discussed patient benefits, patient harms, and clinical utilities associated with diagnostic testing, and the minimum sensitivity and specificity values that it would tolerate. It agreed that sensitivity and specificity estimates at or below 0.50 (or with confidence intervals below 0.50) were not clinically useful as these results indicate that the test could perform the same or worse than chance alone. The GDG agreed that sensitivity and specificity were equally as important in all diagnostic tests, except for ultrasound and liver function tests where the GDG would accept lower sensitivity values. The GDG recognised that patients with false negative results at this early stage of the clinical pathway would be offered subsequent investigations as part of their differential diagnosis. Therefore they could tolerate lower levels of sensitivity, and so a higher number of false negative results, as patients would present to health services with recurrent symptoms at a later time point. The GDG felt that morbidity and mortality related to false negatives was lower at this early stage of investigations, than for people with false positives who would undergo unnecessary treatment. The GDG agreed that in practice other tests are only offered after ultrasound and liver function tests have been performed. Because of this, the GDG felt that for other tests sensitivity and specificity were equally as important, as false positive and false negative results are associated with similar levels of subsequent morbidity and mortality at this stage of the pathway. |
|---|---|
| Trade off between benefits and harms | Abdominal ultrasound For gallbladder stones: A single study provided very low quality evidence about ultrasound for diagnosing gallbladder stones. The study showed that ultrasound was extremely useful for identifying the presence of gallstones (high sensitivity) but was not at all useful for ruling out gallstones in people who don't have them (low specificity). The GDG felt that these results did not match their own experiences of using the test to diagnose gallstones, and they felt that the test is both sensitive and specific. Since the evidence was based on a single study and contradicted clinical experience, the GDG agreed that the results were not reliable. For cholecystitis: Three studies provided very low quality evidence about ultrasound for diagnosing cholecystitis. Point estimates for sensitivity and specificity were satisfactorily high, but confidence intervals fell below 0.50 for sensitivity. The GDG felt that it could accept lower sensitivity values for this test, as it is very low risk and is used routinely in the NHS for a range of conditions. In addition, the GDG noted that the studies included in this meta-analysis had very small populations, so the wide confidence intervals may be because of a lack of statistical power rather than a lack of sensitivity. For common bile duct stones: Five studies provided low quality evidence about ultrasound for diagnosing common bile duct stones. Sensitivity and specificity values were satisfactorily high and supported the GDG's own knowledge and experience of the test. The GDG were confident that ultrasound was clinically useful in diagnosing common bile duct stones. Overall: the GDG felt there was sufficient evidence to recommend ultrasound for diagnosing gallstone disease since it is accurate, widely available throughout the NHS and is of very low risk to the patient. Cholescintigraphy (HIDA scan) No evidence was available. The GDG agreed that this test is not routinely used for diagnosing gallstone disease, and that it is usually used for people for whom other tests are unsuitable. The GDG agreed not to make a recommendation based on the knowledge and experience of the group. The test exposes patients to ionising radiation and the associated risks, and other tests considered in this review have evidence supporting their use for diagnosing gallbladder stones. Computer Tomography (CT) For cholecystitis: A single, small study provided low quality evidence about CT for diagnosing cholecystitis. The study had unreliable estimates of specificity, as the confidence interval was below 0.50. The GDG agreed that there was insufficient evidence and decided not to make a recommendation based on the knowledge and experience of the group. They felt that other tests considered in this review are better at diagnosing cholecystitis. For common bile duct stones: Three studies provided moderate quality evidence about CT for diagnosing common bile duct stones. The meta-analysis demonstrated that CT had satisfactory sensitivity and specificity estimates. Overall: CT uses ionising radiation, and the sensitivity and specificity were found to be inferior to other tests, such as MRCP. Because of this, the GDG felt that CT should not be recommended for diagnosing gallstone disease, as other tests are more sensitive, specific, and safer. CT cholangiography (CTC) For common bile duct stones: Evidence about CTC for common bile duct stones was provided by 4 studies. These showed that sensitivity and specificity were satisfactorily high, with reliable confidence intervals. However, the GDG decided not to recommend this test because the contrast agent needed to perform CTC (iotroxic acid) is not available in the UK. This test also exposes patients to ionising radiation, and the GDG felt that there were more accurate and safer tests for diagnosing common bile duct stones. Endoscopic Ultrasound (EUS) For common bile duct stones: Three studies provided low quality evidence about endoscopic ultrasound for diagnosing common bile duct stones. The meta-analysis demonstrated that endoscopic ultrasound had very high sensitivity and specificity estimates. However, the confidence intervals for specificity fell below 0.50, indicating that there is uncertainty about whether the test is clinically useful. The GDG felt that these results did not reflect clinical reality, as EUS is becoming widely accepted as being an accurate test for diagnosing common bile duct stones. However, when ERCP is used as a reference standard against which to judge the accuracy of EUS, cases that are classified as false positives on EUS might actually be revealed as false negatives on ERCP, were a perfect reference standard with which to judge both tests available. Therefore, the GDG believed that such analyses may underestimate the accuracy of EUS. . Liver function tests For common bile duct stones: Three studies that could not be pooled examined predictors of common bile duct stones and included liver function tests among the variables that were examined. These studies examined different combinations of risk factors and so their results could not be meta-analysed. One study provided moderate quality evidence, and the other 2 provided very low quality evidence. All 3 studies found that liver function tests were clinically useful in predicting common bile duct stones. The GDG agreed that there was sufficient evidence to recommend the use of liver function tests for identifying common bile duct stones, as supportive evidence was provided by all 3 studies. The GDG based its decisions on the results of Shiozawa (2005), based on the moderate quality rating, large sample size, narrow confidence intervals and clinically useful sensitivity and specificity estimates. Like ultrasound, the GDG felt that liver function tests could be recommended because the test is widely available throughout the NHS and is of very low risk to the patient. Magnetic Resonance Cholangiopancreatography (MRCP) For cholecystitis: Two studies provided low quality evidence about MRCP for diagnosing cholecystitis and demonstrated satisfactorily high estimates of sensitivity and specificity. Only one of these studies considered the presence of gallstones when diagnosing cholecystitis. The GDG felt that healthcare professionals would consider the presence of gallstones when making their diagnosis, and if this had been done in both studies included in the evidence, then diagnostic test accuracy would be more certain. The GDG felt that they could explain some of the uncertainty in the sensitivity and specificity estimates for MRCP, and so could be confident that this test is clinically useful. For common bile duct stones: Eight studies provided moderate quality evidence about MRCP for diagnosing common bile duct stones. The meta-analysis demonstrated sufficiently robust sensitivity and specificity values, so the GDG was confident that MRCP was also clinically useful and should be recommended. Magnetic Resonance Imaging- contrast enhanced (MRI- contrast enhanced) For cholecystitis: The GDG agreed there was insufficient evidence to recommend MRI for cholecystitis, as evidence came from a single small study using contrast enhancement, which is not normally done in clinical practice today. Point estimates were satisfactorily high, but confidence intervals for specificity fell below 0.50, so the GDG were unsure if the test was useful. The GDG agreed not to make a recommendation about MRI based on the knowledge and experience of the group, as it felt that other tests considered in this review are better tests for diagnosing cholecystitis and conditions associated with gallstone disease. Summary of all diagnostic methods After reviewing and discussing the evidence, the GDG agreed that ultrasound, liver function tests, EUS and MRCP could be recommended for diagnosing gallstone disease. The GDG then discussed the sequencing of diagnostic tests and based these discussions on the group's knowledge and experience. The evidence review in section 4.1 of this guideline discussed the 2 main ways that patients with gallstone disease present to health care services: either with typical signs, symptoms, and risk factors for gallstone disease, or with vague abdominal or gastrointestinal symptoms that can easily be misattributed to other conditions. Because of this, the group felt that ultrasound and liver function tests should be recommended as first line investigations for patients with suspected gallstone disease. In addition, these tests were recommended for patients with abdominal or gastrointestinal symptoms that are not responding to treatment as expected, since this group of patients may have ‘unsuspected’ gallstone disease. This was felt to be of benefit to both patients and the NHS as these tests are clinically useful, relatively cheap and widely available. Improving the identification of gallstone disease in people without typical symptoms would ultimately enable patients to progress to more appropriate investigations or treatment pathways. The group then discussed if MRCP and EUS should both be offered as second line investigations. MRCP is widely available throughout the NHS, is relatively quick to perform on an outpatient basis, and requires no sedation. It is less effective at diagnosing common bile duct stones than EUS and is not an appropriate test for patients with metal implants or claustrophobia. The GDG agreed that EUS is a more accurate test than MRCP, but EUS is an invasive test and procedural errors could result in adverse effects for the patient (such as perforation and even death). The test is normally performed with the patient under sedation, which carries its own risks to the patient, and requires the patient to be accompanied by a responsible adult for 24 hours after sedation. People who do not have a responsible adult to escort them will need to be admitted to hospital. Furthermore, EUS is not as widely available throughout the NHS as other tests (such as MRCP) are. Although the GDG agreed that EUS was probably superior at diagnosing gallstone disease than MRCP when performed by highly skilled operators, there was acknowledgement that not all operators would be highly skilled. Although skilled operators should also perform MRCP, the GDG felt that there is less potential for harm if less skilled operators performed the test. This is because MRCP is non-invasive, and the entire diagnostic process is recorded. This means that poor quality or equivocal tests can be reviewed by another operator, who may be able to offer a diagnosis without having to investigate the patient again. EUS only records key images from the test, so a second reviewer would not be able to check unclear or poor quality tests and in these cases patients would need to have another test performed. The GDG agreed that in most cases MRCP would be safer and more acceptable to patients than EUS and recommended MRCP as a second line investigation and EUS as a third line investigation. However, the GDG wanted to stress that this should not preclude appropriately trained and experienced operators using EUS instead of MRCP if appropriate. |
| Consideration of Health Benefits and Resource Use | Limited published health economic evidence suggested that MRCP appears cost effective compared with ERCP for diagnosing CBDS. In most situations MRCP appeared more effective and less costly than ERCP. Despite the limitations of the evidence presented, the GDG felt that it supported the primary use of MRCP for diagnosing CBDS. The GDG noted that there was no health economic comparison available for many of the other diagnostic interventions under consideration. |
| Quality of evidence | Overall, the evidence was moderate to very low in quality. For gallbladder stones, only low quality evidence on ultrasound was found. For cholecystitis, low quality evidence on MRCP and CT and very low quality evidence on ultrasound and MRI was found. For common bile duct stones, moderate quality evidence was found on MRCP, CT, CT cholangiography, and for a model predicting common bile duct stones using liver function tests and other predictors. Low quality evidence was found for ultrasound and endoscopic ultrasound, and very low quality evidence was found for 2 models predicting common bile duct stones using liver function tests and other predictors. |
| Other considerations | None |
MRCP and EUS have both been found to be sufficiently accurate for diagnosing common bile duct stones, with EUS regarded as the most accurate test. MRCP is non-invasive and so carries negligible risks to the patient. However, EUS carries a small but significant risk of patient harms, including death. There is insufficient evidence available to determine whether the benefits of improved diagnosis associated with EUS outweigh its procedural risks. Therefore, research is needed to compare MRCP with EUS to evaluate the subsequent management of common bile duct stones.
