論質計酬對於思覺失調症患者急診及急性再住院的影響

The impact of pay for performance on emergency department visits by and acute readmissions of patients with schizophrenia

目標：全民健康保險2010年1月1日實施思覺失調症論質計酬方案，本研究旨在探討有無加入論質計酬方案，比較其急診及6個月內急性再住院發生的風險及其相關因素。方法：資料來源為全民健康保險研究資料庫，擷取2009年至2010年固定就診之思覺失調症重大傷病患者（6,713人），依其所屬院所分為有加入及未加入論質計酬方案兩組，採傾向分數（propensity score）1：1配對後，應用差異中的差異法（Difference-in-Difference）設計，以條件式羅吉斯迴歸模型（conditional logistic regression model）進行分析。結果：急診利用方面，加入論質計酬方案組在方案實施前為12.21%，方案後降為12.01%（p<0.05）；未加入論質計酬方案組在方案前為13.3%，方案後降為11.80%（p<0.05）。6個月內急性再住院加入論質計酬方案組在方案前、後皆為1.71%（p>0.05）；未加入論質計酬方案組在方案前為2.01%，方案後降為1.19%（p<0.05）。控制其他變項後，在條件式羅吉斯迴歸模型中之交互作用變項急診利用之勝算比（OR）為1.15倍（p>0.05）、急性再住院之勝算比（OR）為1.89（p>0.05），顯示加入論質計酬方案組在加入方案後急診利用及再住院下降幅度比未加入論質計酬方案組少。結論：患者經過配對後，院所加入思覺失調症論質計酬方案第一年，降低固定就診病患之急診利用及6個月內再住院之成效不佳。

Objectives: On January 1, 2010, Taiwan began implementation of the Schizophrenia Pay-for-Performance (P4P) program. The objective of this study was to investigate the impact of the P4P program on emergency department (ED) visits and acute readmissions within six months after discharge. Methods: The research data were obtained from the National Health Insurance Research Database. The study sample was made up of schizophrenic patients in the catastrophic illness registry for 2009-2010, whose regular visit institutions did or did not join the P4P program (n=6,713). The groups were matched using propensity scores in a ratio of 1:1. The conditional logistic regression model with difference-in-differences approach was then used. Results: For the P4P patients, the number of ED visits was 12.21% before the program and 12.01% after it (p< 0.05). For the non-P4P patients, the number of ED visits was 13.3% before the program and 11.80% after it (p < 0.05). Acute readmission utilization of the P4P patients both before and after the program was 1.71% (p > 0.05). In the non-P4P patients, acute readmission utilization was 2.01% before the program and 1.19% after it (p < 0.05). After adjustment for other variables, the interaction variable of ED visits in the conditional logistic regression model revealed that the odds ratio (OR) was 1.15 (p > 0.05) and that of acute readmission utilization was 1.89 (p>0.05). This finding showed that the reduction in the P4P patients was less than that in the non-P4P patients. Conclusions: When patients were matched, hospitals participating in the P4P program for patients with schizophrenia were ineffective in reducing emergency care and readmission within 6 months after discharge in the first year.