Methods: The TDH AU Survey is a quarterly assessment of the number of patients who received antibiotics during a 24-hour period, as a percent of daily hospital census. The metric used for NHSN is monthly DOT/days present. All available DOT from institutions reporting into the NHSN AU Option from October 2015-December 2016 was pulled and aggregated by quarter. TDH AU data reported during the same time period was also pulled. A matched data point was usable when data existed from both the TDH Survey and NHSN. Trend lines and coefficients of determination were created using Excel 2010 and SAS 9.4.
Results: Four Tennessee hospitals reported into both the NHSN AU Option and the TDH AU survey during the study period. From those institutions, there were 117 matched data points for nine antibiotics or groups of antibiotics. A positive correlation was observed in the primary analysis of all institutions’ reported antibiotic agents – r2=0.7947; p<0.0001. Variation existed among the nine different drug/classes, with the strongest correlation existing for anti-pseudomonal carbapenems (r2=0.8402, p<0.0001) and the weakest with respiratory fluoroquinolones (r2=0.0449; p=0.487). No strong influences were observed when data were analyzed by subgroups of quarters or institutions.
Conclusion: A positive correlation was found between the two AU metrics, indicating that the TDH AU Survey is a reasonable interim proxy of the NHSN AU data. It is frequently used by TN hospitals for evaluating individual institutional trends, but benchmarking institutions with it is not ideal. This illustrates the need for more hospitals in our state and nationwide to report into the NHSN AU Option.
Z. Gugkaeva, None
B. Stayer, None
C. Simpson, None
P. Talley, None
M. Kainer, None