| 摘要: | 背景與目的：攝護腺癌為男性最常見之惡性腫瘤，近年來罹患人數呈增加趨勢，且於民國99年為第七大癌症死因。電腦化臨床決策支援系統(CDSS)能否促進攝護腺癌醫療照護品質仍屬未知。本研究欲比較有無使用CDSS於介入前後攝護腺抗原(PSA)之差異，並分析CDSS使用與病患死亡、復發及放射線治療併發症之相關性。
方法：採類實驗設計，以某醫學中心746位攝護腺癌患者之2007年至2010年回溯性資料，進行實驗組(使用CDSS共140人)與對照組(無使用CDSS共606人)之雙組前後測比較(nonequivalent comparison guoup, pretest - posttest design)。醫療照護結果品質生化指標為PSA值，而醫療照護結果品質病患指標為死亡、復發及放射線治療併發症。以SAS 9.1統計軟體進行描述性統計、卡方檢定、邏輯斯迴歸分析及重複量測一般線性模式(Repeated measures GLM)。
結果：使用CDSS之病患介入前後PSA數值無顯著差異(p=0.117)；無使用CDSS之病患介入前後PSA數值無顯著差異(p=0.852)；CDSS使用有無與介入前後病患PSA數值未呈現顯著差異(p=0.478)。無使用CDSS之病患於介入前PSA平均值為157.83，介入後PSA平均值為163.28，增加5.45。有使用CDSS之病患於介入前PSA平均值為163.91，介入後PSA平均值為84.16，下降79.75。CDSS使用之顯著相關因素為病患年齡、格里森分級(Gleason's Score)、臨床分級、家族史、治療方式及共病症。CDSS使用與病患死亡統計上達到顯著意義(p=0.0446)。CDSS使用與病患復發及放射線治療併發症未呈現統計上之顯著相關。
結論：CDSS使用於PSA重覆量測值未呈現顯著效果；惟無使用CDSS之病患介入後PSA平均值較介入前為上昇，使用CDSS之病患介入後PSA平均值較介入前為下降，其效果仍值得注意。CDSS使用與病患死亡達顯著相關。未來研究可針對CDSS對其他照護結果品質指標之影響進行探索，以建立專屬本國人攝護腺癌臨床照護品質結果之篩檢與預測模型。
Background: Prostate cancer, the seventh leading cause of cancer death in 2010 in Taiwan, is the most common male malignancy recently. Its prevalence is escalating. It is still unknown that whether computerized clinical decision support systems (CDSS) can improve the quality of prostate cancer care. The present research sought to investigate the effect of CDSS on the outcome quality of care, including PSA, mortality, recurrence, and complication of radiotherapy, among prostate cancer patients.
Methods: A quasi-experimental design with repeated PSA measures for the pretest and posttest periods for a total of 4 years was used. This study analyzed the retrospective data of 746 patients with prostate cancer collected in a medical center during 2007-2010. The subjects were divided into the intervention group (140 CDSS users) and control group (606 non-users). The biochemical PSA serves as an outcome quality indicator. The patient indicators of outcome quality include mortality, recurrence, and the complications of radiotherapy. The descriptive statistics, Chi-square, logistic regression, and repeated measures general linear model (GLM) were performed by SAS 9.1.
Results: The between-group PSA differences in the pretest-posttest comparison were not significant (p=0.478; p=0.117 for the pretest-posttest difference among the CDSS users; p=0.852 for the non-users). However, the CDSS users experienced a 79.75 decrease in PSA, with a contrast to a 5.45 increase in PSA among the non-users. The factors significantly associated with the CDSS use were age, Gleason's Score, clinical stage, family history, treatment, and comorbidity. There was a significant association between the CDSS use and mortality (p=0.0446). No significant effects of CDSS on recurrence and radiotherapy complications were found.
Conclusions: Although no significant effect of CDSS on PSA was found, that the decreased PSA value in the pretest-posttest difference among CDSS users deserves more attention. Besides, this study found a significant effect of CDSS on mortality. The effects of CDSS on other quality indicators might be examined in further studies, in endeavoring to establish the screening and prognostic models of medical care quality for prostate cancer in Taiwan. |
