Effect of sedative premedication with oral midazolam on postanesthesia care unit delirium in older adults: a secondary analysis following an uncontrolled before-after design

Table 3 Binary logistic regression analysis

	OR	95% CI	P
First step
Midazolam (reference: no midazolam)	0.720	0.128; 4.067	0.710
Year of enrollment (per year increase)	0.805	0.429; 1.512	0.500
Propofol (reference: sevoflurane)	2.653	1.395; 5.044	0.003
ORP (reference: RARP)	1.061	0.583; 1.934	0.846
Age (per year increase)	1.098	1.046; 1.152	< 0.001
ASA physical status (reference: ASA III)
ASA I	0.907	0.499; 1.649	0.749
ASA II	1.286	0.595; 2.778	0.523
Estimated blood loss (per ml increase)	1.165	0.828; 1.638	0.381
Duration of surgery (per min increase)	2.028	0.657; 6.260	0.219
Sufentanil (per μg kg⁻¹ h⁻¹ increase)	1.014	1.002; 1.025	0.018
Final step
Midazolam (reference: no midazolam)	0.847	0.164; 4.367	0.842
Year of enrollment (per year increase)	0.829	0.458; 1.499	0.535
Propofol (reference: sevoflurane)	2.805	1.497; 5.256	0.001
Age (per year increase)	1.102	1.050; 1.156	< 0.001
Sufentanil (per μg kg⁻¹ h⁻¹ increase)	1.014	1.005; 1.024	0.004

Factors associated with postanesthesia care unit delirium were analyzed with binary logistic regression. Variables were eliminated stepwise backwards with the first and the final step presented. ‘Midazolam’ and ‘year of enrollment’ were forced into the model. ‘Duration of surgery’ and ‘estimated blood loss’ were transformed to their binary logarithm (ln(x)/ln(2)). RARP robot-assisted radical prostatectomy. ORP open retropubic radical prostatectomy

ISSN: 2047-0525