Fig 1.
Markov model state transition diagram.
Possible state transitions in the Markov model with a cycle length of 3 months. Patients begin in the stable state and may transition to post-recurrent minor stroke (mRS 0–2), post-recurrent moderate-to-severe stroke (mRS 3–5), or death. All living states carry a probability of transitioning to death.
Table 1.
Cost parameters used in the economic model.
Table 2.
Utility parameters used in the economic model.
Fig 2.
Tornado diagram: One-way sensitivity analysis.
Impact of parameter uncertainty on net monetary benefit (NMB). Each bar represents the range of NMB when the corresponding parameter is varied between its low and high values while holding all other parameters at base case. The vertical line indicates the base-case NMB of $57,137. PFO closure remains cost-saving across all parameter ranges tested.
Fig 3.
Cost-effectiveness acceptability curve.
Probability of PFO closure being cost-effective compared with medical therapy alone across a range of willingness-to-pay (WTP) thresholds. At the Chinese WTP threshold of $37,654/QALY (3 × GDP per capita), PFO closure has a 94.2% probability of being cost-effective. At a WTP of $0 (cost-saving threshold), PFO closure has an 87.3% probability of being dominant.
Fig 4.
Probabilistic sensitivity analysis scatter plot.
Cost-effectiveness plane showing results of 10,000 Monte Carlo simulations. Each point represents one simulation iteration. Green points indicate PFO closure is dominant (more effective and less costly); orange points indicate cost-effective (more effective but more costly, below the WTP threshold); red points indicate not cost-effective. The dashed ellipse represents the 95% confidence region. The crosshair indicates the mean estimate (ICER: − $2,868/QALY).
Table 3.
Base-case cost-effectiveness results over 30-year time horizon.
Table 4.
Clinical outcomes comparison over 30-year model horizon.
Table 5.
One-way sensitivity analysis results.