Filtering for the best policy: An economic evaluation of policy options for kidney replacement coverage in the Philippines

Kidney failure patients in the Philippines have free choice on their kidney replacement therapy (KRT), with a majority choosing haemodialysis (HD) over peritoneal dialysis (PD) and transplantation despite the inadequate coverage of HD. Although national health insurance coverage is limited, KRT remains to be one of the top benefits pay‐outs in the country. The study aims to identify the most cost‐effective policy strategy for financing KRT in the Philippines, in the context of a universal healthcare policy.

Kidney failure (KF) is a catastrophic illness that affects morbidity and mortality and brings about significant financial burden in both highand low-income countries. 1 Several options for kidney replacement are available, such as transplantation, haemodialysis (HD) and peritoneal dialysis (PD); however, access remains limited due to a range of social, economic and health system factors. 2 It was estimated that 2.284 million premature deaths worldwide in 2010 are due to the lack of access to any kidney replacement therapy (KRT). 3 A larger treatment gap was seen in low-income countries, particularly in Asia, where around 1.907 million patients need KRT but are unable to receive it. In the Philippines, KF is seen as one of the most burdensome conditions as it is the second largest expense of the national health insurance agency, the Philippine Health Insurance Corporation or PhilHealth 4 and is often a source of out-of-pocket health expenditure among households. By the end of 2016, there were 36 247 patients on dialysis in the country, 5 an increase of 15% from the previous year. The number of patients who have KF but are not receiving any form of dialysis remains unknown.
While a kidney transplant is considered the best option for KF patients, donor kidneys are not readily available, and the cost of postsurgery monitoring and immunosuppression therapy is limiting for most. Patients who are ineligible or are on the waiting list for transplant have to resort to dialysis, where the choice depends on both patient and system level factors such as age, comorbidities, availability, affordability and even personal preferences. 6 Majority of patients in the Philippines are on centre-based HD (94%), while 4% are on continuous ambulatory peritoneal dialysis. Only 2% of KF patients are able to get kidney transplants. PhilHealth currently supports funding for all three modalities but at different levels of coverage (see Appendix S1). The inadequate coverage of 90 sessions of HD annually assumes that patients are able to bear the cost of the remaining sessions needed. However, current practice shows that more than half (52.76%) of the patients undergo dialysis twice weekly or less, to spread out the allocation of 90 sessions each year, resulting in suboptimal outcomes. Most patients still choose HD despite the higher coverage for PD.
Several countries have implemented "PD-First" or "PD-favoured" policies wherein use of PD as first modality of choice for KF patients is encouraged or incentivized. This is due to the growing body of evidence that PD is more cost-saving compared with centre-based HD. 7 Reflections from Hong Kong and Thailand's experience suggest that collecting necessary economic evidence is critical when developing a national dialysis coverage policy. 8

| METHODS
This study utilized economic evaluation methods, specifically, a cost-utility analysis, which is defined as a comparative analysis of alternatives, in terms of both their costs and outcomes (presented in quality-adjusted life years [QALY]). 10 The alternatives being compared are the options for covering KRT vs the current scenario where a vast majority are on twice-weekly HD or less, which is deemed inadequate. These are shown in Table 1 The base-case population is 52 years, which is the mean age of adult patients with KF in the Philippines. 5 All patients are also assumed to undertake either of the three modalities, as palliative care alone is rarely preferred since KRT is covered by PhilHealth. Patients can switch between each dialysis modality and can undergo a transplant pre-emptively or after being on dialysis. For simplicity, we assume that patients can only undergo a transplant once in their lifetime, and then progress to the post-transplant state where they take immunosuppressants for a lifetime or until they require dialysis as a result of chronic graft loss.

