比較完全整合與部分整合個案管理模式對使用者之日常生活功能變化的影響及相關因素：以臺北市立聯合醫院中興院區為例

Abstract

研究背景與目的： 因應我國人口快速老化，台北市立聯合醫院自民國106年起承接三種不同的社區整合照護計畫。依據過去對整合照顧的分類與機制，將台北市立聯合醫院承接之三種整合照護模式分成為兩大類：第一類為部份整合個案管理模式－衛福部長照2.0社區整體照顧服務體系A計畫。衛福部長照2.0之長照服務計畫屬於協調到合作層次的整合程度，並不會常規將居家醫療或是非營利組織提供的資源進行評估與連結。第二類為完全整合個案管理模式－分別由社會局委託及衛生局委託執行之照護計畫，常連結居家醫療與其他非常規連結、非常照2.0支付之服務項目，屬於更高度整合的整合照護計畫。過去研究著重服務類型整合，沒有從不同整合程度的照護模式對於個案的日常生活功能影響之探討。因此，本研究欲以臺北市立聯合院區承接的不同整合程度的整合照護計畫之資料，將個案分為接受部分整合個案管理模式與完全整合個案管理模式兩類，探討兩種整合照護模式對於個案日常生活功能改變的影響。本研究假設完全整合照護模式相較於部分整合照護模式，可以更有效地維持日常生活功能。為此訂定三個研究目標： 1. 比較兩種照護模式中，個案的基本特質與服務使用情形是否有差異。 2. 比較兩種照護模式中，個案的日常生活功能退化速度是否有差異。 3. 探討兩種照護模式，對個案日常生活功能退化之潛在影響因素。 研究方法： 本研究使用次級資料進行縱貫式研究。資料來自衛生福利部照顧服務管理資訊平臺及臺北市立聯合醫院中興院區內部的資料庫，分析西元2020年1月1號至12月30號此時間段中台北市立聯合醫院中興院區社區整合照護科有收案在案之照護對象，排除資料缺失、無追蹤、無複評之個案，共計篩選出541位個案，分為部分整合照護組416人，完全整合照護組125人。針對研究目的一，使用獨立樣本T檢定、卡方檢定（Chi–Square）以及變異數分析比較兩組在照護過程中，個案的基本特質、所使用的服務項目數、非長照支付的服務使用數、以及使用居家醫療的比例是否有差異。針對研究目的二，使用獨立樣本T檢定以及變異數分析，比較兩組的平均日常生活功能變化速度的差異。因應兩組在照顧個案的複雜照顧程度有所不同，我們分別以獨居、中低收入身分、以及身障身分將兩種模式內的個案進行分組（有無身分），以T檢定比較同一照護模式下，有特殊身分者相較於沒有特殊身分者，其日常生活功能退化速度是否有所差異。考量到個案可能兼具上述三種身分，透過複雜照顧程度的定義方式，以T檢定比較同一照護模式下，不同複雜照顧程度個案的日常生活功能改變速度是否有差異。最後再以獨立樣本T檢定，比較兩種模式同樣複雜照顧程度個案的平均日常生活改變速度是否有所差異。針對研究目的三，以多元邏輯斯迴歸分析，分別探討全部個案與部分整合照護模式以及完全整合照護模式中，使用長照2.0支付服務、使用非長照2.0支付服務、年齡、性別、獨居、身障、中低收福利身分、起始CMS等級等多項因子對於個案日常生活功能退化速度的影響。分別探討全部個案與部分整合照護模式以及完全整合照護模式中，影響日常生活功能變化速度的因子。 研究結果： 針對研究目的一，在兩種不同整合程度的照護模式對象中，完全整合照護模式的個案男性比例較高（部分整合照護模式男性15.9%，完全整合照護模式男性47.2%，P< 0.001），經濟上屬於中低收入身分的比例也較高（部分整合照護模式中低收22.6%，完全整合照護模式43.2%，P< 0.001）。