What factors predict which patients with asymptomatic gallbladder stones will develop acute complications?
The aim of this review question was to establish whether some people with asymptomatic gallbladder stones are at a higher risk of developing complications than others.
A systematic search (appendix D.3) retrieved 12,256 references. After removing duplicates and screening the references based on their titles and abstracts, 56 references were obtained and reviewed against the inclusion and exclusion criteria as described in the review protocol (appendix C.3).
Primary research using a randomised controlled trial, cohort or cross sectional design was eligible for inclusion. Two main types of study were expected to be found by the search: prospective studies recruiting a sample of the general population to identify and follow up people with asymptomatic gallbladder stones, and retrospective reviews of people presenting with symptomatic gallstone disease whose prior medical history was examined to identify whether asymptomatic gallbladder stones had previously been diagnosed. Both types of study were eligible for inclusion.
From the review, 55 studies were excluded, mainly because they provided prevalence and/or incidence data only. A list of the excluded studies and reasons for their exclusion is provided in appendix F.
One prospective cohort study met the eligibility criteria and was included. Data were extracted into detailed evidence tables (see appendix G.3) and are summarised in Table 5 below.
Summary of included studies for question 3.
Appropriate methodology checklists were used to appraise the methodological quality of individual studies, and a modified version of the GRADE framework was applied to summarise the overall quality of the evidence (see appendix I.3). In this approach the prospective cohort study was started with a ‘high’ quality rating and was further downgraded as appropriate, according to standard GRADE framework. Overall this evidence was rated as very low in quality.
A literature search was conducted for review question 3 by applying standard health economic filters to the clinical search strategies (see Appendix D). From the search, 1004 references were retrieved, of which 9 were retained after title and abstract screening. No health economic studies were found for question 3. Health economic modelling was not prioritised for this review question.
Very low quality evidence from a single study examined predictors of symptomatic gallstone disease in patients who were asymptomatic at the time the trial was conducted. Insufficient data were reported in the study to validate the author's findings that the following variables are not significant predictors of the development of biliary colic, other complications, or death:
Occurrence of biliary colic was not found to be a predictor of additional complications or death.
| Relative value of different outcomes | The evidence was based on the results of a single study. Only descriptive data and significance levels were provided, so secondary analysis for the purposes of this review could not be performed. As no meaningful results were provided by the evidence, the GDG based its decisions on the knowledge and experience of the group, and alongside evidence presented in section 4.4 of this guideline. |
|---|---|
| Trade off between benefits and harms | For people with asymptomatic gallbladder stones in a normal biliary tree, the GDG agreed that there is currently no way of predicting which people will go on to develop complications. Based on their experience, the GDG agreed that most people with asymptomatic gallbladder stones will not develop complications. However, once a patient has been informed that they have asymptomatic gallbladder stones, they may worry about their risk of developing complications in future, and may ask about having treatment prophylactically. Because of this, the discussions are linked to the evidence about managing asymptomatic gallbladder stones, which is reviewed and has recommendations on asymptomatic gallbladder stones made in section 4.4. After considering the evidence (see ‘Evidence to recommendations’, section 4.4.5), the GDG felt that the prophylactic treatment of asymptomatic gallstones is likely to have a higher risk than leaving them untreated as all of the treatment options carry risks of adverse effects (such as side effects of medication and surgery risks,). Therefore the GDG recommended that healthcare professionals should reassure patients that asymptomatic gallstones do not need treatment unless they develop symptoms (see section 4.4.6 of this guideline). The GDG felt that some people with asymptomatic gallbladder stones will have other abnormalities detected, either in their biliary tree (such as calcification of the gallbladder) or elsewhere in their body at the time of gallbladder stone detection. For this group of people, the GDG agreed that there is a need to perform further tests and investigations to rule in or rule out other diseases and conditions, such as cancer. It is not within the scope of this guideline to describe or review the evidence for the range of other conditions that should be investigated, so the GDG felt that they could not make recommendations about what a healthcare professional should do when other abnormalities coexist with asymptomatic gallstones. |
| Consideration of Health Benefits and Resource Use | No health economic evidence was found. |
| Quality of evidence | A single study providing very low quality evidence was not adequately powered to predict acute complications. |
| Other considerations | The GDG acknowledged that definitions of asymptomatic gallbladder stones may vary and so the following definitions were developed:
|
No specific recommendations were made in relation to this question.
Which strategies should be used for managing asymptomatic gallbladder stones?
This question aimed to establish if prophylactic treatment should be offered to people with asymptomatic gallbladder stones, to prevent them from developing symptoms in the future.
A single search was performed for questions 4a, 4b, 4c and 5, which identified 10,976 references. After removing duplicates and screening the references based on their titles and abstracts, 210 references were obtained and reviewed against the inclusion and exclusion criteria for this review question (appendix C)., and 47 references met the overall inclusion criteria. Details of excluded studies and reasons for their exclusion are in appendix F.4.
None of the 47 included references met the criteria for this specific review question (see study flow chart, appendix E.4).
A literature search was conducted jointly for questions 4 and 5 by applying standard health economic filters to the clinical search strategies (see Appendix D). From the literature search, 1,396 references were retrieved for questions 4 and 5, of which none were retained for question 4a. Health economic modelling was not prioritised for this review question.
No evidence that met the inclusion and exclusion criteria for this question was found.
| Relative value of different outcomes | No evidence was available for this review question. The GDG based its decisions on the knowledge and experience of the group. |
|---|---|
| Trade-off between benefits and harms | The GDG discussed the potential benefits and harms of the various treatment options that were considered, using its knowledge and experience as no evidence was available:
The GDG also felt that it was important to discuss asymptomatic common bile duct stones, which were reviewed as part of question 4c (see section 4.6). No evidence about asymptomatic common bile duct stones was found, but the GDG felt that they were able to make recommendations based on their knowledge and experience. Asymptomatic common bile duct stones are different to asymptomatic gallbladder stones.
Recommendations about managing asymptomatic common bile duct stones are in section 4.6. |
| Consideration of Health Benefits and Resource Use | No health economic evidence was found. |
| Quality of evidence | No evidence was available. |
| Other considerations | The discussions around asymptomatic gallbladder stones that were linked to discussions around asymptomatic common bile duct stones were reviewed as part of review question 4c in section 4.6. The GDG discussed whether it would be useful to make a research recommendation about treating asymptomatic gallbladder stones, since there was no evidence to base their decisions on. The group agreed that an appropriate trial would be one that screened a large population to identify sufficient people with asymptomatic gallbladder stones, who would then need to be followed up for many years. The GDG felt that ideally, people included in such a trial should not be told that they have asymptomatic gallbladder stones identified, as it was felt that this knowledge could confound the findings of a trial. This is because once people are aware they have asymptomatic gallbladder stones they may report symptoms they would not have otherwise have noticed, or seek medical attention for symptoms that they would not have normally sought attention for. Since it would not be ethical to withhold information from patients, the GDG agreed that in reality there is no appropriate way to conduct primary research in this area. Therefore it was decided that a research recommendation should not be made. |
Reassure people with asymptomatic gallbladder stones found in a normal gallbladder and normal biliary tree that they do not need treatment unless they develop symptoms.
