
Methods: We analyzed an electronic research dataset of Becton, Dickinson & Company from 346 USA hospitals in 2015. All non-duplicate EC, KP, and PM isolates (first isolate of a species per 30 day period) from all sources were categorized as ESBLs if confirmed as ESBL-positive per commercial panels or intermediate/resistant to either ceftriaxone, cefotaxime, ceftazidime or cefepime. Positive isolates were categorized into 3 settings by the specimen collection time: a) Admission: within 3 days of an inpatient admission and no previous admission within 14 days; b) Hospital-onset: 3 days post-admission or within 14 days of discharge; c) Ambulatory (neither a or b). Geographic regions were classified into NHSN categories (Sievert, 2013). Region 1, 7, & 8 were grouped into “Other” due to small number of hospitals. We conducted pairwise comparisons between regions using the region with the overall lowest ESBL rate as the reference group.
Results: The overall ESBL rate was 6.9% (56718/820017). The ESBL rates for ambulatory, admission, and hospital-onset settings were 5.7% (36086/631846), 9.2% (9693/105846), and 13.3% (10939/82325), respectively. Compared to Region 10, all regions, except for “Other”, had significantly higher ESBL rates, ranging from 4.4% to 7.9% (all P<0.05) for ambulatory; from 7.8% to 12.6% (all P<0.01) for admission; and 11.0% to 18.7% (all P<0.01) for hospital-onset settings.
NHSN Region |
States |
# of Hospitals |
% ESBL |
||
Ambulatory |
Admission |
Hospital |
|||
2 |
NJ, NY, PR, VI |
30 |
7.1 |
12.6 |
18.7 |
9 |
AZ, CA, HI, Pacific Islands |
28 |
7.9 |
11.5 |
16.6 |
6 |
AR, LA, NM, OK, TX |
59 |
6.9 |
9.6 |
13.4 |
4 |
AL, FL, GA, KY, MI, NC, SC, TN |
98 |
6.2 |
9.2 |
13.2 |
5 |
IL, IN, MI, MN, OH, WI |
85 |
4.4 |
8.2 |
11.5 |
3 |
DE, DC, MD, PA, VA, WV |
13 |
5.2 |
7.8 |
11.0 |
10 |
AK, ID, OR, WA |
22 |
4.0 |
5.5 |
8.6 |
1, 7, 8 |
Other |
11 |
4.1 |
6.0 |
8.0 |
Overall |
346 |
5.7 |
9.2 |
13.3 |
Conclusion: ESBL rates were highest in the hospital setting with significant regional differences. The highest ESBL region had a rate approximately twice that of the lowest region in all three settings. Regional and setting differences in ESBL epidemiology should be considered when making empiric antibiotic treatment decisions.

D. D. Depestel,
Merck & Co. Inc.:
Employee
,
Salary
C. A. Deryke, Merck & Co. Inc.: Employee , Salary
S. Merchant, Merck & Co. Inc.: Employee , Salary
R. S. Johannes, Becton Dickinson: Employee , Salary
P. Moise, Merck & Co. Inc.: Employee , Salary
V. Gupta, Becton Dickinson: Employee , Salary