透過您的圖書館登入
IP:52.14.126.74
  • 期刊

Mathematical Coupling of Data in Periodontal Research: Correlation between Baseline Disease Severity and Treatment Effects in Guided Tissue Regeneration

牙周病學研究的數據資料當中所發生的"數學耦合"現象:用引導組織再生手術治療骨內缺損時,治療前疾病的嚴重性和治療效果的相關性

若您是本文的作者,可授權文章由華藝線上圖書館中協助推廣。

摘要


過去不少的牙周病學文獻應用統計學上的迴歸或線性相關等方法指出,治療前牙周疾病的嚴重性,例如囊袋深度或附連高度,和治療效果有明顯的正相關性。換言之,治療前牙周囊袋越深,治療後囊袋減少越多;治療前附連高度破壞越多,附連高度增加越多。然而,應用統計學上迴歸或線性相關等方法,來證明一個變數的變化和其原來的數值的相關性,犯了兩個統計學上的謬誤:一個是向平均值迴歸,另一個是數學耦合。數學耦合會使得這些統計學方法背後的假設無效,並且產生誤導。本研究用歐德氏所提出的改良方法,重新針對過去引導組織再生手術治療骨內缺損的文獻,加以統計學的分析。結果顯示數學耦合的確會造成誤導,而用引導織再生手術治療骨內缺損時,治療前疾病的嚴重性和治療效果的相關性,事實上比過去牙周病學文獻所聲稱的來的低很多。

並列摘要


Pervious literature showed that the baseline disease severity measured by clinical and radiographic parameters such as probing pocket depth, clinical attachment level, intrabony defect depth etc was strongly associated with the treatment outcome. However, regressing/correlating the changes of the variables on their baseline values had two methodological flaws: one regression toward the mean; the other is mathematical coupling. Mathematical coupling between variables distorts their relationships, and the statistical analysis of testing the null hypothesis the coefficient of correlation or the slope of regression is zero may accordingly be erroneous. To investigate if the baseline disease severity was associated with the treatment outcome of intrabony defects using the principle of guided tissue regeneration as some studies indicated, the data available in previous periodontal literature were reanalyzed by examining two clinical parameters: probing pocket depth and clinical attachment level and using Oldham’s correction for adjusting the mathematical coupling. Results showed that mathematical coupling caused spurious correlation between the baseline disease severity and the treatment effects. Previous evidence based on inappropriate statistical analysis needs to be critically reviewed and cautiously interpreted.

延伸閱讀