FibroScan for assessing liver fibrosis and cirrhosis outside secondary and specialist care

3.2 Patient experts reported that people often travel long distances to access FibroScan, especially in rural areas. Easier access to the test could reduce time and costs associated with this. It could help people attend the test in a more familiar environment. It could also help people with disabilities that make it difficult to travel. Patient experts commented that needing to travel longer distances could be a particular barrier for people from lower socioeconomic groups, who may be at higher risk of liver disease and typically die from the condition much earlier. The committee commented that the benefits outlined may not be seen if multiple appointments are needed to first do the scan and then separately deliver lifestyle advice. Clinical experts responded that lifestyle interventions are often delivered by healthcare assistants or nurses, and that any advice needed based on a FibroScan result would be given in the same appointment as the scan was done (see section 3.12). Clinical experts further commented that the increasing prevalence of liver disease means that secondary care services risk being overwhelmed, and that moving some aspects of care like FibroScan testing to alternative settings could help manage the workload. The extent to which making FibroScan available outside of secondary and specialist care would improve access to testing depends on which locations testing is made available, and what transport links are available (compared with transport links to secondary or specialist care).

3.3 There was no evidence comparing the performance of FibroScan for measuring liver fibrosis when it is used in alternative settings with when it is used in secondary or specialist care. At consultation on the draft guidance, the lack of published evidence was confirmed by the company.

3.4 Clinical experts explained that how well FibroScan works depends on the experience of the user. They stated that if FibroScan is used often enough to make sure it is being used correctly, performance between different care settings would be comparable.

3.5 The committee considered the level of use that would be needed for users to maintain competence with FibroScan. The company commented that it encouraged users to make sure that competency is validated in practice, but that it does not currently provide guidance on requirements for the level of use. Clinical experts highlighted that there is no independent accreditation scheme for users, and that this is also the case for tests done in secondary or specialist care. They explained that FibroScan users outside secondary or specialist care in their areas had close links with local hepatology departments which could provide support when needed. The company explained that pilot schemes in primary care networks typically saw 20 to 30 people a month. The committee noted that it is unclear how many FibroScan tests are currently done in the NHS (see section 3.11). Clinical experts highlighted that there is no clear evidence to define a number or frequency of tests that need to be done to achieve and maintain expertise. The committee considered that sufficient levels of use may not be achieved if the test was available in individual GP practice populations, but use in services that cover larger populations, such as community diagnostic hubs or across a primary care network, would likely mean the users do enough tests to be sure it is being used correctly. The committee concluded that if used outside secondary or specialist care, it would be important to make sure that operators used FibroScan often enough to be able to accurately use the test, and for centres to consider having an accreditation framework in place.

3.6 Clinical experts commented that FibroScan is relatively simple to use, that it indicates if the test has not worked, and that all grades of staff can use the technology if appropriately trained. At consultation, the company proposed several measures they could introduce to make sure that user competency is maintained after the initial training. These included developing a competency checklist and framework for annual assessment, offering on-site assessment, developing online competency assessments, or getting continuing professional development accreditation for FibroScan training. The committee agreed these would be valuable and would build confidence in test results. The committee concluded that if these measures were put in place, it would give reassurance that FibroScan assessment done outside secondary or specialist care would be done effectively.

3.7 The committee recalled that there was no data directly comparing the performance of FibroScan tests done in, or outside, secondary or specialist care (see section 3.3). But, if the test was done in an alternative setting where appropriately trained operators do enough scans to maintain their expertise (see section 3.5), the committee concluded that it was likely that test performance could be maintained outside of secondary or specialist care, if there are ongoing measures to ensure quality such as those proposed by the company (see section 3.6).

3.8 The committee recalled that the population in this assessment was restricted to those who would have FibroScan as in current NHS practice (see section 2.10). The test was only assessed for use in people it is already recommended for. It noted that performance of the test would depend on the population being tested, and that the value of testing would depend on the availability and effectiveness of interventions for the population tested, based on test results. Some consultation comments mentioned a potential benefit of FibroScan outside secondary or specialist care may be that it allows for wider screening for early liver disease. The committee noted that such use had not been assessed in this guidance and expressed concern that using FibroScan outside secondary or specialist care could lead to its use in a wider population than assessed, which could in turn affect its performance. It concluded that if recommended, the test should only be used as recommended in national guidelines (see sections 2.3 to 2.5).

3.9 Clinical experts and committee members emphasised that clear guidance on what to do with the results of FibroScan is vital, particularly if testing is done outside a specialist setting. FibroScan done in alternative settings could reduce the number of unnecessary referrals to hepatology services. But, if there is uncertainty about what to do based on a result, a referral to specialist services, or contact with these services to ask advice, may still be made. Clinical experts highlighted that this could happen often if multiple conflicting test results (including FibroScan) were available. Liver pathways should be designed in agreement with primary and secondary care centres, and incorporate all tests used for detection and characterisation of liver disease, not just FibroScan. The committee concluded that establishing clear care pathways, with guidance for healthcare professionals on what to do based on a FibroScan result based on existing national guidelines, would be essential to ensure appropriate clinical management of liver disease in people who have FibroScan tests done outside secondary and specialist care settings.