No research recommendations were made in relation to this review question.
No references
Which strategies should be used for managing symptomatic gallbladder stones?
This question aimed to establish which management strategies offer optimal outcomes for patients with symptomatic gallbladder stones, including patients with acute cholecystitis.
A single search was performed for review questions 4a, 4b, 4c and 5, which identified 10,976 references. After removing duplicates and screening the references based on their titles and abstracts, 210 references were obtained and reviewed against the inclusion and exclusion criteria for this review question (appendix C), and 47 references met the overall inclusion criteria. Details of excluded studies and reasons for their exclusion are in appendix F.4.
Of the 47 included references, 15 references relating to 9 randomised controlled trials met the inclusion criteria for this review question (see study flow chart in appendix E.4), which specifically focussed on comparisons of the following strategies for managing symptomatic gallbladder stones:
Data from the 9 included randomised controlled trials were extracted into detailed evidence tables (see appendix G) and are summarised in Table 6 below. The included studies were critically appraised using randomised controlled trial methodology checklists. Data were then analysed using meta-analysis wherever appropriate (see appendix H.5 for all data analysis outputs).
Summary of included studies for question 4b.
Outcomes were assessed using the standard GRADE approach (see appendix I.5 for full GRADE profiles). In this approach, randomised controlled trials are started with a ‘high’ quality rating and are further downgraded as appropriate. Overall the evidence was moderate to low in quality.
A literature search was conducted jointly for questions 4 and 5 by applying standard health economic filters to the clinical search strategies (see Appendix D). From the search, 1,396 references were retrieved for questions 4 and 5, of which 1 was retained for question 4b and 1 was retained for question 5. As no existing health economic studies were found that addressed all the comparisons in questions 4 and 5, an original economic model was constructed.
A full description of the health economic model can be found in in Appendix J; a summary is presented here. The model was developed in line with the NICE reference case (National Institute for Health and Care Excellence 2013). A single health economic model structure was developed to address all prioritised comparisons under questions 4b, 4c and 5. For question 4b, no evidence was found to enable the modelling of the percutaneous cholecystostomy comparison.
A single Markov structure with 2-week cycles was used to assess all comparisons (see Figure 1). Not all states and transitions were used in each comparison (see Appendix J for descriptions of the states and transitions available for each comparison). A lifetime horizon was adopted to capture the long term impact of bile duct injuries and mortality differences. The model is a natural history rather than a diagnosis model, so the model “knows” whether a patient has CBDS, irrespective of whether the patient or clinician knows.
Surgical interventions (laparoscopic cholecystectomy and ERCP, or both together) are represented as 2-week states, with short term surgical consequences (including mortality) modelled as cost and QALY impacts within one 2-week state. Laparoscopic cholecystectomy can cause bile duct injury.
Symptoms modelled are mutually exclusive and specific to gallbladder stones (biliary colic, acute cholecystitis) or CBDS (cholangitis, jaundice, pancreatitis [with associated mortality risk], sepsis).
Where possible, model parameters (symptoms at baseline or during treatment, operative consequences as prioritised by the GDG) are sourced from the included clinical studies, with data from standard laparoscopic cholecystectomy or ERCP trial arms combined to increase parameter accuracy.
Costs are based on 2011–12 NHS reference costs (Department of Health 2012). As the GDG identified length of stay as a critical outcome, resource use was based on length of stay differences reported in the included clinical studies.
There are few EQ-5D-based utility values for gallstones, so SF36 data were converted to utility values (Ara and Brazier 2008). The quality of life impact of interventional procedures reflects length of stay, but is assumed to return to normal within 2 weeks, unless a bile duct injury occurred (assumed to have lifelong impact). All symptoms are assumed to have the same utility impact. The utility impact of living with asymptomatic gallbladder stones were taken from a study of urinary stones (Penniston and Nakada 2007).
The health economic model has a number of limitations that should be considered. A number of key parameters are based on the included clinical evidence that has wide confidence intervals; zero event rates in 1 arm are common. A lack of gallstone-specific utility values required a number of assumptions to be made regarding the QALY outcomes. Interventions are costed using average costs – micro-costing of interventions may help identify differences between options, but may limit the generalisability of the model.
The health economic model found that conservative management is more costly and produces fewer QALYs than laparoscopic cholecystectomy and is therefore said to be dominated (see Table 7). This remained true in 100% of probabilistic model runs. It appears that not removing the gallbladder increases the need for and exposure to further ERCPs and also gallbladder cancer.
Cost effectiveness results for laparoscopic cholecystectomy versus conservative management.
The health economic model found that, on average, laparoscopic cholecystectomy with routine intraoperative cholangiography is more costly and produces fewer QALYs than laparoscopic cholecystectomy alone and is therefore said to be dominated (see Table 8). This result is driven by the increased costs of higher laparoscopic to open cholecystectomy conversion rates and uncertain rates of bile duct injury and bile leak in the intraoperative cholangiography arm - neither of the latter were statistically significant. In probabilistic sensitivity analysis, laparoscopic cholecystectomy alone has a 67% chance of being cost effective at a threshold of £20,000 per QALY compared with laparoscopic cholecystectomy with intraoperative cholangiography.
Cost effectiveness results for laparoscopic cholecystectomy with intraoperative cholangiography versus laparoscopic cholangiography alone.
Inpatient laparoscopic cholecystectomy costs more and produces more QALYs than day-case laparoscopic cholecystectomy, at an acceptable cost of QALY of £3568/QALY (see Table 9). The increased costs are driven by the additional length of stay in the inpatient arm. The QALY gains result from a higher estimated bile duct injury rate in the day-case cohort; however, this rate should be interpreted with caution as it is based on 1 event in the day-case arm and 0 events in the inpatient arm. This uncertainty is reflected in probabilistic sensitivity analysis, in which inpatient laparoscopic cholecystectomy has a 56% chance of being cost effective at a threshold of £20,000 per QALY compared with day case laparoscopic cholecystectomy. Therefore, it may be more appropriate to view inpatient laparoscopic cholecystectomy as more expensive than day-case laparoscopic cholecystectomy with comparable QALY outcomes.
Cost Effectiveness Results for Day Case versus Inpatient Laparoscopic Cholecystectomy.
One study provided low quality evidence about conservative management in comparison with laparoscopic cholecystectomy, and there was some evidence that laparoscopic cholecystectomy was superior to conservative management. Readmission rates were lower in the laparoscopic cholecystectomy group than the conservative management group, and 44% of people in the conservative management group also required cholecystectomy
Three studies provided moderate to low quality evidence about laparoscopic cholecystectomy in comparison with laparoscopic cholecystectomy with intraoperative cholangiography and there was uncertainty about whether the addition of intraoperative cholangiography was beneficial or not since there were no significant differences between the groups on any of the outcomes.
No evidence was available comparing cholecystostomy to laparoscopic cholecystectomy, and so no conclusions can be drawn.
Five studies providing moderate to low quality evidence compared day-case laparoscopic cholecystectomy with planned inpatient laparoscopic cholecystectomy. No differences in readmission rate or quality of life were found. Unplanned inpatient admissions were required in 12.1% of planned day case procedures.
A directly applicable original health economic model analysis with minor limitations suggests laparoscopic cholecystectomy is cost effective compared with conservative management. Comparisons of laparoscopic cholecystectomy with laparoscopic cholecystectomy and intra-operative cholangiography and day-case with inpatient laparoscopic cholecystectomy were limited by the quality of the clinical evidence.