SUMMARY AT A GLANCE
This article aims to identify a cost-effective policy strategy for access to kidney replacement therapy in the Philippines, and calls on the government to consider shifting to a Peritoneal Dialysis (PD)-First strategy and support policies that promote kidney transplants among existing PD and haemodialysis patients.
The study took two perspectives: the government purchaser (PhilHealth), which takes into account only the direct medical costs, and societal perspective which includes non-medical costs such as travel, meals and caregiver expenditure. Productivity losses were excluded due to the wide variation in income estimates. Relevant costs and outcomes per health state are estimated and modelled for a lifetime (99 years) considering the treatment duration of dialysis and immunosuppression post-transplant. Both costs and health outcomes were discounted at an annual rate of 3%, as recommended by draft HTA guidelines at the time the study was conducted.
Effectiveness of the policy strategies were measured in terms of treatment-specific survival and quality of life, factoring the distribution of patients per option. Treatment survival data per modality was obtained from the Philippine Renal Disease Registry and the National Kidney and Transplant Institute (NKTI). NKTI is the national referral tertiary facility for kidney disease and has the largest volume of KF patients in the country. Key distribution parameters and rates of complications were obtained from the same hospital. We assumed similar efficacy in terms of survival for PD and adequate HD for the model and used the survival data for twice-weekly HD patients for the status quo option. This assumption is consistent with other evaluations of the same interventions. 11,14,15 The 2-year aggregate survival data from the hospital were modelled using exponential distribution to estimate annual cumulative survival and death risk for twice-and thrice-weekly HD, PD and transplant patients. In addition to the registry and hospital data, a cross-sectional survey using the EuroQol 5D-5L tool 16 was conducted from March to April 2019 among 262 KF patients. The survey was administered to eligible adult patients in NKTI who have been on their current modality (post-transplant, HD or PD) for at least 3 months, to get an estimate of health utility per modality. Utility scores were calculated using the Philippines' EQ-5D-5L value set. 17 In terms of estimating costs and resources used, annual costs per health state were taken from three main sources. First, current coverage rates were obtained from PhilHealth's published case rates as of 2015. This reflects the actual amount paid for by the national health insurance agency, which was used to model the costs of the status quo option. Second, costs used for adequate HD, PD-First and preemptive transplant policies were taken from the prevailing unit costs used in NKTI, which were estimated through a micro-costing exercise.
Rates of peritonitis and vascular access-related conditions were factored in the costing. Because the differences in initiation costs for HD (eg, vascular access surgery) and PD (catheter insertion procedure) were minimal and were incurred only in the first year of treatment, Lastly, non-medical and indirect costs were obtained through an additional questionnaire from the same respondents of the EQ-5D-5L survey. All costs were estimated in 2019 Philippine Pesos (PHP).
PhilHealth Case Rates are from 2015; however, these were not adjusted for inflation since these still reflect the current coverage rates at present. All input parameters used in the study are presented in Table 2.
The cost effectiveness of each policy option is presented as an incremental cost-effectiveness ratio (ICER), which is the ratio of the difference between costs and outcomes of a policy strategy (eg, PD-First) and the status quo. Deterministic and probabilistic sensitivity analysis were conducted to account for structural and parameter uncertainty. A Monte Carlo simulation was conducted where transition probability, cost and utility parameters were assigned a probability distribution and a random estimate was selected for 10 000 iterations. 25 The ICER for each simulation was compared against a ceiling threshold value and presented in a cost-effectiveness acceptability curve. Although there is no explicit threshold in the Philippines, it was recommended by the HTA Council of the Department of   Given these, the study is not without limitations. Some of the survival data, cost and quality of life inputs were taken from a single facility, the NKTI. Although patient characteristics of the NKTI cohort is consistent with the national KF profile based on the national registry data, the cost data may not necessarily be generalizable to the entire Philippines. NKTI is a large tertiary hospital with a high volume of patients, thus they are able to procure supplies and machines at a much cheaper price due to economies of scale. Smaller clinics in the provinces outside the National Capital Region may not be able to get similar prices that were used in the study. However, if PhilHealth does the strategic purchasing and contracting for these facilities, then they may be able to drive prices down and make KRT provision more sustainable and affordable.

| DISCUSSION
Covering kidney replacement therapies in context of universal health coverage remains to be a challenge for most countries. Lessons from the Philippines show that a good understanding of current context in terms of existing coverage and distribution of patients per modality is essential in making a relevant evaluation. There was a need to scrutinize whether partial dialysis coverage result to suboptimal outcomes and possibly widening the inequity gap which goes against the key principles of a universal health coverage policy. From this assessment, shifting to a PD-First policy instead of expanding current HD coverage is the best strategy to make KRT affordable and sustainable for the health system, complemented by increased kidney transplant uptake. Implementation and policy-making considerations are provided to ensure better use of this piece of economic evidence.