5.1 The company presented 2 published economic studies, both of which compared GreenLight XPS with transurethral resection of the prostate (TURP; Thomas et al. 2015a and Benejam-Gual et al. 2014). The external assessment centre did not identify any further studies.
5.2 Thomas et al. (2015a) included 1-year data from the GOLIATH study (Bachmann et al. 2014). The authors used a Markov model with a lifetime horizon to estimate quality-adjusted life years (QALYs) gained and costs from a UK NHS perspective. The main cost driver was found to be the proportion of treatments done as day-case procedures. Using GOLIATH data, sensitivity analyses showed the costs were almost equal for GreenLight XPS and TURP. If more than 32% of patients had day-case procedures with GreenLight XPS, it became cost saving when compared with TURP. However, the external assessment centre found some uncertainties with the risk ratios and day-case rates used in the model. It also highlighted that the capital costs of the GreenLight XPS equipment were not included, only the cost of the fibres. Therefore, the findings apply to the current funding arrangements, where the NHS incurs no capital costs in adopting the technology.
5.3 Benejam-Gual et al. (2014) used retrospective data from 79 patients in 4 centres in Spain to estimate the direct costs of procedures and complications over 3 months. The external assessment centre identified uncertainties with how the resource costs were collected and noted differences in lengths of stay at different hospitals. However, it agreed with the conclusions that GreenLight XPS is associated with shorter lengths of stay than TURP and may therefore be cost saving.
5.4 The company presented 2 de novo cost model analyses comparing the cost consequences of using GreenLight XPS in different populations with different comparators:
A primary analysis compared GreenLight XPS with monopolar/bipolar TURP in a non-high-risk BPH population (patients without urinary retention, not taking anticoagulation therapy or with prostates less than 100 ml).
A secondary analysis compared GreenLight XPS with HoLEP in a high-risk BPH population (patients with urinary retention, taking anticoagulation therapy or with prostates more than 100 ml).
5.5 Both cost models used the same decision-tree structure, in which patients entered the model at the point of having surgery (either GreenLight XPS or monopolar/bipolar TURP or HoLEP) and were then routed through 4 potential pathways. The post-treatment pathways included options for discharge on the day of surgery or after an in-patient stay, as well as the potential to develop new symptoms or post-surgical complications (which may or may not lead to readmission). At the end point of the model, patients were either asymptomatic or continued to have symptoms. The model was constructed from an NHS perspective with a 6‑month time horizon and a discount rate of 3.5% on the capital costs.
5.6 In the primary (non-high-risk) model, the clinical outcome parameters used included IPSS score, the probability of being complication-free and the proportion of adverse events, all derived from the 6-month GOLIATH study (Bachmann et al. 2014). Mean excess bed days were 10.36 days for GreenLight XPS and 10.65 for TURP, sourced from 2014/15 NHS Hospital Episode Statistics (HES). The company model allowed 4 different day-case discharge rates for GreenLight XPS informed by different sources: 35.96% from HES data; 80% from a single UK hospital specialising in GreenLight; 57.71% from French health service data and 71.5% from the US Medicare population. The company used the HES day-case discharge rate of 4.08% for TURP. Resource costs included hospital resource costs (procedure costs, cost per day of hospital stay, excess bed days) and the costs of treating adverse events (acute [classed as grade 3, treated in hospital] or non-acute [classed as grade 2, treated in primary care]), which were derived from national tariffs and NHS reference costs.
5.7 For the secondary (high-risk patients) model, the only additional clinical outcome parameter included was a 1.5% additional risk of bleeding (Woo et al. 2008) for both GreenLight XPS and HoLEP. All other clinical parameters were the same as in the primary model, which assumed that GreenLight XPS and HoLEP have the same day-case rates and efficacy and safety outcomes. The company justified this on the basis that there were no head-to-head clinical trial data comparing GreenLight XPS with HoLEP. In the absence of any UK-specific data for HoLEP, the company assumed that the same HES day-case rate (35.96%) could be used for both GreenLight XPS and HoLEP.
5.8 The company calculated the equipment costs for each technology from internal sales data (GreenLight XPS and HoLEP) and expert opinion (TURP). No capital costs were calculated for TURP or GreenLight XPS: the TURP device was assumed to be already present in NHS hospitals and the GreenLight XPS console can be provided on loan if a minimum number of consumable laser fibres are purchased at £550.00 each. For TURP consumables, it was assumed that 50% of procedures were monopolar and 50% were bipolar, with an average cost of £190.50 per surgery. HoLEP capital costs of the laser and morcellator were included, based on a 5-year lifespan and treating 25 high-risk patients per year. HoLEP consumables were assumed to be 50% multi-use and 50% single-use laser fibres (plus a fibre stripper and cleaver for multi-use only), in addition to a morcellator blade, suction tubing, omni-jugs and Ellik evacuator for all procedures. Maintenance and training costs were assumed to be zero for GreenLight XPS and were not considered for the other comparators.
5.9 The results of the company's primary analysis in non-high-risk patients found that when applying day-case discharge rates of between 36% and 80%, GreenLight XPS was associated with cost savings of between £29 and £443 per patient when compared with TURP.