3.10 In the base-case analysis provided by the company, the economic model used a 1‑year time horizon. The committee commented that this omits potential costs or cost savings that would only appear many years after testing, such as the costs of treating previously undetected liver disease. The committee noted that increased attendance at FibroScan appointments in primary or community care increased costs in the model, because more people were referred for follow-up appointments in hepatology. But, any potential cost savings or health benefits of greater detection of liver disease were not considered (see section 3.1). At consultation on the draft guidance, the company submitted a scenario analysis with a 5‑year time horizon, which estimated lower long-term costs of about £30 less per person if testing was done in primary or community care. The external assessment group (EAG) explained that the lower cost was because there were fewer people with missed liver disease if testing was done in primary or community care, because more people attended scans. The committee considered it was unclear what assumptions were made in the modelling to base this on. Company representatives were not able to provide further clarity in the committee meeting. The company's model did not allow people's liver disease to progress in the 5‑year time period modelled. Clinical experts commented that this may not be appropriate for people with alcohol-related liver disease, whose condition can progress at a faster rate. The committee noted that the effect of lifestyle advice may differ depending on who provides it, for example a GP compared with a liver specialist, but experts said that there was no evidence on this. Clinical experts commented that referrals to hepatology services may increase after adopting FibroScan outside secondary and specialist care, but this may mean that more people who would benefit from specialist care are able to access it. Clinical experts also commented there was uncertainty about the long-term effect of using the test outside secondary or specialist care, for example, on levels of hospitalisation. The committee considered it plausible that testing in alternative settings could lead to longer term cost savings but thought that the company analysis did not allow this to be assessed. In advance of the third committee meeting, the company provided a revised model, and accompanying description, of the long-term implications of missing liver disease. This included allowing liver disease to progress within the modelled time period. This led to lower costs if FibroScan was done in primary or community care because increased attendance at scans was assumed to increase detection of liver disease and reduce progression to more severe stages. The EAG questioned the long-term costs used in the model because they came from a study of antiviral treatment for people with chronic hepatitis C (Wright et al. 2006). It suggested a study in which costs were related to managing non-alcoholic fatty liver disease (Tanajewski et al. 2017) as an alternative source. Some of the results from the updated model provided by the company for the third committee meeting, and further analyses run by the EAG using this model, did indicate that testing in primary or community care reduced long-term costs. Clinical experts said that earlier detection of liver disease could plausibly lead to cost savings. But, the committee also considered that costs could be higher in the long term (although potentially with accompanying improvements in health-related quality of life), particularly if a time frame longer than the 5 years modelled was used. The committee concluded that there is considerable uncertainty about the long-term effect of FibroScan testing outside secondary or specialist care on costs.

3.11 The committee discussed the costs used in the original model submitted by the company, and the revised costs used by the EAG. The EAG removed a cost from the company's model for staff time to do and evaluate FibroScan in secondary or specialist care because this time was already incorporated within an existing cost used in the model. This meant that, using the figure proposed by the company for testing in this setting, the cost of doing FibroScan was greater per scan when done in primary or community care. Experts agreed that the staff costs of doing the scan would be included in the Health Resource Group (HRG) cost used by the company. The company used HRG bundled costs for ultrasound elastography to estimate the cost of FibroScan in secondary or specialist care, at £43.93 in the base case. This cost was also used by the EAG. The company highlighted that a scenario analysis done by the EAG in which a higher cost per use in secondary or specialist care (£61.98) was used, based on a weighted average of 2 different costs attributed to the HRG code, and suggested that this might be more appropriate. The EAG commented that the results of this scenario still indicated that using FibroScan outside secondary or specialist care was cost incurring. In their report, the EAG highlighted difficulties in evaluating the costs of doing FibroScan in the different settings that were a consequence of comparing a bundled HRG cost from secondary care with a cost obtained by micro-costing in a non-hospital setting, where a HRG code does not currently exist. The committee noted that the HRG code for ultrasound elastography was used only 3,561 times for outpatients in 2019 to 2020, which likely underestimated the number of FibroScan tests done in the NHS. Further scans may be done during outpatient appointments and recorded using other HRG codes, potentially at higher cost. At consultation, the company provided further analyses. Its base-case analysis kept the higher cost of testing in secondary care, including additional costs for staff time to do the test as well as the HRG code. Analyses using alternative costs were not cost neutral or cost saving for testing done in primary or community care. The company did not provide any further support for their choice of cost used in the base case or rationale for the most appropriate choice of cost for the test in secondary care. The committee also questioned whether the full costs of a referral for testing in secondary or specialist care had been incorporated. Missed appointments were included as a separate cost in the model. A clinical expert commented that the cost of missed appointments was likely to already be captured in the cost of doing scans used in the company's model. If so, including an additional cost for missed appointments was not appropriate. Clinical experts noted that if a person misses an appointment in secondary care, they may need to restart the referral pathway to access FibroScan, incurring further cost. The committee concluded that there was still considerable uncertainty about the costs of testing in secondary care, and suggested further analysis to address this. In advance of the third committee meeting, the company provided further analyses. This included a micro-costing-based estimate of £40.61 for doing FibroScan in secondary care. The number of scans (610) used to determine this was from a survey of 4 NHS trusts. The EAG noted some limitations in the company's micro-costing approach but stated this was its preferred method for assessing costs. Clinical experts noticed that the company's micro-costing only included costs of doing FibroScan but not the costs of a referral for a hepatologist outpatient appointment that would happen in practice if a GP decided that the scan was needed. The EAG noted that the NHS reference cost for this appointment is £268 (cost in individual trusts may vary). The committee concluded that while there is uncertainty about the exact cost of testing, it is likely that the model underestimates the cost of doing FibroScan in secondary care.

