Evaluating Regional Nursing Home Antibiograms to Advance Stewardship at 233 Skilled Nursing Facilities in Georgia, USA

Background: U.S. skilled nursing facilities (SNFs) must have antibiotic stewardship programs to receive Medicare; this requires access to summary susceptibility data (antibiograms).  Most SNFs test too few clinical samples to generate antibiograms by CLSI standards. We evaluated approaches to combining data across SNFs. 

Methods: Unsuppressed susceptibility testing results from pathogens recovered from urine specimens submitted in 2015-2016 to a regional referral laboratory servicing SNFs in GA were accessed. Study facilities were limited to GA SNFs (via linkage to CMS nursing home compare). No. and % testing susceptible (no. T, %S) of K. pneumoiae, E. coli, and P. mirabilis to cefazolin, levofloxacin, nitrofurantoin, TMP/Sulfa, and ceftriaxone were pooled overall, by geographic area, year, and other facility-characteristics. P. aeruginosa testing to cefazidime, imipenem, and piperacillin/tazobactam (P/T) were also evaluated. Differences in %S between stratum were considered clinically important if difference were >10 percentage points.

Results: Of 345 providers, 233 were confirmed as GA SNFs and were categorized by geography as Central (87 SNFs), 8 metro-Atlanta counties (Atlanta Area, 56 SNFs), North (46 SNFs), or Southern (44 SNFs). No. T didn’t vary by year, but varied slightly by region: range for E. coli was 275-1010, K. pneumonia was 101-409, P. mirabilis 133-439, and P. aeruginosa 35-160. No. T was lowest in North and highest in Atlanta Area. Overall %S was poor to levofloxacin, better for TMP/Sulfa and cefazolin (Table).  %S didn’t vary by year. Regional differences were minor: of 17 drug-bug combinations, only 5 has clinically important differences, 3 of which were among P. aeuginosa when testing was limited to 67 isolates in outlier region. Only the South had clinically worse susceptibility of K. pneumoniae to nitrofurantoin (36% vs. 47%).

Conclusion: Overall, SNFs in Georgia had remarkably similar susceptibility patterns when grouped by geography among common urinary pathogens, except when <75 isolates were tested. Antibiogram data can be combined across facilities in a region to provide SNFs with a reasonable antibiograms for stewardship. Further study is ongoing to assess benefits of bedsize or length of stay based antibiograms.