Stephens R J, Richards R G
J Forensic Sci. 1987 Mar;32(2):503-9.
Upon review of the literature, extensive disagreement was found as to the usefulness of vitreous humor potassium concentration as a predictor of the postmortem interval (PMI). A pilot study of 1427 cases was performed to address this problem. The requisite statistical analysis for the prediction of PMI is inverse prediction. The 95% inverse prediction interval was found to be approximately +/- 20 h. The linear regression equation for the data was y = 0.238 chi + 6.342, with a coefficient of determination (r2) of 0.374. This r2 value means that 62.6% of the variation of potassium is unaccounted for by the variation in PMI. Further studies are required to attribute this unaccounted variation to quantifiable factors. This would narrow the inverse prediction interval and enable vitreous potassium to be a useful aid in the prediction of PMI.
在查阅文献时,发现关于玻璃体液钾浓度作为死后间隔时间(PMI)预测指标的有用性存在广泛分歧。为解决这一问题,进行了一项包含1427例病例的初步研究。预测PMI所需的统计分析是反向预测。发现95%的反向预测区间约为±20小时。数据的线性回归方程为y = 0.238χ + 6.342,决定系数(r2)为0.374。这个r2值意味着钾变化的62.6%无法用PMI的变化来解释。需要进一步研究将这种无法解释的变化归因于可量化因素。这将缩小反向预测区间,并使玻璃体液钾成为预测PMI的有用辅助指标。