. Author manuscript; available in PMC: 2016 Nov 1.

Published in final edited form as: Psychosomatics. 2015 May 15;56(6):644–651. doi: 10.1016/j.psym.2015.05.002

Table 3.

Logistic regression results assessing the association of symptoms of delirium on postoperative day 1 with subsequent delirium on postoperative day 2.

	P value	Mean Stay (95% CI)	Estimate (95% CI)

Number of SSD Features
0 features (ref) (n=256)	---	5.25 (4.55, 5.94)
1 feature (n=64)	.146	6.36 (5.01, 7.71)	--
2 features (n=49)	<.001	8.76 (7.18, 10.33)
>2 features (n=231)	<.001	7.33 (6.61, 8.04)

Charlson Score	.037	-	.275 (.016, .534)

Length of Surgery (hours)	<.001		.943 (.774, 1.11)

Logistic regression was used with diagnosis of delirium on day 2 as the response variable and number of subsyndromal delirium features on day 1 as the explanatory variable (0 features is the reference group).

The features refer to the presence of at least one out of a possible 10 symptoms of delirium, as defined by the CAM (acute onset, inattention, fluctuation, disorganized thinking, altered level of consciousness, disorientation, memory impairment, perceptual disturbances, psychomotor agitation or retardation, altered sleep wake cycle).