Cost-effectiveness of haemodialysis versus haemodiafiltration in Singapore: a health economic simulation

INTRODUCTION In Singapore, haemodialysis (HD) is the most prevalent dialysis modality, accounting for 87% of all dialysis treatments[1] costing SGD 2500 per month per patient.[2] However, patients undergoing HD have a high incidence of mortality as HD has limited capacity to remove large uraemic and protein-bound toxins. These compounds are associated with increased overall and cardiovascular mortalities. Haemodiafiltration (HDF) is a blood purification therapy that combines convective solute transport with diffusion and could enhance both large and protein-bound uraemic toxin removal. Recently, there have been several randomised controlled trials demonstrating that high-efficiency post-dilution online HDF could improve survival.[3] Haemodiafiltration has been reported to demonstrate better clinical outcomes, albeit at a higher price, across the globe.[4-6] The objective of this investigation was to determine the cost-effectiveness of HDF at the willingness-to-pay (WTP) threshold of Singapore (SGD 75,000).[7] The population of end-stage kidney disease (ESKD) patients requiring dialysis in Asia is expanding at a rate higher than other places worldwide. In many Asian countries, including China, the Philippines and Malaysia, the annual growth is more than 10%.[8] In Singapore, the number of new patients who initiated dialysis increased from 1049 per year in 2011 to 1473 per year in 2020. The majority of the new dialysis patients were aged 50–79 years, with close to eight in ten patients belonging to this age group in 2020. The median age at first dialysis increased slightly from 62.5 years in 2011 to 65.3 years in 2020.[1] Like other countries, diabetic mellitus is the major cause of chronic kidney disease (CKD; >55%) in Singapore.[9] METHODS The patient population considered in the analysis were adults with CKD newly initiated on kidney replacement therapy and dialysing via a fistula. A patient’s perspective was considered in the analysis. To determine cost-effectiveness, a Markov model was constructed with three health states: alive on HD or HDF, transplanted and death with 1-year cycle length. Patients could move from alive on HD or HDF treatment to death or to transplant. Patients from transplant could move to death [Figure 1]. The differentiating factors between HD and HDF (namely, mortality, infection rate, hospitalisations, etc.) were quantitatively incorporated in the decision analytic Markov framework utilising rates and probabilities from various peer-reviewed publications and local costs from Singapore. This allowed computational measurement of benefits of HDF over HD in the healthcare system of Singapore.[10-13] A report from the National Institute of Health and Care Excellence (NICE), UK, was identified as the source for transition probabilities of mortality and transplant.[10] A time-dependent annual probability of death and transplant, for patients alive in the model on HD and HDF treatments at years 1–10 after initiating dialysis, was used to simulate the patient journey [Table 1].[10] In the NICE report, a risk ratio of 0.82 was used to model the relative treatment effect of HDF compared to HD on mortality,[10] which has been also suggested in a recent publication.[11] Hence, a risk ratio of 0.82 was used to compute the comparative survival benefits of HDF over HD in the model. Rate of transplant after 10 years was assumed to be zero based on the trend for a decreasing rate of transplant over time observed in the later years of the analysis, similar to NICE 2018.[10] The mortality rate in transplant state was assumed to be similar for HD or HDF arms and was based on Singapore data considering a living donor.[14] Benefits in terms of reduced hospitalisation rate in HDF were parameterised based on a real-world study in Taiwan (25.44% for HD and 20% for HDF).[12] Proportion of bloodstream infections in the event of hospitalisation in HD or HDF was imputed as per a published report.[13] A similar proportion for bloodstream infection in hospitalised patients in HD and HDF arms was assumed for the analysis in the absence of a specific study. The details of the transition probabilities and costs have been presented in Tables 1 and 2. Outcomes were valued in terms of life years (LY) and quality-adjusted life-years (QALYs). Future costs and consequences were discounted at 3%. The results are reported in terms of incremental cost per QALY gained using HDF versus HD.Table 1: Probability of death used in the model for HD and HDF simulated patients over 10 years.