3.12 The committee noted that the cost the company has provided for FibroScan in primary or community care in their original submission is higher (£58.00 per scan, plus £10.50 for staff time to do the test and evaluate FibroScan result) than the HRG code cost used in the EAG's base case and scenario analysis for FibroScan in secondary or specialist care (see section 3.11). This was based on a fixed cost being charged by the company per scan, with no upfront cost for the machine. At consultation, the company submitted an alternative costing model in which the FibroScan device was purchased outright, which included a maintenance contract over the assumed 7‑year lifespan of the device. The average cost per scan, calculated assuming 500 scans per year being done based on Southampton clinical commissioning group use, was £34.29 plus staff time to do the test. The EAG did a threshold analysis and found that the device would have to be used at least 300 times a year for this model to be cheaper than the pay-per-scan model originally suggested. The company stated that their intended use of the tests outside secondary or specialist care is in hubs and diagnostic centres, rather than single GP practices, where use would be expected to be higher. The committee agreed that this usage may be achieved if the device was used in primary care networks or community diagnostic hubs (see section 3.7). But, it noted that only a single estimate of expected use in primary or community care had been provided by the company. The committee recalled that moving FibroScan testing outside secondary and specialist care would potentially move workload to other settings for activities that happen based on test results, such as lifestyle advice, and questioned whether the time taken by healthcare professionals to do this has been adequately captured in costs of doing the test outside secondary and specialist care. They further highlighted that even if a person is not referred to a specialist service after a test done outside this setting, advice from staff in these services may be sought. A clinical expert emphasised that community and primary care staff such as nurses and healthcare assistants are experienced in providing lifestyle and diet advice (see section 3.2) and that any advice could be given in the same appointment as the FibroScan test was done. The committee concluded that there was uncertainty about whether the costs of doing FibroScan outside secondary and community care used in the company's model were an accurate reflection of the true cost of testing. It further noted that if buying the FibroScan device outright, the cost per use would depend on the extent of use, and asked for further information to support estimates of expected use. In advance of the third committee meeting, the company provided further analysis. Using local real-world data and national data sources, the company estimated that 1 FibroScan device shared between 5 primary care networks would be used for 2,500 to 5,000 scans per year. The EAG considered the estimates based on real-world data more robust but stated that using 6 sources of information provided by the company, the EAG found only 1 example where FibroScan was used in as many as 500 to 1,000 people per year per primary care network. But of the 8 clinical experts consulted by the EAG, 5 said sharing 1 device between 5 primary care networks was plausible in some scenarios and all thought a single network would be able to do 500 scans per year. The clinical experts attending the committee meeting supported this view. The committee noted that in its updated submissions, the company had provided the cost per FibroScan done in primary care based on buying the device outright and at least 500 scans per device being done per year (£44.79), rather than the cost per scan based on a pay-per-scan charging model as in its original submission (£58.00 per scan, plus £10.50 for staff time).

3.13 There is still uncertainty about the true cost of doing a test both in secondary and specialist care (see section 3.11) or outside these settings (see section 3.12). The committee recalled that it is likely that the model underestimated the cost of testing in secondary care (see section 3.11). Higher cost of testing in this setting would make testing outside of secondary and specialist care more likely to be cost saving. The committee concluded that, based on buying FibroScan 430 Mini+ outright (see section 2.11) and an expected use of at least 500 scans per year per device as modelled by the company, the immediate costs related to a FibroScan test were likely to be lower outside of secondary and specialist care. The committee also recalled that making sure FibroScan was used enough outside secondary and specialist care was important to make sure operators do enough scans to maintain their expertise (see section 3.7).The committee further recalled that there is considerable uncertainty about the long-term effect on costs of using the test outside secondary and specialist care (see section 3.10). On balance, the committee concluded that there was enough certainty that the immediate costs of using FibroScan for assessing liver fibrosis and cirrhosis outside secondary and specialist care are likely to be lower than the cost of referring people for testing in secondary or specialist care to allow it to recommend use in this setting.

3.14 The committee commented that it would be beneficial to monitor the effect of greater availability of FibroScan outside secondary and specialist care on relevant costs and outcomes to make sure that the proposed benefits are being achieved in practice in the NHS.