4 Committee discussion
4.1 The clinical experts explained that since NICE published guidance on SeHCAT (tauroselcholic [75 selenium] acid) in 2012, awareness of bile acid diarrhoea as a condition has increased. They estimated that 1 in 20 people referred to a gastroenterology clinic because of chronic diarrhoea may have bile acid diarrhoea. Testing means that this condition can be identified and distinguished from diarrhoea-predominant irritable bowel syndrome (IBS‑D), and treatment offered. Clinicians now agree that it is important to be able to test for bile acid diarrhoea and to treat it.
4.2 The patient experts explained that having bile acid diarrhoea can affect quality of life and limit daily activities such as the ability to work. Diagnosing bile acid diarrhoea is important because it explains the person's symptoms and means they can have treatment. Also, it can support a request for reasonable adjustments at work. The committee considered how having a diagnosis affects treatment with bile acid sequestrants. These are unpleasant to take, and many people do not adhere to treatment. The clinical experts explained that people who have a diagnosis of bile acid diarrhoea may be more motivated to continue them than people who start them as a trial of treatment. The committee recognised that having a diagnosis of bile acid diarrhoea is helpful.
4.3 The external assessment group's (EAG) systematic review on the clinical effectiveness of SeHCAT testing (see section 3) found only 1 study in people with Crohn's disease. The committee concluded that SeHCAT testing in this population could be useful but there is not enough data available to understand its benefits and harms. It recommended that further research is done to show the clinical effectiveness of SeHCAT testing in people with Crohn's disease (see sections 5.1 to 5.4).
Evidence on the effects of SeHCAT testing in people with primary bile acid diarrhoea is limited in quality
4.4 There were 9 new studies published since NICE's original guidance in 2012. So, 24 studies in total were available for the primary bile acid diarrhoea population. Most of these provided data on response to treatment after a positive test result. The committee noted that the populations in these studies did not reflect the people who would be seen in NHS clinical practice. This is because people with chronic diarrhoea in the NHS are likely to be offered tests to identify other conditions with similar symptoms first, before SeHCAT testing. The populations in the studies were not likely to have had faecal immunochemical tests and faecal calprotectin tests before SeHCAT because the studies predate their introduction. Also, the committee noted that the studies were often small and had methodological limitations. Most only described limited short-term outcomes. It recommended that further research is done to show the clinical effectiveness of SeHCAT testing in people with primary bile acid diarrhoea (see sections 5.1 to 5.4).
4.5 The committee noted that, although most studies described response to treatment after a positive SeHCAT test result, not everyone with a positive test result was offered bile acid sequestrants. Between 70% and 100% of people who had a positive SeHCAT test at a 15% threshold had bile acid sequestrants. The studies provided no information on how treatment decisions were made, and it was unclear whether some people with negative test results would also have treatment. The committee concluded that research was needed to better understand how the test results affect treatment decisions (see section 5.1).
4.6 The committee noted that, based on the evidence, it was not possible to estimate the effectiveness of the different bile acid sequestrants for treating bile acid diarrhoea. No evidence was found on the long-term effects of the bile acid sequestrants. It was unclear whether they have a sustained effect on bile acid diarrhoea and if they have any negative effects such as reducing vitamin absorption. The patient experts highlighted how important it is to better understand the tolerability of different bile acid sequestrants. The committee noted that many studies reported high rates of treatment discontinuation, but it was not clear whether the rates were the same for the different bile acid sequestrants. The evidence described outcomes only for people who had a positive SeHCAT test result. The committee recognised that although the SeHCAT test is safe, the potential benefits and risks of testing for people who have a negative test result are not clear. The committee concluded that further research is needed to assess tolerability and effectiveness of the treatment options for bile acid diarrhoea (see section 5.3). It further concluded that to fully understand the benefits and risks of the SeHCAT test, evidence from people with a negative test result is needed (see section 5.2).
4.8 The EAG assumed that a threshold of 15% retention of SeHCAT would be used to define a positive test result in its model. The committee discussed whether this threshold was appropriate. The evidence was too limited to estimate how bile acid sequestrants might benefit people who have a positive test result at different SeHCAT thresholds. But it noted that in most studies a 15% threshold was used to define a positive SeHCAT test result. The clinical experts explained that the threshold used in practice varies. Treatment may also work for people with a positive test at higher thresholds, but 15% was a widely used and accepted threshold. The clinical experts noted that the 15% threshold was also supported by 2 recent surveys of SeHCAT use. The committee concluded that although the evidence did not allow the optimal threshold to be explored, it was reasonable to assume a 15% threshold in the model.
