Table 3.

Determinants of Coefficient of Variation: Multivariable Logistic Regression Model to Calculate “r” ^a

	Age	p value	Modified APACHE II score	p value	Tests	p value	DM	p value
PRE NON	0.033	.2070	0.328	<.0001	0.235	<.0001	n/a
PRE DM	0.040	.1380	0.154	.0077	0.095	.1495	n/a
GC NON	0.150	<.0001	0.285	<.0001	0.104	.0186	n/a
GC DM	0.009	.6689	0.220	<.0001	0.006	.2906	n/a
All PRE	0.056	.0633	0.291	<.0001	0.201	<.0001	0.218	<.0001
All GC	0.142	<.0001	0.277	<.0001	0.076	.2109	0.259	<.0001
All NON	0.103	<.0001	0.301	<.0001	0.151	<.0001	n/a
All DM	0.019	.2872	0.196	<.0001	0.033	.8338	n/a
All	0.108	<.0001	0.281	<.0001	0.119	<.0001	0.242	<.0001

^aAge, modified APACHE II score (age component deleted from the total score), tests (the number of glucose tests obtained during ICU stay), and presence of DM were found to be significantly associated with CV using univariable logistic regression analysis. This table displays the regression coefficients and associated p values for these factors using a multivariable logistic regression analysis. Different subgroups were created on the basis of admission date (before or after the implementation of glycemic control protocol) and presence or absence of DM.