1.
Patients
a.
Inclusion criteria
During the period from March 1st,
2009 to February 27th,
2011,
a patient wherein a PMCT was performed at Kameda Medical Center (Kamogawa,
Chiba,
JAPAN) was entered into our study. At Kameda Medical Center,
PMCT studies were performed on patients who received pathological autopsy following their death,
and patient's who deceased shortly after admission to our emergency room. In these former cases,
it was utilized as a screening method before autopsy and,
in the latter cases,
it was utilized to help determine cause of death.
b.
Exclusion criteria
The cases excluded from our study were 1) patients who didn't receive CPR,
2) patients younger than 60 years old,
3) patients with hepatic tumor,
4) patients who were under percutaneous cardiopulmonary support system (PCPS),
5) patients who had bowel necrosis,
6) patients with biliary tract infections,
7) traumatic patients.
2.
PMCT
Data collection protocol: Multislice CT was performed on a Philips Brilliance CT 64-slice unit (Royal Philips Electronics,
Netherlands). Axial slices were acquired with a collimation of 0.625mm x 64 detectors; 2 mm sections,
and 0.923 mm helical pitch by 120 kV 250 mA,
sagittal and coronal reformations were calculated as well. The PMCT scans took between 10 to 15 min.
3.
Image Analysis
A board-certified radiologist evaluated the CT scans. The cases which had round or tree-like shaped negative radio density in their liver were assessed as existing IHG.
Also IHG volume was classified as grade 0 to 3 according to our original semi-quantitative grading system (Figure.
2).
The presence of BD (stomach or small bowel or both distended with gases) were also recorded (Figure.
3).
4.
Data collection
Six attributes of patient's data were collected from medical records.
The attributes were,
whether their death were Bath related or not,
age,
gender,
interval time from the time from their death to PMCT study (Time),
interval time of CPR (CPR),
their body temperature on hospital arrival (BT).
5.
Statistical analysis
Weltch's t-tests were calculated between Bath related group and non-Bath related group by using R for windows version 2.10.0 (the R Development Core Team.
http://cran.r-project.org/). Also,
Fisher's exact test,
and Rigid analysis test were calculated. The p-value,
under 0.05,
was assessed as significant difference.