4.9 Since NICE published the original guidance in 2012, the place of SeHCAT in the care pathway has changed. The EAG included the possibility of having a colonoscopy after SeHCAT testing in the model for people with suspected or diagnosed IBS‑D or functional diarrhoea. The committee considered whether this reflected current practice and whether using SeHCAT could help reduce the number of colonoscopies done. The clinical experts explained that most colonoscopies are avoided because blood and stool tests are used to exclude inflammation before a SeHCAT test (see section 4.4). They noted that clinical practice varies but variation in the timing of colonoscopies (before or after SeHCAT testing) was adequately reflected in the modelling. The committee noted that the cost of colonoscopy was included in the model. It concluded that the resource impact of preventing colonoscopies was adequately captured. The committee questioned whether the model should have also included disutility associated with colonoscopy. The EAG explained that the disutility was not included because of a lack of data. But it noted that the model assumes colonoscopy occurs only in the 6 months immediately after a SeHCAT test. So, the committee concluded that it was unlikely that including a disutility for colonoscopy would change the overall conclusions of the cost-effectiveness analyses.
4.10 The committee discussed whether all the relevant costs involved in providing SeHCAT testing had been included in the model. The clinical experts explained that because the level of radioactivity in SeHCAT is low, other investigations using radioactivity could interfere with the test results. So, nuclear medicine departments cannot do other investigations when doing SeHCAT testing. The committee concluded that the cost of providing SeHCAT testing was unlikely to be fully captured in the model.
4.11 The long-term Markov model included a health state for people who have diarrhoea (because the treatment did not work) and a health state for people who do not have diarrhoea (because the treatment worked). The committee recalled its discussion on the severity of bile acid diarrhoea and how this could affect quality of life (see section 4.8). It concluded that the model did not capture the effects of variable diarrhoea severity and treatment response.
4.12 The committee discussed whether it was appropriate to assume that response to bile acid sequestrants in the trial of treatment strategy was lower than in the SeHCAT strategy. This assumption was based on expert opinion. It recalled that the clinical experts highlighted that they were not confident of their answers to the questionnaire used to obtain values for the model. Assuming a lower probability of treatment response in the trial of treatment strategy would bias the model results towards SeHCAT and make it appear more cost effective. The committee concluded that using expert opinion without accounting for the discrepancies in treatment response for each of the strategies affected the internal validity of the model. As a result, this affected the comparison between the modelled strategies.
4.13 The committee considered the assumptions used in the economic model. It noted that, because of the lack of clinical outcome data, most of the model inputs were estimated based on expert opinion from a small group of clinicians. The clinical experts explained that they were not confident that their estimates captured the variability of bile acid diarrhoea seen in practice. The committee was not certain that the analyses presented had fully quantified the uncertainty caused by the lack of clinical outcome data that links the results of testing to treatment outcomes. It concluded that, in the absence of key clinical outcome data, the results of the economic model cannot be used to inform recommendations about whether SeHCAT can be adopted.
4.14 The committee considered that there is an unmet clinical need for a test to diagnose bile acid diarrhoea. It recognised the value of having a diagnosis and access to treatment, but acknowledged that the evidence to support both the use of the test and the treatment is highly uncertain. So, the full benefits and potential harms of widespread use of SeHCAT testing cannot be reliably quantified. The committee recognised that the British Society of Gastroenterology clinical guidelines for the investigation of chronic diarrhoea in adults recommend the use of SeHCAT testing. It considered that this NICE assessment aimed to evaluate the test as well as its link to longer-term clinical outcomes to show whether it was cost effective. There is no robust data on this link and so the clinical utility of SeHCAT testing, that is:
how well it predicts response to treatment
how it influences clinical decision making
the longer-term clinical outcomes with treatment.
Without this, SeHCAT's cost effectiveness cannot be adequately assessed. The committee concluded that it was unable to recommend the routine adoption of SeHCAT testing. It strongly encouraged further data collection and research to address the limitations in the evidence.
4.15 There is a lack of robust data on the link between the SeHCAT test and longer-term clinical outcomes preventing the test from being recommended for routine adoption in the NHS (see section 4.14). Because of the unmet clinical need for a test to diagnose bile acid diarrhoea, the committee emphasised that there is an urgent need for further data collection and research to:
support future evaluations of the test and
improve outcomes for people with bile acid diarrhoea.
The committee's recommendations for further data collection and research are described in section 5.