其他在年齡、起始長照失能CMS等級、獨居、身障、外傭使用上均無顯著差異。兩組在持續照護時間上完全整合照護模式較部分整合照護模式長（部分整合照護模式549.01天，完全整合照護模式842.09天，P=0.053），但未達統計顯著差異。在服務使用上，部分整合照護模式傾向於接受長照2.0支付之服務為主，使用非長照2.0支付之服務之比例上相較於完全整合照護模式顯著較少（部分整合照護模式有使用非長照2.0支付之服務4.8%，完全整合照護模式70.4%， P< 0.001）。其中居家醫療的使用比例也有顯著差異（部分整合照護模式居家醫療使用2.4%，完全整合照護模式38.4%， P< 0.001）。此結果顯示在臺北市立聯合醫院中興院區完全整合照護模式相較於部分整合照護模式更能連結多元的長照服務個案接受非長照2.0支付之服務。針對研究目的二，兩種不同整合程度照護模式下的日常生活功能改變速度並沒有顯著的差異。以複雜照護的風險因子分組，包含獨居、中低收、或是身障，在部分整合照護照護模式組內的比較結果顯示，獨居比起非獨居者會有較快的退化速度（部分整合照護模式獨居vs 非獨居之ADL改變速度:0.83分/月，0.31分/月， P< 0.001），中低收入者比起非中低收者也會有較快的退化速度（中低收vs. 非中低收之ADL改變速度:0.49分/月 vs.－0.38分/月， P=0.04）。身障身分則沒有顯著差異。在完全整合照護模式中，獨居、中低收入身分有無者的日常生活功能變化速度均沒有顯著差異。以複雜照顧程度分組比較個案日常生活功能退化速度，結果顯示部分整合照護模式中同時具有獨居且中低收且身障身分的個案較沒有獨居或中低收或身障的個案為快（部分整合照護模式複雜照顧程度最高與最低組之日常生活功能退化速度為:0.98分/月 vs.0.44分/月， P=0.04）。完全整合照護照護模式中，複雜照顧程度1～3分組間個案日常生活功能退化速度相較於0分組均無顯著差異。此結果顯示完全整合照護模式對於複雜照顧程度高的族群可能有正向的幫助。針對研究目的三，多元邏輯式迴歸分析結果顯示，全部個案的日常生活功能顯著受到非長照支付服務的使用、獨居、以及起始CMS等級的影響，其中有使用非長照服務每個月會增加日常生活功能0.643分（P=0.010），具備有獨居身分者每個月下降0.761分（P<0.001），起始CMS等級每增加一級個案的日常生活功能每月下降0.27分（P<0.001）。其中獨居與CMS起始等級的影響主要來自於部分整合照護模式的個案，非長照2.0支付之服務的效益來自完全整合照護模式的個案。 結論： 完全整合照護模式提供個案較多元的服務介入，其中在非長照2.0支付之服務有顯著的差異。而非長照2.0支付之服務項目大部分來自使用居家醫療。對獨居、中低收、及身障身分因子增加的個案，單純使用部分整合照護模式的個案日常生活功能退化速度會比完全沒有這些身分的個案來的顯著增加。完全整合照護模式對於複雜照顧程度較高的個案可以抵銷此狀況讓其日常生活功能不會快速退化。多項式分析的結果也顯示，非長照支付對於個案日常生活功能有顯著增加的效果，獨居與起始CMS等級對日常生活功能則是顯著下降。建議政策應該針對高度複雜照顧或是失能程度較高的個案，評估連結多元的服務（包含居家醫療或其他社區非政府組織公益服務）的可能，建立評估複雜照顧需求的個案，提供高度整合的照護模式，避免個案日常生活功能的快速退化。

