3.1 The clinical evidence comprises 3 studies reported in 4 publications including a total of 341 adults in cardiac or general intensive care. One of these studies was a randomised controlled trial comparing PneuX with a standard endotracheal tube (ETT) without drainage (Gopal et al. 2014). The other studies were non-comparative (Smith et al. 2014, Doyle et al. 2011 and Hodd et al. 2009). For full details of the clinical evidence, see section 3 of the assessment report, which is in the supporting documents for this guidance.
3.2 The Gopal et al. (2014) study recruited patients who needed ventilation during and after cardiac surgery. Patients were randomly selected to be ventilated using PneuX or a standard ETT without subglottic drainage. All patients in the study were classified as high risk (over 70 or with a left ventricular ejection fraction of under 50%, or both) and were therefore considered more likely to contract ventilator-associated pneumonia (VAP). People in this study were ventilated for a relatively short period of time, 15 hours and 13 hours (median) in the treatment and control groups respectively.
The other 2 studies include people needing ventilation with a wider range of health conditions but do not compare PneuX to any other ETT
3.3 The evidence from the non-comparative studies is more generalisable to people needing ventilation with a wider range of health conditions. Nonetheless, the lack of a control group makes it difficult to draw any conclusions about the efficacy of PneuX. There was also wide variation in the outcomes measured in these studies (for example, mortality was 1.6% to 35.8% and unplanned tube removal 0.1% to 17%). However, the rates of VAP and unplanned tube removal are very low in these studies. All 3 studies used different definitions for diagnosing VAP.
3.4 The company identified 2 relevant studies, Andronis et al. (2018) and NHS Innovation Accelerator (2017). No additional economic analyses were identified by the external assessment centre (EAC). Both the studies compared PneuX with standard ETTs and were carried out in the UK.
The company's economic model compares PneuX to ETT without subglottic drainage in a cardiac surgery population
3.5 The company model uses a simple decision tree structure based on the model published in Andronis et al. (2018; see figure 2 of the assessment report in the supporting documents for this guidance). The population modelled is adult patients requiring mechanical ventilation following major heart surgery. The model compares PneuX with conventional ETT without subglottic secretion drainage. The key clinical parameter used in the model is the risk of VAP as reported in the comparative study Gopal et al. (2014), which was 10.8% for PneuX and 20.8% for ETT without subglottic drainage. For full details of the cost evidence, see section 4 of the assessment report.
The model is appropriate for people who have had cardiac surgery but may not be generalisable to all people
3.6 The EAC considered the simple model structure to be adequate to capture the costs and consequences of the technology and did not make any changes. It said that all assumptions were acceptable except for the generalisability of the results from people who had cardiac surgery to a broader population of patients for whom PneuX is intended. The EAC also said the costs of treating VAP may not be generalisable to a wider population given the shorter stay in intensive care in the cardiac surgical studies (Gopal et al. 2014 and Luckraz et al. 2018) compared with all people who might need ventilation.
3.7 The results of the company model indicate a cost saving of £738 per patient after cardiac surgery when PneuX is used instead of an ETT without subglottic drainage. This saving is from an absolute reduction in the risk of VAP of around 10% for PneuX and the associated reduction in resource consumption based on avoided costs of around £9,000 per VAP prevented. In the model, the expected cost of needing to treat VAP is around £900 less for patients given PneuX than for those having ETTs without subglottic drainage. This cost saving is substantially greater than the additional cost of using PneuX instead of ETT without subglottic drainage (PneuX costs £150 and ETT without drainage £5).
3.8 The company did scenario analyses by varying 3 parameters:
reduction in baseline risk of VAP from 20.8% to 10%
reduction in cost of standard ETT from £5 to £1.12
inclusion of a training cost to use PneuX of £10 per patient.
PneuX remained cost saving for all 3 scenarios. The company also reported:
a one-way sensitivity analysis of the cost of treating VAP
a two-way sensitivity analysis of the baseline risk of VAP (0% to 50%)
the relative risk of VAP with PneuX (0 to 1)
a probabilistic sensitivity analysis to characterise the impact of uncertainty in the model parameters.
The one-way sensitivity analysis indicated that PneuX is cost saving even if the cost of treating VAP is as low as £4,000. In the two-way analysis, PneuX remained cost saving for most combinations of the 2 parameters. The probabilistic sensitivity analysis indicated that there is a 96% likelihood that PneuX is cost saving compared with ETT without subglottic drainage.
Additional analysis by the EAC shows PneuX may be slightly cost saving compared with an ETT with subglottic drainage
3.9 There are other ETTs with subglottic drainage but there are no trials available that directly compare these with PneuX. There was only 1 other study that compared an ETT with subglottic drainage (Portex Blue Line, Smiths Medical) with an ETT with no drainage (Jena et al. 2016). The EAC did an additional cost analysis using results from the Gopal and Jena studies to indirectly compare PneuX and Portex ETTs. Portex Blue Line costs less than PneuX (£20, compared with £150 for PneuX), and the relative risk reduction of VAP in the 2 studies in question was 0.52 for PneuX and 0.60 for Portex Blue Line. This led to a slight cost saving for PneuX of £18. The EAC cautioned that the relative risk of VAP for Portex Blue Line came from a very small trial, the results of which were not statistically significant, although they were consistent with data from a large meta-analysis (Mao et al. 2016).