4 Committee discussion
- The clinical care pathway is complex
- Balloon tamponade is an appropriate comparator
- The evidence shows that Danis stent improves short-term clinical outcomes
- The evidence does not reflect the potential use of Danis stent in the UK care pathway
- Side effects and adverse events
- Outcome measures
- NHS considerations
- Other patient benefits or issues
- Cost-modelling overview
4.1 The clinical experts explained that oesophageal variceal bleeding is an acute clinical emergency and clinical care is managed on a case-by-case basis. First-line treatment is an endoscopy followed by band ligation. However, in rare cases, balloon tamponade may be done to control the bleeding before an endoscopy. If first-line therapy fails, balloon tamponade or Danis stent is used as a bridging therapy to stabilise the patient before a definitive treatment can be done, such as transjugular intrahepatic portosystemic shunt (TIPS) insertion.
4.2 People who cannot have definitive treatment are given palliative care. Clinical experts explained that Danis stent can be used to control bleeding as a component of palliative care after all other lines of treatment have failed or if the patient cannot have definitive treatment with TIPS or transplant surgery. The committee concluded that the care pathway is complex, and that practice varies depending on the individual circumstances of each patient.
4.3 The committee noted that TIPS was included as a comparator in the scope and that in the randomised controlled trial TIPS was done within 72 hours of presentation. The experts explained that when TIPS is done this early it could be considered a comparator to Danis stent. However, the experts explained that it usually takes between 5 days and 7 days to deliver TIPS in the UK and that Danis stent would be used before this timepoint. The committee concluded therefore that, in the UK NHS setting, balloon tamponade is the best comparator for Danis stent.
4.4 The comparative evidence reported that Danis stent improves control of bleeding for patients in the short term (15 days). The committee recognised there were some key limitations in the evidence. For example, the population included in the retrospective case-controlled comparator study (Maiwall et al. 2018) was limited to patients with acute-on-chronic liver failure, and the randomised controlled trial (Escorsell et al. 2016) was underpowered and was not done in the UK. The committee considered the randomised controlled trial to be the most robust evidence for Danis stent and recognised the difficulties in doing controlled studies and generating evidence in this patient population. While the committee acknowledged the limitations in the studies it concluded that on balance the evidence shows that Danis stent improves short-term clinical outcomes.
4.5 The randomised controlled trial (Escorsell et al. 2016) was done in Spain and the clinical experts advised that TIPS was more accessible in this trial than it would be in the UK. They explained that TIPS procedures are arranged and done in regional tertiary centres in the UK and that this procedure may only be suitable for people with less severe liver disease (Child–Pugh score A). In contrast, the experts explained that in the randomised controlled trial patients with more severe liver disease (Child–Pugh score B or C) were given a TIPS procedure, and were given it at an earlier timepoint than in the UK. The committee concluded that the protocol used in the randomised controlled trial does not accurately reflect UK practice.
4.6 Stent migration is reported to happen in 20% of cases. Clinical experts explained that, in their experience, this figure is likely to be an overestimation and that stent migration happens rarely if operators are fully trained in using the device. They advised that there is a small risk of lung aspiration with the procedure and that the patient's condition should initially be managed in an intensive care setting after stent insertion. The risk of oesophageal perforation is higher in patients who have had balloon tamponade first. The committee concluded from the evidence and expert advice that using Danis stent does not increase the risk of an adverse event.
4.7 The evidence is limited to a follow-up time of 6 weeks or less. The committee recognised that some patients who have definitive treatment will live beyond 6 weeks, however it noted that the clinical evidence did not report a significant difference in mortality at 6 weeks between the Danis stent group and the balloon tamponade group. The committee also understood that people with oesophageal variceal bleeding have other comorbidities that are likely to affect survival beyond 6 weeks. So it concluded that the time horizon to definitive treatment was appropriate for the cost modelling.