5.10 The company performed a deterministic sensitivity analysis in which clinical costs were varied by upper and lower limits of the 95% distribution and other costs were varied by 20% in each direction. The analysis determined that when the lowest day-case rate was applied, the most sensitive cost drivers were inpatient procedure costs, GreenLight XPS consumable costs and day-case procedure costs. Varying these costs resulted in GreenLight XPS being slightly more or less costly than TURP, which led the company to conclude that GreenLight XPS is cost neutral when compared with TURP.
5.11 The results of the company's secondary analysis in high-risk patients showed that GreenLight XPS produced cost savings of between £591 and £1,059 per patient compared with HoLEP. Because all clinical parameters were assumed to be the same between treatments, the key driver in the high-risk model was the capital cost of HoLEP (compared with zero capital costs for GreenLight XPS).
5.12 NICE has published a resource impact report on GreenLight XPS. Assuming that around 6,800 people have GreenLight XPS, the estimated annual cost saving across the NHS in England ranges from £1.3 million when 36% of GreenLight XPS cases are done as day cases to £3.2 million when 70% are day cases.
5.13 The external assessment centre considered the economic evidence for the company's primary analysis in non-high-risk patients to be robust. However, it considered the adverse event parameters in the company's model to be unclear, so revised the model using mean inpatient costs and simplified adverse event data. The revised model allowed for multiple adverse events per patient and used average cost estimates of treating a typical adverse event in different settings.
5.14 The external assessment centre found that GreenLight XPS
produced a cost saving of £60.19 per patient compared with TURP, when using the company's assumptions of a 36% day-case rate and zero capital costs for GreenLight XPS. The greater savings compared with the company's model were due to the greater adverse event-related treatment costs with TURP than with GreenLight XPS.
5.15 Sensitivity analyses by the external assessment centre determined that GreenLight XPS becomes cost saving compared with TURP in non-high-risk patients when the day-case rate is 30% or higher.
5.16 In response to consultation comments, the external assessment centre carried out additional sensitivity analyses exploring how varying the ratio of monopolar to bipolar TURP comparator procedures affected the cost model. When compared with monopolar TURP in 100% of cases, the cost saving with GreenLight XPS fell from £60 to £5 per patient. The external assessment centre also added the cost of a laser bridge to the cost of GreenLight XPS (£700 for 50 cases, based on expert opinion). Including the cost of a laser bridge, the cost saving with GreenLight XPS was £46 per patient.
5.17 The external assessment centre considered that there was uncertainty about the clinical and cost assumptions of the company's secondary model in high-risk patients. In response to consultation comments, the external assessment centre revised the high-risk model with the following changes to HoLEP treatment:
increasing the lifespan from 5 to 10 years
increasing the number of patients having the procedure from 25 to 76 per year
removing the cost of the Ellik evacuator and fibre stripper
assuming that the morcellator blades and holmium fibres were reusable.
In this revised model, using the company's capital cost for HoLEP, GreenLight XPS was found to be cost incurring compared with HoLEP by £315 per patient. Using an alternative UK supplier's capital cost for HoLEP, GreenLight XPS was found to be cost incurring by £141 per patient. Because all clinical parameters were assumed to be the same (in the absence of clinical data to the contrary), the main cost driver was how many patients could have HoLEP per year. If 25 or fewer patients have HoLEP each year, GreenLight XPS becomes cost saving. Experts noted that some NHS centres do not re-use HoLEP fibres because of concerns about infection control, which would add to the costs of HoLEP.
5.18 The external assessment centre concluded there was insufficient information to develop a robust cost case for GreenLight XPS
compared with HoLEP.
5.19 The committee noted the results of the cost modelling that suggested that cost savings with GreenLight XPS are dependent on rates of day-case treatment and are potentially realised once this exceeds 30%. The committee accepted expert advice that this threshold for cost saving would be achievable for most urology centres.
5.20 The committee noted that the cost of adopting GreenLight XPS
does not involve a capital outlay for the console. It was informed that such an agreement is negotiated with the company on the basis of a minimum number of laser fibres purchased over a defined time period and is a practicable and realistic arrangement in the context of current NHS practice.
5.21 The committee was advised that the 10-minute difference in procedure time between GreenLight XPS and TURP would not influence the cost modelling conclusions.
5.22 The committee considered the potential need for using more than 1 laser fibre per GreenLight XPS procedure. The committee was reassured by experts about the durability of fibres and was informed that additional fibres would only usually be needed when vaporising extremely large prostates. The company also stated that the need for more than 1 fibre per procedure is rare, and if additional fibres are needed it has a fibre replacement programme that typically allows the additional fibres to be given to the hospital at no additional charge. The committee therefore concluded that this issue was unlikely to influence the cost case.
5.23 The committee considered the different costs of monopolar and bipolar TURP procedures in light of the results of the additional external assessment centre analysis and concluded that using monopolar or bipolar TURP does not fundamentally alter the cost case supporting the use of GreenLight XPS.