Table 3.

Adjusted incidence rate ratios (aIRR) of intervention versus baseline periods for the outcome CLABSI with any pathogen stratified by incidence densities (ID) of CLABSI at baseline < or ≥ 0.8 CLABSI/1000 CL-days according to study group (post-hoc analysis)

Outcome/group	Chlorhexidine			Octenidine			Routine care (control arm)
	aIRR (95% CI)	P value	P value (Type III)	aIRR (95% CI)	P value	P value (Type III)	aIRR (95% CI)	P value	P value (Type III)
CLABSI with any pathogen
ICUs with incidence density of CLABSI at baseline < 0.8 CLABSI/1,000 CL days	n = 11 ICUs			n = 13 ICUs			n = 12 ICUs
Baseline period	1 = reference			1 = reference			1 = reference
Intervention period	0.88 (0.42–1.86)	0.7390	0.7373	2.59 (1.61–4.17)	< .0001	0.0383*	0.67 (0.3–1.52)	0.3390	0.3626
Mechanical ventilation use (per 1%)	1.01 (0.99–1.03)	0.4830	0.4275	1.01 (0.98–1.04)	0.3760	0.3927	0.99 (0.96–1.03)	0.7080	0.7800
Length of stay (per 1 day)	0.74 (0.53–1.04)	0.0830	0.1320	1.05 (0.65–1.68)	0.8520	0.8602	1.13 (0.93–1.38)	0.2070	0.3666
ICUs with incidence density of CLABSI at baseline ≥ 0.8 CLABSI/1000 CL days	n = 13 ICUs			n = 11 ICUs			n = 12 ICUs
Baseline period	1 = reference			1 = reference			1 = reference
Intervention period	0.61 (0.42–0.88)	0.0080	0.0226*	0.98 (0.58–1.64)	0.9380	0.9365	1.06 (0.63–1.78)	0.8210	0.8242
Mechanical ventilation use (per 1%)	1.02 (1–1.03)	0.0160	0.0912	1 (0.99–1.02)	0.8200	0.8405	1.01 (0.99–1.02)	0.4090	0.4278
Length of stay (per 1 day)	0.95 (0.85–1.06)	0.3540	0.3631	1.09 (1.02–1.17)	0.0100	0.2590	1.14 (1.02–1.27)	0.0250	0.0505

CLABSI central line associated bloodstream infection. 95% CI 95% confidence interval, aIRR adjusted incidence rate ratios. Separate GEE models were based on monthly aggregated data, negative binomial distribution, accounted for clustering effects and were calculated with the log number of CL days as offset variable. Parameters considered in all models were mechanical ventilation use and length of stay. *P values (type III test) < 0.05 were considered significant