摘要

【背景】脑卒中是我国成人致死、致残的首位病因。及时的早期康复治疗能够有效改善脑卒中患者的功能预后。康复服务的高质量发展需要医保资金支持，相应的医保支付制度尚不完善成为制约脑卒中早期康复服务发展的重要原因。目前我国正在推行以按病种付费为主的多元复合式医保支付方式，鼓励医疗康复服务等长期住院按床日付费，而目前各地确定的床日基本费率存在较为粗放的问题，不能准确反映脑卒中早期康复住院的成本。以病例组合为基础的医保支付制度探索是国内外共同的热点课题。

【目的】了解脑卒中早期康复患者的住院费用现状和构成，并进一步探究影响脑卒中早期康复患者日均总费用和日均医药、康复及护理费用的影响因素，建立脑卒中早期康复患者床日费用病例组合模型，为脑卒中早期康复医保支付制度改革提供一定的参考。

【方法】通过文献研究法和专家咨询法了解脑卒中早期康复领域的临床治疗和成本控制的理论研究进展，系统收集国内外脑卒中康复住院费用影响因素和病例组合研究经验。采用非参数检验和多元线性逐步回归分析的统计学方法分析脑卒中早期康复阶段日均住院费用的影响因素。采用分类回归树的方法构建脑卒中早期康复患者床日费用病例组合模型。使用方差减小系数RIV和变异系数CV值对模型进行验证。

【结果】

脑卒中早期康复患者次均住院总费用为15072.09±8865.98元，平均日均住院总费用为1202.30±1232.30元。总费用中医药费用占比最高，为75.73%，康复费用约占13.01%，护理费用约占7.01%。康复费用中，中医康复费用和物理治疗费用占比最高，分别为47.87%和42.18%，言语治疗和作业治疗费用占比较少，分别为5.71%和4.23%。

影响脑卒中早期康复患者日均住院总费用的主要因素为医疗机构等级、年龄、气管切开和使用呼吸机情况、ADL等级、并发症与合并症情况、认知障碍、吞咽障碍和抑郁状态(P<0.05)；影响日均医药费用的主要因素为气管切开和使用呼吸机情况、认知障碍、ADL等级、并发症与合并症和吞咽障碍(P＜0.05)；影响日均康复费用的主要因素为ADL功能评分、吞咽障碍、年龄、并发症与合并症(P＜0.05)；影响日均护理费用的主要因素为认知障碍、并发症与合并症、气管切开和使用呼吸机情况(P＜0.05)。

构建了以日均总费用为分组轴心的8个病例组合，分组节点变量包括气管切开和使用呼吸机情况、认知障碍、并发症与合并症、ADL等级、年龄五个变量。CV值均＜1，RIV值为55.36%。

构建了基于患者导向模型（PDPM）思路的床日费用病例组合，即对日均医药服务、日均康复服务、日均护理服务、日均设施服务四部分费用分别建立病例组合。各部分病例组合分组CV值基本＜1，RIV值在40%-60%之间。

【结论】按PDPM分组思路构建的床日费用病例组合模型以患者的需求为导向，对脑卒中早期康复阶段医药、康复、护理服务成本决定因素进行区分再分别分组，相较于以日均总费用为分组轴心的病例组合模型可以更加准确地反映不同类型服务的成本，各组组内同质性和组间变异度表现良好。

Abstract

Background: Stroke is the leading cause of death and disability among adults in China. Timely early rehabilitation therapy can effectively improve the functional prognosis of stroke patients. The high-quality development of rehabilitation services requires support from medical insurance funds, and the incomplete medical insurance payment system has become an important constraint on the development of early stroke rehabilitation services. Currently, China is implementing a multi-component medical insurance payment method primarily based on diagnosis-related group (DRG), encouraging exploration of payment based on bed-days for long-term rehabilitation hospitalization. However, the currently determined basic bed-day rates in different regions have significant limitations and fail to accurately reflect the costs. The exploration of a case-mix-based medical insurance payment system is a common research focus both domestically and internationally.

Objective: To understand the composition of hospitalization costs during the early-stage rehabilitation phase of stroke and further exploring the factors influencing the average daily total cost, as well as the average daily medical costs, average daily rehabilitation costs, and average daily nursing costs during this phase. Constructing a case-mix model for bed-day costs of early-stage stroke rehabilitation, aiming to provide insights for the reform of medical insurance payment systems in early-stage stroke rehabilitation.

Methods: Literature review method and expert consultation method were used to understand the theoretical research progress of clinical treatment and cost control in the field of early stroke rehabilitation, and the research experience of influencing factors and case portfolio of stroke rehabilitation hospitalization costs at home and abroad was collected. Non-parametric test and multiple linear stepwise regression analysis were used to analyze the influencing factors of average daily hospitalization cost in early stroke rehabilitation stage. The classification and regression tree(CART) was used to construct the case-mix model of daily cost of early stroke rehabilitation. RIV and CV were used to verify the model.

Results: 

The average total hospitalization cost of early stroke rehabilitation was 15072.09±8865.98 yuan, and the average daily total hospitalization cost was 1202.30±1232.30 yuan. Among the total expenses, medical expenses accounted for 75.73%, rehabilitation expenses accounted for 13.01% and nursing expenses accounted for 7.01%. Among the rehabilitation expenses, TCM rehabilitation expenses and physical therapy expenses accounted for the highest proportion, 47.87% and 42.18%, while speech therapy and occupational therapy expenses accounted for relatively less, 5.71% and 4.23%.

The main factors affecting the average daily hospitalization cost were the level of medical institution, age, tracheotomy and ventilator use, ADL, complications, cognitive impairment, dysphagia and depression (P<0.05). The main factors affecting the average daily medical cost were tracheotomy and ventilator use, cognitive impairment, ADL, complications and dysphagia (P<0.05). The main factors affecting the average daily rehabilitation cost were ADL, dysphagia, age and complications (P<0.05). The main factors affecting daily nursing cost were cognitive impairment, complications, tracheotomy and ventilator use (P<0.05).

Eight case-mix of total daily cost were constructed. Required variables included tracheotomy and ventilator use, cognitive impairment, complications, ADL, and age. The CV were all lower than 1, and the RIV were 55.36%.

A case-mix model of daily bed expense was constructed with reference to PDPM, including medical service, rehabilitation service, nursing service and facility service. The CV of each part were basically less than 1, and the RIV were between 40%-60%.

Conclusions: The case-mix model constructed according to the grouping idea of PDPM shows good homogeneity and variation among groups. The case-mix model can distinguish and medicine, rehabilitation and nursing services based on patient needs. Compared with the case-mix model grouped according to the average daily total cost alone, it can more accurately reflect the cost of different types of services and is more consistent with the needs of early rehabilitation.