| Relative value of different outcomes | In general, studies were underpowered for the outcomes that the GDG had chosen and so confidence intervals relating to effect estimates for many outcomes were wide. This was expected, since events such as mortality and bile duct injury are relatively rare and sample sizes would need to be very large to observe sufficient rare events. There was also a lack of randomised controlled trial (RCT) evidence for some of the comparisons that were considered as part of this review. The GDG felt that it is generally accepted that cholecystectomy is safe and effective, and that conducting clinical trials where participants for whom cholecystectomy is appropriate are randomised to a treatment that may be less effective may be seen as unethical, thus limiting the amount of RCT evidence that is available. |
|---|---|
| Trade off between benefits and harms | Laparoscopic cholecystectomy compared with conservative management The evidence comparing conservative management to laparoscopic cholecystectomy was limited. Only 1 underpowered study met the inclusion criteria. The evidence showed that disease progression was statistically significantly lower in patients offered cholecystectomy compared with conservative management, but there were no statistically significant differences between the groups in terms of readmission, or mortality. Other outcomes that the GDG felt were critical to their decision making (requirement for additional intervention, and length of stay) were not reported by the study. The GDG agreed that the results imply that conservative management is only suitable as a temporary management strategy, as patients offered conservative management continued to have recurrent biliary events. This supports the GDG's views that once a gallbladder develops gallstones, it will continue to do so, and that the only way of stopping gallstone formation and preventing recurrent biliary events is cholecystectomy. The GDG agreed that laparoscopic cholecystectomy has been established as a relatively safe procedure, and the risks of disease progression outweigh the risks of this type of surgery. Of course all surgeries carry some risks, particularly for people with complex, multiple conditions and for these patients the risk of surgery may well outweigh the risk of disease progression. Furthermore, some patients may choose not to opt for surgical treatment. Thus, the GDG agreed that conservative management should only be considered when cholecystectomy is not appropriate, and felt that this is standard practice across the NHS already and so did not require a specific recommendation to be made. However, the GDG did feel that sometimes older patients and those with comorbidities may be assumed to be inappropriate for laparoscopic cholecystectomy before they have even been assessed. Therefore the group felt the need to reinforce the message that laparoscopic cholecystectomy is becoming increasingly safer and that it is inappropriate not to offer this treatment to someone where the risks of disease progression outweigh the risks of surgery. The GDG felt justified in making a strong recommendation to offer laparoscopic cholecystectomy to all patients with symptomatic gallbladder stones, since there is sufficient acceptance of this procedure in the clinical community which is supported by the GDG's experience, health economic evidence, and by a small amount of clinical evidence found in this review. Laparoscopic cholecystectomy compared with percutaneous cholecystostomy No evidence that met the inclusion criteria was available about comparing percutaneous cholecystostomy to laparoscopic cholecystectomy and so the GDG discussed this treatment option by reflecting on its knowledge and experience. Percutaneous cholecystostomy is widely accepted to be a safe and effective treatment for patients with a gallbladder obstruction. The GDG felt that percutaneous cholecystostomy is only used when cholecystectomy is not an appropriate treatment option (either because of patient preferences or clinical judgement), and that this is standard practice across the NHS. However, the GDG agreed that there is inappropriate variation in the way that patients are managed once a percutaneous cholecystostomy has been placed. This was attributed to the fact that patients presenting with symptomatic gallbladder stones will usually be admitted to a surgeon, and when surgery is identified as not being an appropriate treatment option, the patient is then considered for percutaneous cholecystostomy. The GDG felt that some patients are never reassessed for cholecystectomy, even when their clinical condition has improved making surgery an appropriate option for them. The GDG agreed that there is a high incidence of recurrent gallbladder sepsis in this group of patients which has a high risk of mortality associated with it. Thus, the GDG agreed to recommend that patients with percutaneous cholecystostomy should be reviewed for reconsideration for cholecystectomy once adequate drainage has been established and the acute episode has resolved in order to minimise the risk of recurrent sepsis and mortality. Laparoscopic cholecystectomy compared with laparoscopic cholecystectomy with intraoperative cholangiography (IOC) Intraoperative cholangiography is performed during laparoscopic cholecystectomy to improve the safety of the procedure, by improving visualisation of the biliary anatomy and therefore reducing bile duct injuries, and it can also identify common bile duct stones that weren't suspected at earlier diagnostic work up. The evidence review considered the use of routine IOC, selective IOC, and not using IOC at all, but there were no statistically significant differences for any of the outcomes that the GDG had chosen to base its decisions on. The GDG felt that this was because the studies were underpowered, and so were not confident in these findings. Thus, the discussions about IOC were based on the knowledge and experience of the GDG. The GDG acknowledged that the individual studies included in the review drew mixed conclusions about whether IOC was useful or not, and this reflected current differences in clinical opinions about the usefulness of the procedure. The GDG reflected on its knowledge and experience and agreed that using IOC to improve visualisation of biliary anatomy offers clear benefits during laparoscopic cholecystectomy. This is because improved visualisation improves the surgeon's ability to plan and progress with laparoscopic cholecystectomy, making it safer for the patient. Because of this the GDG agreed that it is desirable for all surgeons performing laparoscopic cholecystectomy to be competent in IOC. However, there was further debate in relation to the benefits and harms of IOC for detecting common bile duct stones, and these discussions linked to evidence about managing common bile duct stones which is presented in section 4.6 of this guideline, where it was agreed that all common bile duct stones need to be removed, regardless of whether they are causing symptoms or not. Common bile duct stones found during IOC can be removed laparoscopically by a surgeon who has undergone specialist training. Most surgeons trained to do laparoscopic cholecystectomy do not have this specialist training and will not remove bile duct stones laparoscopically. Alternatively, common bile duct stones can be removed endoscopically. This can be done pre-, intra-, or post-operatively by someone who has undergone specific endoscopy training. Although some surgeons are appropriately trained for performing endoscopies, the majority of surgeons trained to perform laparoscopic cholecystectomy are not, and surgical lists are not usually planned with both a surgeon and an endoscopist in attendance. Thus the GDG agreed that few teams would be equipped to deal with common bile duct stones that might be identified. The GDG felt that once common bile duct stones are discovered, they need removing either at the time of discovery or within 2 or 3 days. Thus, if an unequipped team discovered stones the team would need to make an unplanned and urgent call for an appropriately trained person to remove the stones during the operation. Alternatively stones would need to be left in situ and the patient rescheduled for another intervention, thus exposing the patient to the risk of the stone developing serious complications during the wait, and the risks associated with undergoing a second intervention (such as risks of anaesthesia and procedural errors). Thus the group agreed that there was insufficient evidence of benefit to support a recommendation about using IOC for diagnosing common bile duct stones, or when inadequate preoperative diagnostic tests have been performed. Instead, the GDG felt that good preoperative examinations using diagnostic methods outlined in section 4.2 would improve the chances of common bile duct stones being identified preoperatively so that subsequent cholecystectomy can be planned to deal with common bile duct stones by appropriate surgeons and/or endoscopists. Overall, the GDG felt that IOC is beneficial for visualising the biliary anatomy and making laparoscopic cholecystectomy safer, and because of this surgeons performing laparoscopic cholecystectomy should be competent to use the procedure. However, the use of IOC varies in current clinical practice as there are uncertainties about its potential harms when unexpected common bile duct stones are discovered by surgeons who do not have specialist training to remove the stones laparoscopically. Thus the GDG decided not to make a recommendation about IOC. Instead a research recommendation was made to support the development of more research to explore the potential benefits and harms of this procedure. The GDG also wanted to stress that the lack of a recommendation should not preclude suitably trained individuals from using IOC at their discretion in the appropriate clinical situations. Day-case laparoscopic cholecystectomy compared with inpatient stay laparoscopic cholecystectomy The choice about whether to perform laparoscopic cholecystectomy as a day-case procedure or as planned inpatient admission was also reviewed. The evidence revealed that there were no statistically significant differences for any of the outcomes that the GDG had chosen to base their decisions on. Furthermore, the GDG acknowledged that some of the evidence included in this comparison came from older studies that might not accurately reflect continuing improvements in surgical outcomes that have occurred over time. Thus the discussions about day-case or planned inpatient admission were based on the knowledge and experience of the GDG. The GDG agreed that there are no differences in important outcomes between day-case and inpatient procedures. The GDG discussed whether day-case surgery should always be offered to patients, since it offers the benefits of enabling the patient to return home sooner, reducing the risk of hospital acquired infections, and reducing the use of NHS resources. However, there are various patient and clinical factors that need to be taken into consideration. For example, patients require adult supervision for 24hours after cholecystectomy and people who do not have someone at home to supervise them will require an inpatient stay. Similarly, some patients with comorbidities may need admitting to hospital before or after the operation to manage their other conditions. Thus the GDG felt that decisions about whether to plan day-case or inpatient cholecystectomy should be decided at a local level on a case by case basis to take the various patient and clinical factors into account. |
| Consideration of Health Benefits and Resource Use | For laparoscopic cholecystectomy compared with conservative management, the economic model suggests the costs and health benefits of surgery outweigh the health detriment of leaving the gallbladder in situ and having recurrent biliary events. The GDG noted the economic model considered this question as a binary choice – either removing the gallbladder or leaving it in situ – that is unlikely to occur in clinical practice. In sensitivity analyses, the economic model almost always produces higher costs and lower QALYs, but did not specifically consider older patients with comorbidities. For laparoscopic cholecystectomy versus laparoscopic cholecystectomy with intra-operative cholangiography, the economic model suggests laparoscopic cholecystectomy is a more cost-effective option. However, this reflects clinical evidence of small non-statistically significant differences in rates of operative consequences, in which the GDG were not confident. Probabilistic analysis of the model suggested laparoscopic cholecystectomy was the cost-effective option in around 2/3 of cases over a wide range of assumed values of 1 QALY, so intra-operative cholangiography may be cost effective in some situations and when done by appropriately trained surgeons Whilst the comparison of day case to inpatient laparoscopic cholecystectomy suggests inpatient laparoscopic cholecystectomy is more cost effective, this again is based on clinical evidence of small non-statistically significant differences (in bile duct injury rates) in which the GDG were not confident. Increased costs of inpatient laparoscopic cholecystectomy were driven by higher lengths of stay. It may be more appropriate to view inpatient laparoscopic cholecystectomy as more expensive than day-case laparoscopic cholecystectomy with comparable QALY outcomes. |
| Quality of evidence | Evidence was of moderate to low quality, and some comparisons had no evidence at all. |
| Other considerations | None |
Offer laparoscopic cholecystectomy to people diagnosed with symptomatic gallbladder stones.
