A full economic evaluation, cost-utility analysis, was done to compare two dialysis modalities - HD and PD - among ESRD patients in Iran from the health system perspective. To conduct a proper economic evaluation study, we followed the various steps as per the reference guidelines of the National Institute for Health and Care Excellence (NICE) [18]. The costs and outcomes were estimated based on a 10-year timeframe and health system perspective. Based on this perspective, we included the costs of equipment, facilities, supplies, medications, and human resources associated with providing each dialysis modality. We excluded indirect costs and direct non-medical costs borne by patients and caregivers, such as travel expenses and productivity losses. With regard to effectiveness, the health system perspective does not affect the effectiveness of the study. Specifically, the effectiveness outcomes of mortality, hospitalization rates, and quality-adjusted life years are clinical and patient-centered results that are independent of the perspective taken [19, 20].

The economic model was constructed according to the nature of the disease, literature review, the process of performing dialysis modalities, state transition probabilities, clinical outcomes in terms of QALY, and patient costs. The simple diagram of the Markov model used in the study with a one-year cycle length and 10-year timeframe is illustrated in Fig. 1. Four health states were considered, including HD, PD, kidney transplant, and death. The survival and mortality rates for HD and PD were obtained from a national cohort study [21], and the survival rate for kidney transplant was extracted from a meta-analysis study [22].

The probability of transition from PD and HD to kidney transplantation was obtained from the health insurance database. Also, the probability of rejecting the kidney transplant and returning to the PD state was extracted from a cost-utility study [23]. The probability of patients transitioning to the two states of HD and PD after transplant rejection was assumed to be the same.

Markov model for CEA of HD vs. PD

In this study, four scenarios were examined. The base scenario was defined based on available data (rate of PD in ESKD patients), which was 3% for PD compared to 97% for HD. These figures were extracted from the health insurance information. The other scenarios were determined assuming an increase in PD patients compared to HD: the second scenario (30% PD, 70% HD), the third scenario (50% PD, 50% HD), and the fourth scenario (70% PD, 30% HD).

To estimate the costs of dialysis modalities, first, the list of HD, PD, and kidney transplant patients covered by the Iranian Health Insurance Organization (IHIO) was obtained from the database of the National Institute for Health Insurance Research in 2022. The healthcare system in Iran relies on several health insurance funds to provide coverage for the entire population as mandated by law. The largest of these is the IHIO, which is legally required to cover over half of the country’s citizens. Eligibility for coverage under this organization is defined by legislation, and enrollment is compulsory for most people meeting the criteria. Once enrolled, there is little flexibility to change funds or insurers. The system is structured so that each eligible citizen must remain in their assigned fund based on the guidelines. This means that for the majority enrolled in the IHIO, they cannot opt out or select alternate coverage even if desired. Next, according to the number of patients with CKD in seven provinces (Tehran, Yazd, Fars, West Azarbaijan, East Azarbaijan, Hamadan, and Qazvin) and the total population of the provinces, 760 patients were selected as the final sample. The average age was 57.7 years, with a standard deviation of 16.6. The sample contained more males (60%) than females (40%). The largest proportion of patients were on HD (43.6%), followed by those with a kidney transplant (42.6%), and PD (13.8%). The average annual direct medical costs per patient, including medications, physician visits, lab tests, imaging services, dialysis service, and hospitalization, were estimated. The total direct medical costs of PD, HD, kidney transplant in the first year, and kidney transplant in the second year were 1,143,654,799 IRR, 848,855,549 IRR, 538,750,671 IRR, and 64,458,254 IRR, respectively. All cost data were calculated according to 2022–2023 prices. According to the NICE guidelines [18], the Iranian version of the EQ-5D-5 L questionnaire [24] was used through direct interviews with 518 patients to extract the utility values for patients with PD (n = 76), HD (n = 312), and kidney transplantation (n = 130). The questionnaire includes 5 dimensions (mobility, self-care, usual activities, pain/discomfort, and anxiety/depression), and each dimension has 5 levels (no problems, slight problems, moderate problems, severe problems, and extreme problems). The mean (SD) utility values were 0.550 (0.468) for PD patients, 0.423 (0.549) for HD patients, and 0.695 (0.341) for kidney transplant patients.

The inclusion criteria for the study comprised all patients with ESKD who were treated with PD or HD and covered by the Health Insurance Organization in the selected provinces, were at least 18 years and above, and finally for whom at least three months had passed since the start of their treatment with either PD or HD.

The incremental cost-effectiveness ratio (ICER) was used to determine the most cost-effective scenarios as follow:

Where C shows the costs and E shows the effectiveness.

A willingness-to-pay threshold is needed to analyze cost-utility results. In developing countries, the most commonly used threshold is the one recommended by the WHO, which is calculated based on GDP per capita [25]. According to this recommendation, if the ICER for a healthcare intervention in a country is less than the GDP per capita, that intervention is chosen as very cost-effective. In addition, if the ICER falls between 1 and 3 times the GDP per capita, the intervention is considered cost-effective. Finally, interventions with an ICER more than 3 times the GDP per capita are identified as not cost-effective. In this study, we used the WHO recommendation and the GDP per capita was $4,100 US at the time of this study. In 2022, the GDP per capita for Iran was equal to US $4,100 according to International Monetary Fund (IMF) data [26]. In 2022, US$1 was almost equal to 3,000,000 IRR [27].

Considering uncertainty regarding parameters included in the model, including utility values, costs, and probability of transition within and between states, a probabilistic sensitivity analysis using Monte Carlo simulation with 1,000 repetitions was carried out. A gamma distribution for cost data and a beta distribution for other variables such as transition probabilities and utility values were considered. All data analyses were done through TreeAge software 2020.