Screening for older inpatients at risk for long length of stay: which clinical tool to use?

BMC Geriatrics

Table 2 Multiple regressions showing the association of length of stay in days (dependant variable) with the PRISMA-7 high risk level and the 6-item BGA risk levels (i.e., low, moderate and high) (independent variable, separated model for the PRISMA-7 and the 6-item brief geriatric assessment) adjusted with reasons for admission (n = 166)

	Linear regression*			Logistic regression^†			Cox regression*
	ß	[95% CI]	P-value	OR	[95% CI]	P-value	HR	[95% CI]	P-value
PRISMA-7 Abnormal score	3.3	[−2.9;9.6]	0.295	2.6	[0.8;7.9]	0.107	1.2	[0.8;1.9]	0.398
A priori 6-item BGA levels of risk
Low	−4.9	[−9.3;-0.5]	0.030	Ref.			Ref.
Moderate	−0.3	[− 4.6;4.0]	0.887	0.4	[0.7;3.2]	0.279	1.3	[0.9;1.8]	0.195
High	9.9	[4.0;15.9]	0.001	1.1	[1.1;8.5]	0.028	2.1	[1.3;3.4]	0.004

PRISMA-7: the seven questions “Program of Research on the Integration of Services for the Maintenance of Autonomy”
BGA Brief geriatric assessment
ß Coefficient of regression beta
OR Odds ratio
HR Hazard ratio
CI Confidence interval
Ref Referent condition
*Length of stay used as a continuous dependant variable
^†Length of stay used as a discontinuous variable defined as the highest tertile of lengths of stay (i.e., > 13 days)
All models adjusted with reason for admission, including mobility disorders (i.e., gait and/or balance disorders and/or fall, defined as unintentionally coming to rest on the ground, floor, or other lower level), neuropsychiatric disorders (i.e., delirium, dementia or behavioral disorders), organ failure (i.e., acute organ decompensation) or social issue (i.e., absence of symptoms of acute diseases combined with an acute increase of the use of formal and/or informal home and social services, leading to an inability to stay at their place of residence for life)
Coefficient of regression beta, odds ratio, hazard ratio and P-value significant (i.e.; < 0.05) indicated in bold

ISSN: 1471-2318