Offer day-case laparoscopic cholecystectomy for people having it as an elective planned procedure, unless their circumstances or clinical condition make an inpatient stay more appropriate.
Offer percutaneous cholecystostomy to manage gallbladder empyema when:
Reconsider laparoscopic cholecystectomy for people who have had percutaneous cholecystostomy once they are well enough for surgery.
What are the benefits and harms, and cost effectiveness of routine intraoperative cholangiography in people with low to intermediate risk of common bile duct stones?
In the evidence reviewed for this guideline, there was a lack of randomised controlled trials of intraoperative cholangiography. Therefore, there is a need for large, high-quality trials to address clinical questions about the benefits and harms of intraoperative cholangiography.
Which strategies should be used for managing common bile duct stones?
A single search was performed for questions 4a, 4b, 4c and 5 which identified 10,976 references. After removing duplicates and screening the references based on their titles and abstracts, 210 references were obtained and reviewed against the inclusion and exclusion criteria for this review question (appendix C), and 47 references met the inclusion criteria overall. Details of excluded studies and reasons for their exclusion are in appendix F.4.
Of the 47 included references, 24 references relating to 24 randomised controlled trials were included in this review question (see study flow chart, appendix E.4) which specifically addressed the following comparisons for managing common bile duct stones:
Data were extracted into detailed evidence tables (see appendix G.4) and are summarised in Table 10 below. Each study was assessed for methodological quality using randomised controlled trial checklists, and some studies had methodological flaws which impacted on the overall quality of the evidence. For example, some studies didn't report randomisation procedures and those that did sometimes used inappropriate methods (patient identifying numbers - Hong, 2006), or had methods that led to differences in baseline characteristics of the groups (for example, Hui, 2002) which could bias results.
Summary of included studies for question 4c.
Data from multiple studies were available for most of the comparisons; therefore, meta-analysis was performed wherever possible (see appendix H.7). In addition, network meta-analysis was possible to compare the different methods for ERCP clearance of the duct (pre-, intra-, or post- operative) in comparison with surgical clearance. Standard GRADE approaches were used to assess pairwise comparisons, and a modified version of the GRADE framework was used to assess the evidence analysed using network meta analysis (see appendix I.6 for full GRADE profiles). In the modified approach, the standard GRADE criteria still apply but additional factors are also considered such as how each ‘link’ or pairwise comparison within the network applies to the others (see NICE clinical guideline CG173, appendix D for details of the additional factors that may be considered in a network meta analysis). However, since the network of evidence for this review question was relatively small, very few of the additional criteria actually applied. Details of the specific reasons for downgrading are stated in the footnotes of the relevant GRADE profiles in appendix I.6.
A literature search was conducted jointly for questions 4 and 5 by applying standard health economic filters to the clinical search strategies (see Appendix D). For questions 4 and 5, 1,396 references were retrieved, of which 1 was retained for question 4b and 1 was retained for question 5.
Gurusamy et al. (2012) (see Table 11) used a decision tree to compare intraoperative and preoperative ERCP for patients with gallbladder stones and CBDS. The study modelled the success or failure (and subsequent repeat ERCP) of CBDS extraction, complications of ERCP, consequent cholecystectomy (laparoscopic or open) and mortality from symptoms and open operations. Whilst this analysis was directly relevant to the UK NHS, the model did not compare all the interventions included in this guideline question (as postoperative ERCP and intraoperative bile duct exploration were not included).
Health Economic Evidence – Intra-Operative Versus Pre-Operative ERCP for CBDS.
As no existing health economic studies were found that addressed all the comparisons in questions 4 and 5, an original economic model was constructed. A full description of the health economic model can be found in in Appendix J; a summary is presented in section 4.5.3). For questions 4c (managing symptomatic CBDS), the GDG did not prioritise the biliary stents comparison and no evidence was found to enable the modelling of the day case versus inpatient ERCP comparison.
The health economic model found that conservative management is more costly and produces fewer QALYs than ERCP and is therefore said to be dominated (see Table 12). This remained true in 100% of probabilistic model runs. The increased costs and decreased QALYs are driven by patients retaining their gallstones, having further symptomatic episodes and requiring additional ERCPs that would be undertaken non-electively in the conservative management arm.
Cost effectiveness results for ERCP versus conservative management.
ERCP and laparoscopic cholecystectomy is more costly and produces more QALYs than ERCP alone, at an acceptable cost per QALY of £4680 (see Table 13). The increased costs and QALYs are driven by all patients having a laparoscopic cholecystectomy that avoids patients living with asymptomatic gallstones or having further symptomatic episodes. In probabilistic sensitivity analysis, ERCP with laparoscopic cholecystectomy has a 98.5% chance of being cost effective at a threshold of £20,000 per QALY, compared with ERCP alone.
Cost effectiveness results for ERCP with laparoscopic cholecystectomy versus ERCP alone.
In an incremental analysis of the 4 options for managing CBDS with laparoscopic cholecystectomy, laparoscopic cholecystectomy with intraoperative ERCP was less costly and produced more QALYS than either laparoscopic cholecystectomy with bile duct exploration, laparoscopic cholecystectomy with preoperative ERCP or laparoscopic cholecystectomy with postoperative ERCP. Therefore, laparoscopic cholecystectomy with intraoperative ERCP is the dominant option (Table 14). In probabilistic sensitivity analysis, laparoscopic cholecystectomy with intraoperative ERCP is cost effective in 84.6% of simulations (see Table 15).
Cost effectiveness results for laparoscopic cholecystectomy with bile duct exploration versus laparoscopic cholecystectomy with pre-, intra- or postoperative ERCP.
Probabilistic Sensitivity Analysis Results.
Pre- and postoperative ERCP options have increased costs as they require 2 hospital admissions. The differences between intraoperative ERCP and bile duct exploration are driven by higher rates of bile leaks and extra ERCPs required to clear the CBDS, both of which are worse for bile duct exploration. However, the QALY differences are small and both these parameters are based on non-statistically significant differences from 1 or 2 RCTs.
The model does not take account of any additional implementation costs or wider opportunity costs that occur to facilitate laparoscopic cholecystectomy with intraoperative ERCP or bile duct exploration.
Low to very low quality evidence provided by up to 8 randomised controlled trials showed that there were no significant differences between ERCP and conservative management. The only exception to this was a statistically significantly lower requirement for additional ERCP in the ERCP group than the conservative management group.
Moderate quality evidence was provided by 2 randomised controlled trials and showed that ERCP followed by laparoscopic cholecystectomy was preferable to ERCP alone.
Very low quality evidence provided by up to 11 randomised controlled trials comparing ERCP clearance of the bile duct plus laparoscopic cholecystectomy with surgical clearance of the bile duct plus laparoscopic cholecystectomy was inconclusive. Where network meta-analyses were conducted, all options had wide credibility intervals meaning there was uncertainty about which option was best. Furthermore, some outcomes had evidence that couldn't be meaningfully compared.
Low quality evidence was provided by a single study and showed there were no statistically significant differences between biliary stenting and clearing the bile duct. The only exception to this was a statistically significantly lower requirement for additional ERCP in the stent group than in the cleared duct group.
No evidence comparing day-case ERCP to planned inpatient ERCP was found
A directly applicable published cost–utility study with potentially serious limitations suggested that intraoperative ERCP is cost saving and generates more QALYs when compared with preoperative ERCP.
A directly applicable original health economic model analysis with minor limitations suggests ERCP is cost effective compared with conservative management and ERCP with laparoscopic cholecystectomy is cost effective compared with ERCP alone. Based on limited and inconclusive clinical evidence, laparoscopic cholecystectomy with intra-operative ERCP dominates other options but without taking account of any additional implementation costs.