4.8 Danis stent can be left in place for up to 7 days compared with a balloon tamponade, which needs to be removed after 24 hours. The clinical experts commented that, in secondary or tertiary care settings, this additional time allows healthcare professionals the time to stabilise and monitor patients and arrive at a carefully considered clinical decision about the next stage of treatment. They highlighted that, when patients need to be transferred to a tertiary care centre for definitive treatment, using Danis stent can increase patient safety during the transfer. The committee recognised that there are limitations in the evidence and accepted expert advice about the additional patient and system benefits.
4.9 Healthcare professionals are trained to use Danis stent. Clinical experts stated that training is straightforward for healthcare professionals with experience of endoscopic procedures and it is also possible to gain the necessary skills even without this experience. They described how maintaining clinical competence in a large team is challenging because of the limited number of patients needing this procedure each year, which necessitates regular reskilling. The committee considered that a lack of clinical confidence in using Danis stent during a medical emergency situation might serve as a barrier for adoption and suggested that the company should make sure centres that use the device have access to training and reskilling support.
4.10 The clinical experts advised that, for a small proportion of the patient cohort, estimated at between 5% and 6%, further definitive treatment may not be appropriate. The clinical experts advised that Danis stent can be used if all previous lines of therapy have failed, and in patients for whom either transplant surgery or a TIPS procedure is not suitable. After the stent is inserted, patients can be extubated, moved off a high dependency ward and managed in a more comfortable and less intensive environment where interaction with family and friends is more possible. The clinical experts explained that when used in this way Danis stent can stay in place for at least 7 days and even longer. This scenario was not included in the cost modelling, however the committee recognised that data collection is unrealistic in this population. It concluded that, in the proportion of people for whom definitive therapy is not appropriate, using Danis stent may offer substantial patient benefits in alleviating suffering and allowing compassionate care.
4.11 The EAC base-case cost model used clinical parameters based on the randomised controlled trial (Escorsell et al. 2016) and case series data. The committee received the following expert advice:
definitive treatment is decided on a case-by-case basis in the NHS and is not affected by the choice of bridging treatment
a second endoscopy is needed in the balloon tamponade group, which was not included in the base-case model
use of multiple healthcare resource group costs could result in an overestimation of procedure costs
the incidence of hepatic encephalopathy should not differ between arms.
The committee concluded that the assumptions used in the base case did not accurately reflect the cost of using Danis stent in the UK.
4.12 The EAC developed scenarios 5A and 5B based on clinical advice so that the cost model better reflected UK practice. Scenario 5A reported that using Danis stent was cost saving by £2,423. Scenario 5B included the cost of transferring a proportion of people from secondary to tertiary care, although this cost had little effect on the results. This scenario reported Danis stent to be cost saving by £2,426. The main cost drivers of these scenarios were the risk of re-bleeding, the procedure costs and the estimated length of intensive care unit (ICU) stay. The committee recognised that there were uncertainties in the parameters because they were primarily based on expert advice, but concluded that scenarios 5A and 5B were the most appropriate models for estimating the cost of Danis stent in the NHS.
4.13 The committee noted that the estimated difference in length of ICU stay had the greatest effect on the direction of the cost case results. Clinical experts estimated that length of ICU stay for the Danis stent group is 3.6 days, and 6 days for the balloon tamponade group. The EAC did a threshold analysis for this parameter and reported that Danis stent would be cost neutral or cost saving when the balloon tamponade group had an increased length of ICU stay of 0.6 days or more compared with the Danis stent group. The committee accepted the expert clinical advice that the clinical effectiveness of Danis stent is likely to affect the length of ICU stay and concluded that in all probability, Danis stent will reduce time in ICU by more than approximately 1 day (that is, more that 0.6 days) in UK clinical practice. The committee concluded that Danis stent is very likely to be cost saving compared with balloon tamponade in people being considered for definitive treatment for oesophageal varices.
4.14 The committee recognised that there are uncertainties in the cost case because of the limited information available about resource use. The committee noted that a planned multicentre UK randomised controlled trial was not completed because of difficulties with recruitment. It recognised that generating evidence is challenging in this small and heterogenous population and accepted that expert advice was an appropriate alternative to definitive UK evidence. Further data collection is welcomed by the committee to inform more accurate assessments of the cost savings associated with using Danis stent in the future.