Cost-effectiveness analysis of stroke management under a universal health insurance system

• Published in: Journal of the neurological sciences Vol. 323; no. 1; pp. 205 - 215
• Main Authors: Chang, Ku-Chou, Lee, Hsuei-Chen, Huang, Yu-Ching, Hung, Jen-Wen, Chiu, Hsienhsueh Elley, Chen, Jin-Jong, Lee, Tsong-Hai
• Format: Journal Article
• Language: English
• Published: Amsterdam Elsevier B.V 15.12.2012; Elsevier
• Subjects: Acute Disease; Adult; Aged; Biological and medical sciences; Carcinoembryonic antigen; Cohort Studies; Cost of illness; Cost-Benefit Analysis; Disease Management; Female; Health Resources - economics; Health Resources - utilization; Hospital Costs - standards; Hospital Costs - statistics & numerical data; Hospital Departments - economics; Humans; Male; Medical sciences; Middle Aged; Models, Economic; Neurology; Neurosurgery; Patient Readmission - economics; Patient Readmission - statistics & numerical data; Patients' Rooms - economics; Rehabilitation Centers - economics; Retrospective Studies; Sampling Studies; Severity of Illness Index; Stroke - classification; Stroke - economics; Stroke - epidemiology; Stroke - therapy; Stroke Rehabilitation; Stroke/mortality/therapy; Survival Analysis; Taiwan - epidemiology; Universal Coverage - economics; Vascular diseases and vascular malformations of the nervous system; General/miscellaneous wards; OPD; Retrospective studies; Transient ischemic attack or other unspecified cerebrovascular disease; FYMC; HS; Ischemic stroke; AUC; Longitudinal Health Insurance Database 2000; Stroke/mortality/therapy; Computed tomography; LHID 2000; ROC curve; CEA; Cerebrovascular disease; ED; Length of stay; Incremental cost-effectiveness ratio; Cerebral infarction; AE; Subarachnoid hemorrhage; TIA/unspecified; Hemorrhagic stroke; IS; NHIRD; Charlson Comorbidity Index; CVD; LOS; National Health Insurance Research Database; Survival analysis; Bureau of National Health Insurance; Risk-adjustment index; National Health Insurance; SAH; Magnetic resonance image; MRI; NS; Cox proportional-hazards model; Cox-PH model; NW; NHI; BNHI; Health resources/utilization; Emergency department; Area under the curve; Adverse outcome events; First-year medical cost; Receiver Operating Characteristic curve; Cost of illness; ICER; CI; CCI; Intracerebral hemorrhage; Neurosurgery wards; ICH; GW; Cost-effectiveness analysis; CT; Outpatient Department; Neurology/rehabilitation wards; RAI; Health system; Health insurance; Stroke; Costs; Nervous system diseases; Mortality; Cardiovascular disease; Survival; Retrospective; Cerebral disorder; Vascular disease; Clinical management; Treatment; Central nervous system disease; Cost analysis; Cost efficiency analysis
• ISSN: 0022-510X; 1878-5883
• DOI: 10.1016/j.jns.2012.09.018

Abstract Objective Cost-effectiveness analysis (CEA) of stroke management was evaluated in three care models: Neurology/Rehabilitation wards (NW), Neurosurgery wards (NS), and General/miscellaneous wards (GW) under a universal health insurance system. Methods From 1997 to 2002, subjects with first-ever acute stroke were sampled from claims data of a nationally representative cohort in Taiwan, categorized as hemorrhage stroke (HS) including subarachnoid hemorrhage (SAH) and intracerebral hemorrhage (ICH); or, ischemic stroke (IS), including cerebral infarction (CI), transient ischemic attack/ unspecified stroke (TIA/unspecified); with mild-moderate and severe severity. All-cause readmissions or mortality (AE) and direct medical cost during first-year (FYMC) after stroke were explored. CEA was performed by incremental cost-effectiveness ratios. Results 2368 first-ever stroke subjects including SAH 3.3%, ICH 17.9%, CI 49.8%, and TIA/unspecified 29.0% were identified with AE 59.0%, 63.0%, 48.6%, 46.8%, respectively. There were 50.8%, 13.5%, 35.6% of stroke patients served by NW, NS and GW with AE 44.9%, 60.6%, 56.0%, and medical costs of US$ 5,031, US$ 8,235, US$ 4,350, respectively. NW was cost-effective for both mild-moderate and severe IS. NS was the dominant care model in mild-moderate HS, while NW appeared to be a cost-minimization model for severe HS. Conclusions TIA/unspecified stroke carried substantial risk of AE. NS performed better in serving mild-moderate HS, whereas NW was the optimal care model in management of IS.