| Relative value of different outcomes | In general, studies were underpowered for the outcomes that the GDG had chosen and so confidence intervals relating to effect estimates for many outcomes were wide. This was expected, since events such as mortality and bile duct injury are relatively rare and sample sizes would need to be very large to observe sufficient rare events. There was a lack of randomised controlled trial evidence for some of the comparisons that were considered as part of this review. The GDG felt that this is because surgical removal of the gallbladder, and surgical and/or endoscopic stone extraction of stones in the common bile duct are accepted as best practice. Alternative management options are only considered for patients for whom surgery/endoscopy is not appropriate, such as those who have complex, comorbid conditions, or those who make an informed choice not to opt for surgery/endoscopy. Continual improvements in clinical outcomes for surgery and endoscopy has meant that surgery and/or endoscopy can be offered to an increasing number of people for whom traditionally these options would not be appropriate. It would be seen as highly unethical to randomise patients for whom surgery and/or endoscopy is appropriate to inferior treatments, and thus randomised controlled trial evidence in this area is limited. The absence of evidence and the wide confidence intervals surrounding most effect estimates meant the GDG relied on their knowledge and experience for most of the discussions of the interventions under review. |
|---|---|
| Trade off between benefits and harms | ERCP compared with conservative management Evidence about conservative management in comparison with ERCP for common bile duct stones was inconclusive. There were no significant differences in any of the critical or important outcomes and this was attributed to lack of statistical power in the evidence. GDG experience is that ERCP is a safe procedure and is being increasingly offered to a wide range of clinically compromised and complex patients, but there was no evidence to say that ERCP should always be performed. The GDG agreed that there will always be patients where the risks of intervening with ERCP outweigh the risks associated with not intervening, and for these patients conservative management is an appropriate treatment option. The GDG felt that decisions about who should be offered conservative management should be taken locally, since such cases will be complex and dependent on the skills and experience of local health care professionals. The GDG felt it was not necessary to make a recommendation about conservative management, as doing so may make it seem that conservative management is an appropriate option for all patients, which they agreed it wasn't. ERCP + laparoscopic cholecystectomy compared with ERCP alone The GDG discussed whether laparoscopic cholecystectomy should always be offered to people who have ERCP, in order to prevent recurrent disease. Evidence about this came from 2 studies, where statistically significant differences on 2 of the 3 critical outcomes that the GDG had specified were found. The evidence showed significant reductions in disease progression/recurrence, and in the requirement for additional ERCP in the group who received laparoscopic cholecystectomy after ERCP, compared with the group that had ERCP alone. Furthermore, subsequent cholecystectomy was required for 25% of patients who had ERCP alone. The evidence supported the knowledge and experience of the group, where it was agreed that disease recurrence is common, and the only way to prevent this is to remove the gallbladder to stop new gallstones forming. Thus the GDG recommended that laparoscopic cholecystectomy should be offered to patients who have common bile duct stones since the harms of recurrent events outweighed the harms of laparoscopic cholecystectomy. Pre/post/intra operative ERCP + LC compared with BDE+LC The evidence relating to these comparisons came from 9 studies that were analysed using network meta-analysis whenever possible. In some instances there was insufficient evidence to form a network so pairwise meta-analysis was performed instead. For some outcomes it wasn't possible to meta analyse at all. This was usually when an outcome was possible in one arm of the comparison but not in the other (for example, requirement for cholecystectomy isn't possible in people who have already had cholecystectomy, but is possible in those waiting for cholecystectomy). In these cases simple rates were presented. Network meta-analysis was possible for one critical outcome (length of stay) and three important outcomes (missed stones, failed procedure, conversion to open surgery). The results of the network analyses were inconclusive. All of the options had wide credibility intervals on all of the outcomes, indicating there was uncertainty about which options offered the best outcomes and which offered the worst. Mortality was also chosen as a critical outcome by the GDG, and although mortality data were reported by most of the studies included in these comparisons, the event rates were often zero in both arms of the trial. Since zero event data cannot be meaningfully analysed, only evidence from studies where deaths were observed were compared. It was only possible to compare mortality data in a pairwise comparison between preoperative ERCP+LC versus BDE+LC. The results of this comparison were not statistically significant, and this was attributed to a lack of statistical power in the analysis. The point estimate did indicate that deaths were lower in preoperative ERCP group than the bile duct exploration group. Requirement for more than 1 ERCP was chosen as an important outcome by the GDG, but this outcome was not possible to observe in the surgical bile duct exploration arm, since ERCP was not part of treatment. Furthermore, not all studies reported this outcome so it was not possible to meta-analyse the data. Rates of multiple ERCP were provided but it was not possible to calculate confidence intervals around the rates, and so this outcome had unknown levels of uncertainty. Thus, the GDG agreed that the mixed and unclear evidence was a reflection of current clinical reality as there are lots of different options for managing common bile duct stones that are used in different clinical scenarios. However, the GDG acknowledged that the studies used within this evidence often excluded patients who were older, complex, and had comorbidities from their protocols. Thus, the range of clinical scenarios that the evidence covered does not reflect reality. Since there was insufficient evidence, decisions were based on knowledge and experience of the group. A discussion about the merits of ERCP+LC and BDE+LC took place. ERCP+LC requires two separate procedures, ERCP to remove the stone and LC to remove the gallbladder. Usually these procedures are done days or weeks apart from each other. Thus the patient will require a prolonged hospital stay during the wait, or will be admitted to hospital twice for each individual procedure. Furthermore, the patient would be exposed to the risks of anaesthesia/sedation twice, and there is a risk that their clinical condition could deteriorate in the time between the two procedures. BDE+LC on the other hand is performed as a single procedure and would usually result in the patient having a shorter total hospital stay (since multiple or prolonged admission is not necessary), is only exposed to one anaesthesia, and has their condition treated definitively in one procedure. Thus, the GDG agreed that in principle BDE+LC is superior to ERCP+LC. The GDG acknowledged that only a minority of surgeons can perform BDE+LC safely, as this is highly complex surgery that requires specialist knowledge and experience. The GDG felt that surgeons not trained to do bile duct exploration would not and should not undertake this procedure. Conversely, laparoscopic cholecystectomy performed after bile duct stones have been removed does not require a surgeon trained in bile duct exploration, and can be performed by any surgeon trained to perform laparoscopic cholecystectomy. Since ERCP is widely available throughout the NHS, the GDG felt that it could recommend that bile duct stones should be removed by ERCP before laparoscopic cholecystectomy, or by surgical bile duct exploration during laparoscopic cholecystectomy in order to reduce the risk of subsequent surgical complications. The GDG felt that the decision to use ERCP or bile duct exploration to clear the duct should be taken locally depending on the skills and experience of the team. Uncleared duct + biliary stent compared with cleared duct Only one study was identified that met the inclusion criteria. The study had a low rate of stone clearance which the GDG felt reflected the age of the paper as it was published in 1996, and higher rates of clearance are usually observed today. The GDG also acknowledged that the study randomised patients to biliary stents without attempting to clear the duct first, which is not ethical and wouldn't be done in practice today. The study was underpowered and so no clear conclusions could be drawn about the efficacy and safety of biliary stenting in comparison with ductal clearance. Biliary stenting is widely accepted to be an appropriate management strategy when stones cannot be cleared from the bile duct and surgery is not an appropriate treatment option. Biliary stenting is associated with complications that can be life-threatening, and clinicians must balance the risks of using stents with those associated with persevering in the attempt to clear the duct. The GDG felt that in general, most clinicians would strive for a cleared duct and the decision to abort ductal clearance would be taken locally depending on the individual patient's condition and the clinical experience of the team. Thus, the GDG did not feel it would be of use to make recommendations about when biliary stenting should be considered. However, the GDG felt that there was inappropriate variation in the management of patients once a biliary stent has been placed. Often biliary stents are used as long term treatment option with patients never getting reassessed for further surgical or endoscopic ductal clearance, even when their clinical condition has improved. Instead patients will go on to have their stent replaced indefinitely. The GDG were concerned about this, since patients would be at increased risk of recurrent events and would be at risk of experiencing cholangitis, one of the most serious complications associated with the use of stents. Although the evidence was inconclusive since the single study it was based on was underpowered, the point estimates showed that risk of death and disease progression were higher in the stent group than in the cleared duct group. Furthermore, the study did find a statistically significant difference in risk of cholangitis, with those in the stent group having a significantly greater risk of cholangitis than those in the cleared duct group. The GDG had not specified cholangitis as a specific outcome for this evidence review, instead it chose a composite outcome that included all types of progressive disease including cholangitis, pancreatitis and jaundice. So although the individual study observed a statistical difference in cholangitis risk, this was not observed in the analysis for this review, since the use of the composite outcome ‘disease progression’ had diluted the evidence. Since cholangitis is such a serious and life threatening complication, the GDG could not ignore the study findings and were satisfied that the evidence did support their knowledge and experience. Thus recommendations about managing stents once they have been placed were made. Day-case ERCP vs planned inpatient ERCP There was no evidence in relation to this comparison and so the GDG based its discussions on the knowledge and experience of the group, and also extrapolated from evidence presented in chapter 4.6 relating to day-case versus inpatient laparoscopic cholecystectomy. The GDG discussed whether day-case surgery should always be offered to patients, since it offers the benefits of enabling the patient to return home sooner, reducing the risk of hospital acquired infections, and reducing the use of NHS resources. However, there are various patient and clinical factors that need to be taken into consideration. For example, patients require adult supervision for 24hours after sedation and people who do not have someone at home to supervise them will require an inpatient stay. Similarly, some patients with comorbidities may need admitting to hospital before or after the operation to manage their other conditions. The GDG also acknowledged that definitions of inpatient admission and day-cases may vary, and that considerations of other aspects of hospital performance may influence decisions around length of stay. Thus, the GDG felt that decisions about whether to plan day-case or inpatient ERCP should be decided at a local level on a case by case basis to take the various patient and clinical factors into account. |
| Consideration of health benefits and resource use | For ERCP compared with conservative management, the economic model suggests that leaving patients with common bile duct stones results in net health losses (due to symptoms) and cost increases (due to extra non-elective interventional procedures being required). The GDG noted that the economic model considered a binary choice between ERCP and no intervention that is unlikely to occur in clinical practice. The economic model suggests that ERCP with laparoscopic cholecystectomy is cost effective compared with ERCP alone with an ICER of around £4,700 / QALY. The GDG agreed that the extra cost of laparoscopic cholecystectomy would be an appropriate trade off against the extra QALYs gained by relief of recurrent biliary symptoms. The economic analysis of laparoscopic cholecystectomy with laparoscopic bile duct exploration compared with laparoscopic cholecystectomy with pre-, intra- or postoperative ERCP suggests that intraoperative ERCP dominates other options. However, the GDG interpreted this result cautiously. They noted the underlying network meta-analysis was inconclusive. They also noted options modelled with 1 (intraoperative ERCP, laparoscopic bile duct exploration) rather than 2 (pre- and post-operative ERCP) hospital admissions are cheaper and also limited the time patients could become symptomatic and lose QALYs. However, the GDG were concerned that the economic model did not reflect implementation costs that would be associated with recommending intra-operative ERCP or laparoscopic bile duct exploration. The GDG felt that the economic analysis supported their view that laparoscopic cholecystectomy with bile duct exploration is superior to laparoscopic cholecystectomy with ERCP, but the decision on which method should be used to clear the bile duct should be taken locally based on the skills and experience of staff. The economic model did not consider biliary stents or day-case versus inpatient ERCP treatment. |
| Quality of evidence | The overall quality of the evidence was mixed, with the majority of evidence being low in quality. |
| Other considerations | None |
Offer bile duct clearance and laparoscopic cholecystectomy to people with symptomatic or asymptomatic common bile duct stones.
