Table 2.

Patterns of antibiotic escalation and de-escalation, by timeliness of initial antibiotic therapy.

	Timely Appropriate Therapy (n = 438)	Delayed Appropriate Therapy (n = 26)	All Patients (n = 464) *
Until day five/discharge
De-escalation	159 (36.3%)	4 (15.4%)	163 (35.1%)
Escalation	19 (4.3%)	6 (23.1%)	25 (5.4%)
No change	135 (30.8%)	6 (23.1%)	141 (30.4%)
Unknown	125 (28.5%)	10 (38.5%)	135 (29.1%)
Until last day of treatment/discharge
De-escalation	240 (54.8%)	9 (34.6%)	249 (53.7%)
Escalation	42 (9.6%)	6 (23.1%)	48 (10.3%)
No change	154 (35.2%)	9 (34.6%)	163 (35.1%)
Unknown	2 (0.5%)	2 (7.7%)	4 (0.9%)

* Patients included in escalation/de-escalation analyses were required to have (1) ≥2 days of antibiotic therapy, (2) LOS of ≥3 days following in-hospital antibiotic therapy administration, and (3) no evidence of death within five days following initiation. As a result, overall n (n = 464) is lower than that of the overall hospital cohort prior to weighting (n = 517). Notes: All outcomes were measured from the end of day two alternatively to day five/discharge (whichever occurred first) and to the last day of treatment/discharge (whichever occurred first). Patients who did not receive at least five days of antibiotic therapy were excluded from these analyses. Patients who received ≥5 days of antibiotic therapy but did not have available data regarding escalation/de-escalation/unchanged were labelled as “Unknown”.