[ 10 , 14 ]Table 2: Clinical and cost parameters for assessing the cost-effectiveness of HD versus HDF.Figure 1: Markov model to assess the cost-effectiveness of haemodialysis (HD) versus haemodiafiltration (HDF).Considering three sessions of dialysis each week, each HD session in Singapore can be computed to be SGD 192 (based on the report, SGD 2500 for 13 sessions/month/patient).[2] Thus, the cost per session/patient as ‘price anchor’ for each session of HD was SGD 192. Considering a different price for HDF, owing to the higher costs of disposable equipment and water purity control, we estimated it to be marginally higher than the ‘price anchor’ for HD. However, in the absence of any published ‘price per session per patient’ of HDF in Singapore, we followed an arithmetic progression sequence of incremental prices per session for the analysis. Thus, incremental price points, estimated across a price corridor of 2.5%, 5%, 7.5%, 10%, 12.5%, 15%, 17.5% and 20%, were higher than the HD price per session of SGD 192. RESULTS The deterministic incremental cost-effectiveness ratio (ICER) of SGD 40,328, SGD 52,849, SGD 65,370 and SGD 77,891 was evident at 2.5%, 5%, 7.5% and 10% price premium above the ‘price anchor’ of SGD 192. Thus, HDF was found to be cost-effective at the WTP threshold of Singapore (SGD 75,000) up to a price premium of 10%, considering the price of HD as SGD 192 per patient per session. At the highest price premium (10% above the price anchor), for HDF, the total costs, LYs and QALYs for HDF were SGD 17,501.25, 0.37 and 0.21, respectively (per patient), while for HD, the total costs, LYs and QALYs for HDF were SGD 16,317.83, 0.34 and 0.19, respectively (per patient). The ICER was 72,203/QALY at the WTP of Singapore at the highest price premium. The incremental cost per patient (vs. HD) over the time horizon of analysis was SGD 11,834 to achieve 0.165 incremental QALYs. Probabilistic sensitivity analysis using a Monte Carlo simulation was conducted to assess the impact of the joint uncertainty around the key parameters, and the results have been shown in Figure 2. A cost-effectiveness acceptability curve [Figure 3] was created, considering the WTP of Singapore as SGD 75,000. The analysis showed that 60% of simulations demonstrated HDF as the cost-effective strategy [Figure 3].Figure 2: Incremental cost-effectiveness ratio (ICER) scatter plot demonstrates haemodiafiltration (HDF) as a cost-effective strategy as compared to haemodialysis (HD) at 10% price premium of HDF over HD in Singapore. QALY: quality-adjusted life-yearFigure 3: Cost-effectiveness acceptability curve (CEAC) plot demonstrates haemodiafiltration (HDF) as a cost-effective strategy as compared to haemodialysis (HD) at 10% price premium of HDF over HD in Singapore. QALY: quality-adjusted life-year, WTP: willingness to payDISCUSSION Our results are comparable to those of other studies in Asia. In South Korea, HDF was reported to be cost-effective compared to HD under a WTP threshold of KRW 35 million/QALY, with an ICER of KRW 1.23 million/QALY,[5] and in Japan, the ICER was reported as 641.7 (WTP threshold of JPY 10,000/QALY).[21] In other continents, HDF has been reported to be cost-effective, with an ICER of CAD 53,270 per QALY gained (WTP threshold CAD 90,000 per QALY) in Canada[6] and EUR 6982/QALY gained in Italy (WTP threshold EUR 40,000/QALY).[22] Our findings corroborate the findings in Korea, Japan, UK, Italy and Canada. Our analysis demonstrates the public health benefits of HDF in patients in Singapore. A recent report has also underlined that the potential barriers to greater utilisation of HDF therapies include concerns regarding additional costs of HDF, for example, for the preparation and microbial testing of substitution fluids for HDF.[23] However, as HDF achieves more efficient reduction of the uraemic toxins compared to HD, providing survival benefits to patients, it is the need of the hour to ensure better utilisation of HDF in the healthcare systems at a fraction of incremental costs in Singapore. Financial support and sponsorship Nil. Conflicts of interest There are no conflicts of interest.

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.