Clear the bile duct:
If the bile duct cannot be cleared with ERCP, use biliary stenting to achieve biliary drainage only as a temporary measure until definitive endoscopic or surgical clearance.
Use the lowest-cost option suitable for the clinical situation when choosing between day-case and inpatient procedures for elective ERCP.
What models of service delivery enable intraoperative endoscopic retrograde cholangiopancreatography (ERCP) for bile duct clearance to be delivered within the NHS? What are the costs and benefits of different models of service delivery?
Evidence reviewed for this guideline identified that intraoperative ERCP is both clinically and cost effective, but it is unclear whether delivery of this intervention is feasible in the NHS because of the way current services are organised. It is also unclear whether intraoperative ERCP will remain cost effective if services are reorganised.
In adults with acute cholecystitis or symptomatic common bile duct stones, should cholecystectomy be performed during the acute episode (early) or should intervention be delayed until the acute episode has subsided (delayed)?
A single search was performed for questions 4a, 4b, 4c and 5 which identified 10,976 references. After removing duplicates and screening the references based on their titles and abstracts, 210 references were obtained and reviewed against the inclusion and exclusion criteria for this review question (appendix C), and 47 references met the inclusion criteria overall. Details of excluded studies and reasons for their exclusion are in appendix F.4
Of the 47 included references, 8 references relating to 8 randomised controlled trials were included in this review question (see study flow chart) which specifically addressed the timing of intervention for gallstone disease, and for the purposes of this review the following definitions were used:
Studies that didn't use these definitions were not eligible for inclusion.
Of the 8 included studies, 7 related to acute cholecystitis (Gul, 2013; Johansson, 2003; Kolla, 2004; Lai, 1998; Lo, 1998; Macafee, 2009; Yadav, 2009) and one related to common bile duct stones (Reinders, 2010). All studies were assessed for methodological quality using randomised controlled trial checklists, and meta-analysis was performed wherever possible (see appendix H.7).
Standard GRADE processes were applied to assess the quality of the evidence in relation to the following outcomes:
Full GRADE profiles are available in appendix (I.7).
Overall, the individual studies had good methodological quality, with adequate randomisation and appropriate study conduct. Evidence was available in relation to readmission and length of stay, but outcomes such as mortality and quality of life were not reported by any of the studies. There was a lack of evidence in relation to common bile duct stones.
Summary of included studies for question 5.
A literature search was conducted jointly for questions 4 and 5,by applying standard health economic filters to the clinical search strategies (see Appendix D). 1,396 references were retrieved for questions 4 and 5, of which 1 was retained for question 4b and 1 was retained for question 5.
Wilson et al. (2010) (see Table 17) used a decision tree to compare early and delayed laparoscopic cholecystectomy for acute cholecystitis. The model estimated the development of 4 symptoms at 9 weeks during an 18-week delay period, with a time horizon of 1 year.
Health Economic Evidence – Early versus Delayed Laparoscopic Cholecystectomy For Acute Cholecystitis.
Whilst Wilson et al. (2010) was directly relevant, it had a number of limitations, including the delay length, time horizon, no consideration of CBDS and utility data.
Because Wilson et al.'s (2010) analysis had potentially serious limitations, and did not address all relevant questions in this area, an original economic model was constructed. A full description of the health economic model can be found in in Appendix J, a summary is presented in section 4.5.3.
The health economic model found that delayed laparoscopic cholecystectomy is more costly and produces more QALYs than early laparoscopic cholecystectomy, but an ICER of over £200,000 per QALY is above that which is usually accepted as being cost effective (see Table 18). In probabilistic sensitivity analysis, early laparoscopic cholecystectomy has an 88% chance of being cost effective at a threshold of £20,000 per QALY compared with delayed laparoscopic cholecystectomy. The increased costs are driven by the additional length of stay and the small QALY gains are sensitive to estimated rates of bile duct injury.
Cost effectiveness results for early laparoscopic cholecystectomy versus delayed laparoscopic cholecystectomy.
Delayed laparoscopic cholecystectomy after ERCP is more costly and produces fewer QALYs than early laparoscopic cholecystectomy after ERCP and is therefore said to be dominated (see Table 19). In probabilistic sensitivity analysis, early laparoscopic cholecystectomy post ERCP has a 100% chance of being cost effective at a threshold of £20,000 per QALY compared with delayed laparoscopic cholecystectomy after ERCP. The increased costs are driven by the additional length of stay and the QALY differences are small.
Cost effectiveness results for early laparoscopic cholecystectomy after ERCP versus delayed laparoscopic cholecystectomy after ERCP.
Moderate to low quality evidence was provided by 6 randomised controlled trials comparing early versus delayed laparoscopic cholecystectomy for acute cholecystitis, and demonstrated that readmission rates and length of stay were lower in the early compared with the delayed group and quality of life was higher.
A published cost–utility analysis and an original health economic model both suggest that early laparoscopic cholecystectomy for acute cholecystitis is cost effective compared with delayed laparoscopic cholecystectomy.
One randomised controlled trial provided low quality evidence comparing early versus delayed laparoscopic cholecystectomy following ERCP for common bile duct stones. Evidence was provided for only 1 of the 5 outcomes for this comparison. This is because the study either didn't report the data, or because data were reported but zero events occurred. The 1 outcome for which moderate quality evidence was available showed that there was no statistically significant difference in length of stay between the 2 groups.
A directly applicable original health economic model analysis with minor limitations suggests that early laparoscopic cholecystectomy following ERCP for common bile duct stones is cost effective compared with delayed laparoscopic cholecystectomy.
