台灣失智症病患不同病程之醫療費用及利用情形

研究目的：本研究目的將探討我國失智症之不同共病與病程與醫療利用及費用之相關性。茲簡述具體之研究目的如下：1. 以年齡、性別配對後，探討失智症與對照組之共病症罹病情形與差異；2. 分析不同共病對失智症醫療利用及費用之影響；3. 傾向分數配對共病後，比較失智症與對照組在不同病程的醫療利用與費用之差異。研究方法：資料來源為衛生福利部之衛福資料科學中心所提供的健保資料庫，研究對象為選取2002~2010年之40歲以上被診斷為失智症新發的病患，且必須於門診或住院曾有任一次「臨床失智症量表」或「簡式智能評估量表」或「認知功能篩檢量表」的醫令處置。非失智症對照組則為研究期間沒有失智症診斷碼的個案，使用 1：1 傾向分數配對，依年齡、性別、投保類別、都市化程度、共病等因子配對，選取最佳的對照組。統計方法部分，使用SAS 9.4 版進行描述性及推論性統計，描述性以平均數及標準差、樣本數與百分比呈現，推論性統計使用McNemar’s test檢定、成對樣本t檢定、Standardized differences檢定研究架構中各變項之相關性。醫療費用及利用部分，則使用廣義估計方程式，估算門急住診之利用與費用。而急診或住院的估算會進一步使用 GEE 之Two-Part Model 進行分析。研究結果：結果顯示，失智症病患將近9成合併有CCI(charlson comorbidity index)慢性共病，明顯高於年齡及性別配對後之對照組。不同共病對失智症病患醫療費用與利用的影響各有不同，影響最高的前三名分別是中度或重度肝臟疾病、後天免疫缺乏症候群，及轉移性腫瘤。失智症確診前一年之平均總醫療費用為NT$1,135,108元是對照組的2.2倍(NT $502,204)。失智症平均門、住、急診次數與住院天數分別為對照組的1.3倍、3.9倍、3.3倍及1.9倍，門、住、急診費用則為1.4倍、2.3倍及3.5倍。失智症確診後一年之失智症平均總醫療費用為NT$1,217,672元，是對照組的2.4倍(NT $517,621)。失智症平均門、住、急診次數與住院天數分別為對照組的1.4倍、4.1倍、4.4倍及2.7倍；門、住、急診費用則為1.7倍、2.4倍及3.5倍。失智症死亡前一年之失智症平均總醫療費用為NT$1,709,531元是對照組的2.8倍(NT $619,427)。失智症平均門、住、急診次數與住院天數分別為對照組的1.1倍、4.1倍、22倍、2.4倍；門、住、急診費用則為1.00倍、3倍及28倍。結論：失智症病患相對於對照組有較多共病，且共病會顯著增加其醫療費用與利用。失智症不論是確診前一年或確診後一年的醫療費用及利用都高於對照組；失智症死亡前一年的醫療費用及利用都高於對照組。

關鍵字

失智症；醫療費用；醫療利用；Two-Part-Model

並列摘要

Objectives: The objectives of this study are described as following: 1) to examine the risk of comorbidities of patients with dementia compared to age- and sex-matched controls.2) to estimate the effect of different comorbidities on healthcare utilization and expenditure in dementia patients; and 3) to compare the health utilization and expenditure of dementia patients with non-dementia matched controls based on the propensity score matching approach. Methods: This study use the National Health Insurance Claims data, provided by Health and Welfare Data Center, Ministry of Health and Welfare. The study time frame is selected mainly from 2002 to 2010, and the study group focuses on newly diagnosed patients with dementia who were aged 40 year-old and older. Additionally, we limited to the patients with at least an outpatient or inpatient claims of Mini-Mental Status Examination (MMSE) or Clinical Dementia Rating (CDR) or Cognitive Abilities Screening Instrument (CASI). The potential controls were the individuals with no diagnostic claims of dementia during the study period. Propensity score matching of 1:1 was used to select the best match controls based on age, gender, insurance category, urbanization and comorbidities. All statistical analysis was done by using SAS version 9.4. The descriptive statistics included mean, standard deviation and percentage and the inferential statistical analysis included McNemar's test, paired t-test, and standardized differences. General linear model was used to estimate the health care utilization and expenditure. For the data with excessive zero, Two-Part Model of GEE analysis was performed. Results: The result showed that nearly 90% of patients with dementia had a diagnosis of chronic diseases, which was significantly higher than non-dementia matched controls. Different comorbidities had different effect on the healthcare utilization and expenditure in dementia patients. Comorbidities had the highest impact were moderate to severe liver disease, AIDS and metastatic solid tumors. The mean total healthcare expenditure one year before dementia diagnosis in study group (NT$1,135,108), was 2.2 times higher than the controlled group (NT$502,204). The mean outpatient, inpatient and emergency visits and length of stay in study group was higher than controlled group by 1.3, 3.9, 3.3 and 1.9 times, respectively; while the expenditure of outpatient, inpatient and emergency department in study group was higher than controlled group by 1.4, 2.3 and 3.5 times, respectively. The mean total healthcare expenditure after one year after diagnosis in study group (NT$1,217,672) was 2.4 times higher than controlled group (NT $517,621). The mean outpatient, inpatient and emergency visits and length of stay in study group was higher than controlled group by 1.4, 4.1, 4.4 and 2.7 times, respectively; while outpatient, inpatient and emergency department expenditure in study group was higher than controlled group by 1.7, 2.4 and 3.5 times, respectively. The mean total healthcare expenditure before one year of the death in study group (1,709,531) was 2.8 times higher than the controlled goup (NT $619,427). The mean outpatient, inpatient and emergency visits and length of stay in study group was higher than controlled group by 1.1, 4.1, 22 and 2.4 times, respectively; while outpatient, inpatient and emergency department expenditure in study group was higher than controlled group by 1.0, 3.0 and 28 times, respectively. Conclusions: Dementia patients had higher comorbidities than controlled group, moreover, comorbidities were associated with an increasing effect on healthcare utilization and expenditure. Healthcare utilization and expenditure are significantly higher in study group than in controlled group in different disease status.

並列關鍵字

Dementia；Healthcare utilization；Healthcare expenditure；Two-Part Model

參考文獻

Charlson, M. E., Pompei, P., Ales, K. L., & MacKenzie, C. R. (1987). A new method of classifying prognostic comorbidity in longitudinal studies: Development and validation. Journal of Chronic Diseases, 40(5), 373-383.

中文部分