Background and Objectives: In response to the rapid aging of Taiwan's population, Taipei City Hospital has implemented three community-based integrated care programs since 2017. Based on previous classifications of care integration mechanisms, these programs are categorized into two models: (1) Partially Integrated Case Management Model – represented by the Ministry of Health and Welfare’s Long-Term Care (LTC) 2.0 Tier A Community-Based Care Program, which primarily provides four types of LTC 2.0-funded services: personal and professional care, transportation, assistive devices, and home environment modifications. Although it features financial integration and case management, it typically serves 100–120 clients and rarely incorporates home-based medical care or non-governmental resources. (2) Fully Integrated Case Management Model – commissioned by the Department of Social Welfare or the Department of Health, this model not only integrates LTC 2.0 services but also actively connects with home-based medical care and non-LTC-funded services. Prior studies using the National 10-Year LTC Plan data suggested that home-based medical and personal care (MpC) services were more effective in improving functional outcomes than other service groups, while community-based care resulted in greater functional decline. However, those studies focused on service types rather than the degree of integration. This study aims to explore how different levels of integration affect clients' functional status by comparing two care models implemented at Taipei City Hospital.

The study tests the hypothesis that a fully integrated care model better maintains clients’ functional ability than a partially integrated model. The objectives are: 1. To compare client characteristics and service utilization between the two care models. 2. To evaluate differences in the rate of functional decline between the two care models. 3. To identify factors influencing functional decline within each model.

Methods: This longitudinal secondary data analysis used records from the Ministry of Health and Welfare's Care Service Management Platform and Taipei City Hospital's internal database. The study included 541 clients enrolled between January 1 and December 30, 2020, in the Department of Community Integrated Care at the Zhongxing Branch. Clients were excluded if data were incomplete, lacked follow-up, or did not undergo re-evaluation. Clients were categorized into two groups: partially integrated care (n=416) and fully integrated care (n=125). For Objective 1, independent sample t-tests, chi-square tests, and ANOVA were used to compare baseline characteristics, the number of LTC 2.0 and non-LTC 2.0 services used, and the use of home-based medical care. For Objective 2, differences in the average rate of functional decline were analyzed across care models. Subgroup analyses were conducted within each model based on client risk factors (living alone, low income, and disability status). A complexity score (0–3) was calculated based on the presence of these factors, and differences in functional decline were assessed using t-tests. Objective 3 employed multiple linear regression to examine the impact of various factors—including service types, demographics, and baseline CMS level—on functional decline.

Results: Regarding Objective 1, clients in the fully integrated care group had a significantly higher proportion of males (47.2% vs. 15.9%, p < 0.001) and low-income individuals (43.2% vs. 22.6%, p < 0.001). No significant differences were found in age, initial CMS level, living alone, disability, or use of foreign caregivers. The duration of care was longer in the fully integrated group (842.09 vs. 549.01 days, p = 0.053). Use of non-LTC 2.0 services was significantly higher in the fully integrated group (70.4% vs. 4.8%, p < 0.001), as was the use of home-based medical care (38.4% vs. 2.4%, p < 0.001), indicating more diverse service connections. For Objective 2, no significant difference was found in the average rates of functional decline between the two models. However, in the partially integrated model, clients living alone or with low-income status showed faster declines in activities of daily living (ADLs) than those not living alone (0.83 points/month vs. 0.31 points/month, p < 0.001). In the fully integrated model, no such differences were observed across risk factors. Clients with a complexity score of 3 (all three risk factors) in the partially integrated model experienced the fastest decline (−0.98 vs. −0.44 points/month, p = 0.04), while no significant differences were found in the fully integrated group. These findings suggest that full integration may mitigate functional decline among high-risk clients. For Objective 3, regression analysis showed that use of non-LTC 2.0 services (+0.643 points/month, p = 0.010) was significantly associated with improved function, while living alone (−0.761, p < 0.001) and higher baseline CMS levels (−0.27 per level, p < 0.001) were associated with greater decline. The impact of living alone and CMS level was more evident in the partially integrated group, while the benefit of non-LTC 2.0 services was mainly seen in the fully integrated group.

Conclusions: The fully integrated care model provided more diverse interventions, especially non-LTC 2.0-funded services, primarily through home-based medical care. Among clients with complex care needs (living alone, low-income, or disabled), those in the partially integrated model experienced faster functional decline, while those in the fully integrated model maintained function more effectively. Multivariate analysis confirmed the protective effect of non-LTC-funded services and the detrimental impact of social isolation and higher initial disability. These findings support the need for highly integrated care models, especially for high-complexity clients, to prevent rapid functional deterioration. Policymakers should consider expanding access to non-LTC services and home-based medical care for complex care need populations.