Nursing homes (NHs) are implementing antibiotic stewardship programs due to recent
federal regulations. We sought to evaluate if NH antibiotic stewardship policies change
antibiotic use in NH residents.
Antibiotic stewardship policy data were collected from nationally-representative samples
of NHs in 2013 and 2017. An aggregate antibiotic stewardship intensity index, ranging
from 0-100, was calculated from five antibiotic stewardship policy measures. Survey
data were merged with concurrent MDS 3.0 data and antibiotic use was defined as a
one-day prevalence from annual and quarterly assessments. Multivariable models, overall
and stratified by infection type, were used to estimate antibiotic use as a function
of antibiotic stewardship intensity, NH and resident characteristics, and time. The
total adjusted change in antibiotic use was calculated as the difference in predicted
use after setting time and stewardship intensity to 2013 values and then to 2017 values,
adjusting for 2017 characteristics. The isolated contribution of antibiotic stewardship
intensity was calculated as the predicted antibiotic use when time is set to 2013
and policy intensity to 2017 divided by the total adjusted change in antibiotic use.
Data from 584,540 resident assessments from 1,825 NHs were analyzed. On average, antibiotic
stewardship intensity in NHs increased significantly from 2013 to 2017 (µ=33.08 to
µ=67.94, p < 0.001). The observed prevalence of antibiotic use decreased by 2.9% (8.29%
to 8.05%, p=0.001). The total adjusted decline in antibiotic use given similar characteristics
was 5.3% from 2013 to 2017. Increases in policy intensity accounted for 37% of this
decline. Antibiotic stewardship intensity was negatively associated with antibiotic
use overall and in cases with active diagnoses of urinary tract infection or septicemia.
Improved antibiotic stewardship programs have resulted in a modest decline in antibiotic
use, suggesting that the full extent of the translation from antibiotic stewardship
policy to practice remains to be seen.