ISSN: 0022-1198
CODEN: JFSCA
Page Count: 6

A New Perspective in the Estimation of Postmortem Interval (PMI) Based on Vitreous [K⁺]
Costas, E
Institute of Legal Medicine, University of Santiago de Compostela,

Muñoz, JI
Institute of Legal Medicine, University of Santiago de Compostela,

Suárez-Peñaranda, JM
Institute of Legal Medicine, University of Santiago de Compostela,

Concheiro, L
Institute of Legal Medicine, University of Santiago de Compostela,

Rodríguez-Calvo, MS
Institute of Legal Medicine, University of Santiago de Compostela,

Otero, XL
Statistics Unit, Faculty of Medicine, University of Santiago de Compostela,

Miguéns, X
Institute of Legal Medicine, University of Santiago de Compostela,

(Received 4 February 2000; accepted 24 April 2000)

Abstract

The relation between the potassium concentration in the vitreous humor, [K⁺], and the postmortem interval has been studied by several authors. Many formulae are available and they are based on a correlation test and linear regression using the PMI as the independent variable and [K⁺] as the dependent variable. The estimation of the confidence interval is based on this formulation. However, in forensic work, it is necessary to use [K⁺] as the independent variable to estimate the PMI.

Although all authors have obtained the PMI by direct use of these formulae, it is, nevertheless, an inexact approach, which leads to false estimations. What is required is to change the variables, obtaining a new equation in which [K⁺] is considered as the independent variable and the PMI as the dependent. The regression line obtained from our data is [K⁺] = 5.35 + 0.22 PMI, by changing the variables we get PMI = 2.58[K⁺] − 9.30. When only nonhospital deaths are considered, the results are considerably improved. In this case, we get [K⁺] = 5.60 + 0.17 PMI and, consequently, PMI = 3.92[K⁺] − 19.04.

Keywords:
forensic science, forensic pathology, postmortem interval, vitreous humor, potassium, regression

Paper ID: JFS14950J
DOI: 10.1520/JFS14950J
ASTM International is a member of CrossRef.