| Relative value of different outcomes | The GDG chose readmission due to symptoms and readmission due to surgical complications as outcomes that were critical to their decision making, since it was felt that these outcomes would capture important differences between groups in relation to the complications of surgery. Readmission due to symptoms was reported in some of the studies relating to acute cholecystitis but was not reported in the study relating to common bile duct stones. Readmission due to surgical complications was not reported by any of the studies included in this review question. The GDG felt that the studies that didn't report these critical outcomes probably didn't observe any readmissions, although there was no way that they could be certain of this. The GDG felt that in reality, readmission rates are low and the individual studies were underpowered to observe these relatively rare events. Likewise, mortality was reported by most of the included studies, but no events were observed since the studies were underpowered to detect this relatively rare event. The GDG felt that it is more likely for studies to report zero mortality events than zero readmission events due to the greater importance placed on mortality than readmission. The GDG felt that the lack of evidence supports their views that readmission and mortality are rare, and focused their attention on the evidence relating to length of stay which was felt to be the best measure of clinically important differences between early and delayed laparoscopic cholecystectomy (LC) for the majority of patients. |
|---|---|
| Trade off between benefits and harms | The earlier LC is performed the less potential there is for recurrent events during the wait for surgery, therefore there are expected benefits in terms of patient quality of life and resource utilisation. However, there are concerns that early intervention could cause harm to patients by increasing the complication rate and extending length of stay. Thus delayed surgery may have greater benefits in the longer term if surgery is safer, although patients may experience recurrent events during the period of delay which could necessitate emergency intervention. In the evidence relating to acute cholecystitis, the early LC group had significantly fewer readmissions than the delayed group as well as a significantly shorter length of stay, which the GDG interpreted as meaning that early intervention does not increase complications. The patient representatives on the group also expressed a preference for early LC due to their personal experiences of having to wait for surgery. The patient members expressed that their quality of life was compromised during the wait by avoiding foods that triggered their symptoms, worrying about social activities and not feeling confident to return to work until treatment has been completed due to worries about pain and other complications, although no evidence relating to quality of life was found in this review. The GDG also felt that the absence of evidence relating to harm, especially mortality, probably reflected the fact that these events are rare, and in the majority of cases the benefits of early intervention would far outweigh the risks. Thus the GDG felt that there was sufficient evidence relating to the benefits of early LC to enable them to recommend that early LC should be offered to patients with acute cholecystitis. There was limited evidence in relation to common bile duct stones. The only evidence available for this population came from one small study for the outcome ‘length of stay’. The study did not find a significant difference and this was attributed to a lack of statistical power. Furthermore the GDG identified that the study had excluded patients with biliary pancreatitis, which represents a large proportion of people with common bile duct stones. Thus the paper lacks transferability to routine clinical settings. Overall, the lack of evidence meant the GDG were uncertain about whether early or delayed LC was the optimum management strategy for common bile duct stones. The group felt that in reality there was great variation in when people were offered LC following removal of bile duct stones, and agreed that the delay is not necessarily based on clinical decision making and is more often due to a lack of capacity. The GDG felt that it probably wasn't appropriate to recommend the same time frame for LC for common bile duct stones as for acute cholecystitis. This is because common bile duct stones are more complicated and associated with higher risks than gallbladder stones that cause acute cholecystitis, particularly when severe pancreatitis has developed. Performing early surgery may therefore increase the risk of bile duct injury and mortality. Since there was no evidence to support or refute this notion, the GDG were not confident about the optimum time frame in which to offer LC and did not make a specific recommendation. |
| Consideration of Health Benefits and Resource Use | In both health economic comparisons, the higher cost of delayed laparoscopic cholecystectomy is driven by the longer length of stay associated with delayed laparoscopic cholecystectomy. QALY differences are small, but suggest that delaying surgery is associated with similar or worse net quality of life than early surgery. |
| Quality of evidence | The GDG agreed that the evidence was of moderate to low quality. |
| Other considerations | The GDG acknowledged that decisions about when to perform LC may be based on capacity issues rather than clinical appropriateness. Thus, the GDG wanted to emphasise that the recommendations made here should improve patient flow and release capacity, rather than put additional constrains on it. This is because the number of people requiring LC in a given year is a finite, and operating on them earlier will not increase the number of people requiring LC. Instead, offering early cholecystectomy may actually reduce the number of readmissions, emergency operations, and length of stay. The GDG felt the recommendations are achievable but acknowledged that some providers may need to focus on redesigning their services in order to implement them. |
Offer early laparoscopic cholecystectomy (to be carried out within 1 week of diagnosis) to people with acute cholecystitis.
In adults with common bile duct stones, should laparoscopic cholecystectomy be performed early (within 2 weeks of bile duct clearance), or should it be delayed (until 6 weeks after bile duct clearance)?
In the evidence reviewed for this guideline, there was a lack of randomised controlled trials of intraoperative cholangiography, and the evidence that was available did not support the knowledge and experience of the Guideline Development Group. Therefore, there is a need for large, high-quality trials to address clinical questions about the benefits and harms of intraoperative cholangiography
What are the information and education needs of patients and carers of people with gallstone disease?
The aim of this review question was to identify areas for which information and education should be provided. This question did not aim to compare the effectiveness of different education programmes/strategies.
A systematic search was conducted (see appendix D) which identified 7,862 references. After removing duplicates the references were screened on their titles and abstracts and 61 references were obtained and reviewed against the inclusion and exclusion criteria as described in the review protocols (appendix C).
Overall, 56 studies were excluded as they did not meet the eligibility criteria. A list of excluded studies and reasons for their exclusion is provided in appendix F.
Five studies met the eligibility criteria and were included. Data were extracted into detailed evidence tables (see appendix G) and are summarised in the table below.
Summary of included studies for question 6.
Studies were quality assessed using methodology checklists. For qualitative studies, the NICE qualitative checklist was used. The NICE guidelines manual does not provide a checklist for surveys, and so a checklist originally published in the British Medical Journal (see appendix K) was used to aid the quality assessment of these studies.
The GRADE framework for assessing quality was modified for this review in that studies, rather than outcomes, were assessed for quality (see appendix I.8) on the following basis. Overall, the studies were of very low quality and had serious limitations; Survey studies lacked detailed analysis and qualitative studies lacked rigour as they did not utilise adequate research design and methodology, and they failed to provide rich data to support findings; studies were limited because 4 of the 5 studies were conducted outside the UK and so lack transferability, and most studies were from the perspective of elective cholecystectomy, meaning that non elective surgery and other treatment options are not adequately represented in the evidence base. Specific details about the reasons for this approach and the judgements made are provided in the footnotes of the profiles in appendix I.8.
A literature search was conducted for review question 6, by applying standard health economic filters to the clinical search strategies (see Appendix D). From the search, 504 references were retrieved, of which none were retained at title and abstract screening. Health economic modelling was not prioritised for this review question.
Very low quality evidence from 5 studies on people waiting for or undergoing surgery for gallbladder disease showed that patients requested more information on diet, wound management, pain management, and resuming normal activities. Some patients had no memory of the information that was provided to them, and some consulted the internet to acquire additional information. Some people did not know why they had to wait for elective surgery.
| Relative value of different outcomes | The GDG considered all themes/outcomes that arose from the evidence, but focused its discussions on the theme/outcome of diet. This is because most themes that arose such as wound management, resuming activity, memory and so on, were not specific to gallstone disease and the GDG agreed that other NICE products that already exist or that are in development (such as Patient experience in adult NHS services, Hospital admission pathway, Surgery pathway) would cover these general themes more appropriately. The group chose to focus discussions on diet because this is the area the patient representatives felt was most important, because diet was a key issue in the evidence, and because this issue was specific to gallstone disease. Previous review questions did not find sufficient evidence to support or refute a link between diet and gallstone disease, and diet as an intervention for managing gallstone disease was not evaluated in any of the review questions. Thus, discussions about diet were based on the knowledge and experience of the GDG and a co-opted expert dietician. |
|---|---|
| Trade off between benefits and harms | Some patients with gallstone disease find that eating and/or drinking triggers gallbladder pain and discomfort. This could be due to the mechanics of digestion because the gallbladder contracts on consuming food and drink in order to release bile to aid digestion, and when a person has gallstone disease the gallbladder can become inflamed and infected, and this contracting process can be painful. The GDG decided that it would not be useful to provide information to patients about food and drink that may potentially trigger symptoms because not everyone with gallstone disease experiences pain and discomfort when eating or drinking, not all food and drink triggers gallbladder pain, and the type of triggers vary for different individuals. Instead, the GDG felt that individual patients are best placed to identify which food and drink triggers their own pain. The GDG agreed that once a gallbladder has been removed the mechanism for experiencing pain is no longer there (since there is no gallbladder to contract and cause pain). The time between diagnosis and treatment for gallbladder disease is relatively short (usually less than 3 months), so the GDG felt that there was no harm in recommending that patients should avoid food and drink that triggers their pain between diagnosis and treatment, since harms to health or quality of life were expected to be very low. However, long term avoidance of specific foods, especially once gallstones or the gallbladder have been removed could be detrimental and are not recommended. The group acknowledged that some patients with gallstone disease continue to report symptoms or experience new symptoms once their gallbladder has been removed. The GDG felt this is probably due to some underlying condition that the patient either didn't notice before they experienced gallstone disease, or that coincidentally began at the same time as the symptoms of gallstone disease or cholecystectomy. The GDG felt that these symptoms are often misattributed to the absence of a gallbladder by both health professionals and patients themselves, and because of this patients may not receive appropriate investigations to identify the real cause of the symptoms. Thus, people may continue to experience unnecessary symptoms, or restrict trigger food and drink over the long term to avoid ongoing symptoms. Since long term dietary restriction may be harmful, and because patient quality of life is affected by on-going symptoms, the GDG felt that people experiencing new or existing symptoms after they have recovered from cholecystectomy should be encouraged to return to their health professional, and health professionals should investigate the patient appropriately. |
| Consideration of Health Benefits and Resource Use | Health economics were not considered as a priority for this review question. No economic studies were found. |
| Quality of evidence | The quality of the evidence was very low, and the studies contributing to the evidence did not provide adequate information about the particular aspects of dietary advice that should be provided. Therefore the GDG developed the recommendations using their knowledge and experience. |
| Other considerations | None. |
Advise people to avoid food and drink that triggers their symptoms until they have their gallbladder or gallstones removed.
Advise people that they should not need to avoid food and drink that triggered their symptoms after they have their gallbladder or gallstones removed.
Advise people to seek further advice from their GP if eating or drinking triggers existing symptoms or causes new symptoms to develop after they have recovered from having their gallbladder or gallstones removed.
What is the long-term effect of laparoscopic cholecystectomy on outcomes that are important to patients?
There is a lack of information on the long-term impact of cholecystectomy on patient outcomes. Many patients report a continuation of symptoms or the onset of new symptoms after laparoscopic cholecystectomy, and these affect quality of life. Research is needed to establish the long-term patient benefits and harms, so that appropriate information can be provided to patients to aid decision-making and long-term management of their condition.
All rights reserved. This material may be freely reproduced for educational and not-for-profit purposes within the NHS. No reproduction by or for commercial organisations is allowed without the express written permission of the National Institute for Health and Care Excellence.
Your browsing activity is empty.
Activity recording is turned off.
See more...
