Table 3.

Assessment of the appropriateness of antimicrobial therapy for microbiologically documented infections

Parameter	Appropriate AT	Inappropriate AT	P
	(n = 203)	(n = 58)
AT protocol available in the ICU	79 (61.1%)	35 (60.3%)	0.91
Timing of new AT prescription
Day shifts	97 (47.8%)	30 (51.7%)	0.59
Out-of-hours	106 (52.2%)	28 (48.3%)
Category of MD prescriber
Fellow	17 (8,4%)	7 (12.1%)	0.88
Senior physician	148 (72.9%)	41 (70.7%)
Medical team decision	38 (18.7%)	10 (17.2%)
Time of initiation of new AT
Suspicion of infection	120 (59.1%)	29 (50.0%)
Gram-stained direct examination available	65 (32.0%)	12 (20.7%)	<0.0001
Microbiologic identification available	18 (8.9%)	3 (5.2%)
Susceptibility testing available	0	14 (24.4%)
Change of AT
None	107 (52.7%)	14 (24.1%)
Gram-stained direct examination available	11 (5.4%)	4 (6.9%)	0.001
Microbiologic identification available	32 (15.8%)	11 (19.0%)
Susceptibility testing available	53 (26.1%)	29 (50.0%)
Number of AT changes	0.5 ± 0.6	0.9 ± 0.7	0.05
Non-microbiologic reason for AT change	38 (18.7%)	10 (17.2%)	0.79
Clinical worsening	4 (2.0%)	1 (1.7%)
New site of infection	5 (2.5%)	4 (6.9%)
Aminoglycoside stopped	23 (11.3%)	4 (6.9%)
AB side effect	3 (1.5%)	1 (1.7%)
De-escalation	26 (12.8%)	4 (6.9%)

Data are presented among the patients receiving new AT (n = 509), and expressed as mean ± SD or as number (proportion). AT, antibiotic therapy; ICU, intensive care unit; MD, medical doctor.