Pro 2018

The development of this model adhered to the recommendations for the conduct of economic evaluations in osteoporosis (Hiligsmann et al., 2019 (link)). We used an updated version of the previously validated individual-level state-transition model (You et al., 2020 (link)) to estimate the impact of the compliance and persistence on the cost-effectiveness of alendronate treatment for Chinese postmenopausal osteoporotic women aged 65 and older. The model estimated the outcomes including the number of fracture quality-adjusted life-years (QALYs); direct societal costs in 2018 US dollars (USD); and incremental cost-effectiveness ratios (ICERs) per QALY gained. Costs and health outcomes beyond the first year were discounted at an annual rate of 3%, which is consistent with Chinese guidelines for pharmacoeconomic evaluations (Liu, 2011 ). We assessed cost-effectiveness from the health care payer perspectives and considered three times per capita gross domestic product of China in 2018 (USD 29,340) as the threshold. We used TreeAge Pro 2018 (TreeAge Software Inc., Williamston, MA, USA) to perform our analyses.

We performed several one-way sensitivity analyses, in which the cost-effectiveness ratio was calculated by altering the following parameters identified from targeted literature reviews: 1) the probabilities of developing CVD in 10 years; 2) the costs of CABG and PTCA procedures; 3) the cost of almonds; and 4) the LDL-C response among participants with existing CVD. In the 10-year model, we applied a 3% per year discount rate to costs and effectiveness [35 , 36 ].
We further conducted Monte Carlo probabilistic sensitivity analysis (PSA) with 10,000 simulations to address uncertainty. We extracted data from Pikula et al. and other literature [13 (link), 17 (link), 18 (link)] to estimate the distributions of key (parameters)
Preferred alternative was chosen based on the net monetary benefit (NMB):
$$\widehat{\mathit{NMB}}=\lambda \times \mathrm{\Delta }\overline{E}-\mathrm{\Delta }\overline{C}$$
where λ is the maximum WTP for health care, ∆ $\overline{E}$ is the difference in the mean effectiveness of two strategies, and ∆ $\overline{C}$ is the difference in the mean cost of two strategies [37 (link)]. TreeAge Pro 2018 was used to conduct the analyses.

Statistical analysis and Markov model were performed using TreeAge Pro 2018, and graph-plotting was done with Excel software. All future costs and QALYs were discounted at 5% (3–5%) per year.
Incremental cost-effectiveness ratios (ICERs) as the ratio of the difference in costs between treatment strategies divided by the difference in QALYs were calculated. A strategy producing an ICER of US$29,295 per QALY, as 3-times per capita gross domestic product (GDP) of China in 2018 [43 ], was considered as cost-effective. A strategy producing an ICER of US$9765 per QALY, as one-time per capita GDP of China in 2018, was considered as being highly cost-effective.
One-way sensitivity analysis was conducted to determine the effects of parameters on the ICER. Probabilistic sensitivity analysis based on a second-order Monte Carlo simulation with 1000 iterations was then conducted to ascertain the model stability. Results were reported as cost-effectiveness acceptability curves. The range, distribution, and source for each parameter were shown in Supplementary Table S1.

A Markov model was constructed on the basis of outcomes data from the CheckMate 227 trial to evaluate the costs and effectiveness of using nivolumab plus ipilimumab vs. chemotherapy as first-line therapy for advanced NSCLC from the US payer's perspective. The Markov model cycle length was 6-weeks and the time horizon were 20-years. We adopted a 3% discount rate per year for both costs and outcomes (20 (link)). The total costs, life years (LYs), quality adjusted life years (QALYs), and incremental cost-effective ratios (ICERs) were calculated in each treatment strategy. The Markov model was constructed via TreeAge Pro 2018 (TreeAge Software Inc., Williamstown, MA).
The model structure included three states to represent the progression of advanced NSCLC: progression-free survival (PFS), progressive disease (PD), and death (Supplementary Figure 1). Patients were treated with nivolumab plus ipilimumab or chemotherapy in the PFS state until progression. All patients could continue subsequent treatment until death if any disease progression or unacceptable toxic effects occurred. Grades 3 or 4 adverse events (AEs) with a ≥1% frequency reported in CheckMate 227 trial were included.

A comprehensive Markov model was constructed to estimate health and economic outcomes of different treatments for locoregionally advanced NPC patients using real-world clinical data (Supplementary Figure 1). As shown in Figure 1, the model structure included three exclusive health states to represent different characteristics of locoregionally advanced NPC: disease-free survival (DFS), progressed disease (PD) and death. The Markov cycle length was three weeks, which is consistent with a clinical treatment schedule (15 (link)). All patients entered the model in the DFS state and immediately commenced treatment. During each cycle, patients either remained on DFS, progressed to PD or death.
The main outcomes of the study were total costs, life-years (LYs), quality-adjusted life-years (QALYs), and incremental cost-effectiveness ratios (ICERs). A 3% annual discount rate was used for survival and cost estimates. All costs were converted to 2020 US dollars (1 USD = 6.8606 RMB) (20 ). Three times of the per capita gross domestic product (GDP) in China in 2019 ($31,008.16) was used as a willingness-to-pay (WTP) threshold according to World Health Organization’s criteria (21 , 22 (link)). The model was performed using the decision analytic software TreeAge Pro 2018 (TreeAge Software, Williamstown, MA).

A Markov health state transition model was constructed using (TreeAge Pro 2018, TreeAge, Willliamstown, MA) to compare six treatment time windows among a base-case cohort of patients with AIS aged 66 [1 ]. Outcomes within treatment initiation time windows of 61–120 min, 121–180 min, 181–240 min, 241–300 min, 301–360 min, and 361–420 min from onset were simulated over a lifetime horizon. Incremental cost-effectiveness ratios (ICER) defined as incremental costs/QALY and net monetary benefit (NMB) were calculated to evaluate cost-effectiveness. We used a willingness-to-pay (WTP) threshold of ¥71,000 per QALY (US$10,280/ QALY), which is the 2019 gross domestic product (GDP) per capita in China [12 ].