Cost-Effectiveness Of Inpatient Rehabilitation Compared To Home Discharge Following Longer Stay In General Or Neurological Ward, After Admission For The First Ever Stroke In Greece

OBJECTIVES: To assess the cost-effectiveness of Inpatient Rehabilitation (IR) versus home discharge following longer stay in general/neurological ward (non-IR) for patients being hospitalized for the first acute stroke in Greece, from a third-party payer perspective (EOPYY). METHODS: A decision analytic model consisting of a 1-month decision tree and 5-year Markov Model was developed in Excel. Decision tree consisted of four nodes: 'Home Independent', 'Home Dependent', 'Hospital', 'Death'. Patients enter Markov model at the health state they exit the decision tree. In the Markov model, they may transit among five health states, at 1-year cycles, with assigned probabilities extracted from literature. The efficacy of IR was taken into consideration only at the 1st month. Resource utilization (i.e. medication, monitoring tests, outpatient rehabilitation, equipment etc) that depends on the health state/node of the model was extracted from experts using a questionnaire developed to serve the purpose of the present study. The respective unit costs, as well as the cost of IR and hospitalization for non-IR were obtained official local sources (e,2016). Life-years (LYs), quality-adjusted life-years (QALYs), and cost-effectiveness in terms of life-years gained (LYG) and QALY gained were evaluated. One-way (OWSA) and probabilistic sensitivity analysis (PSA) were conducted to evaluate the robustness of base-case analysis. RESULTS: An average patient transferred to IR, following hospitalization for 1st stroke, was estimated to have higher survival by 0.368 LYs and gained 0.288 QALYs compared to non-IR, at a higher cost of €1,934. Incremental cost-effectiveness ratios were €5,258/LYG and €6,728/QALY gained relative to non-IR. The proportion of patients being home dependent following IR and re-hospitalized (non-IR) influenced the results. The probability of IR to be cost-effective exceeds that of 97% at a willingness-to-pay of €34,000. CONCLUSIONS: Given the assumptions and limitations of this analysis, IR seems to be a cost-effective option in Greece, for patients experiencing their first ever stroke.