Use of electronic pharmacy transaction data and website development to assess antibiotic use in nursing homes

Days of Antibiotic Spectrum Coverage (DASC): A Novel Metric for Inpatient Antibiotic Consumption

BACKGROUND

Days of Therapy (DOT), the most widely used benchmarking metric for antibiotic consumption, may not fully measure stewardship efforts to promote use of narrow-spectrum agents and may inadvertently discourage the use of combination regimens when single-agent alternatives have greater adverse effects. To overcome DOT's limitations, we developed a novel metric, Days of Antibiotic Spectrum Coverage (DASC), and compared hospitals' performances using this novel metric with DOT.

METHODS

We evaluated 77 antibiotics in 16 categories of antibacterial activity to develop our spectrum scoring system. DASC was then calculated as cumulative daily spectrum scores. To compare hospital benchmarking using DOT and DASC, we conducted a retrospective cohort study of adult patients admitted to acute care units within the Veterans Health Administration system in 2018. Antibiotic administration data were aggregated to calculate each hospital's DOT and DASC per 1,000 days present (DP) for ranking.

RESULTS

The spectrum score for each antibiotic ranged from 2 to 15. There was little correlation between DOT per 1,000 DP and DASC per DOT, indicating that lower antibiotic consumption at a hospital does not necessarily mean more frequent use of narrow-spectrum antibiotics. The differences in each hospital's ranking between DOT and DASC per 1,000 DP ranged from -29.0% to 25.0%, respectively, with 27 (21.8%) hospitals having differences >10%.

CONCLUSIONS

We propose a novel composite metric for antibiotic stewardship, DASC, that combines consumption and spectrum as a potential replacement for DOT. Further studies are needed to evaluate whether benchmarking using the DASC will improve evaluations of stewardship.

Description of antibiotic use variability among US nursing homes using electronic health record data

Background:

Antibiotics are frequently prescribed in nursing homes; national data describing facility-level antibiotic use are lacking. The objective of this analysis was to describe variability in antibiotic use in nursing homes across the United States using electronic health record orders.

Methods:

A retrospective cohort study of antibiotic orders for 309,884 residents in 1,664 US nursing homes in 2016 were included in the analysis. Antibiotic use rates were calculated as antibiotic days of therapy (DOT) per 1,000 resident days and were compared by type of stay (short stay ≤100 days vs long stay >100 days). Prescribing indications and the duration of nursing home-initiated antibiotic orders were described. Facility-level correlations of antibiotic use, adjusting for resident health and facility characteristics, were assessed using multivariate linear regression models.

Results:

In 2016, 54% of residents received at least 1 systemic antibiotic. The overall rate of antibiotic use was 88 DOT per 1,000 resident days. The 3 most common antibiotic classes prescribed were fluoroquinolones (18%), cephalosporins (18%), and urinary anti-infectives (9%). Antibiotics were most frequently prescribed for urinary tract infections, and the median duration of an antibiotic course was 7 days (interquartile range, 5–10). Higher facility antibiotic use rates correlated positively with higher proportions of short-stay residents, for-profit ownership, residents with low cognitive performance, and having at least 1 resident on a ventilator. Available facility-level characteristics only predicted a small proportion of variability observed (Model R2 version 0.24 software).

Conclusions:

Using electronic health record orders, variability was found among US nursing-home antibiotic prescribing practices, highlighting potential opportunities for targeted improvement of prescribing practices.