Differences in Use of the Antibiotic Timeout Between Hospitalists and Housestaff at an Academic Medical Center

Background:

Curbing antibiotic overutilization has become a priority in hospitals attempting to address the growing problem of antimicrobial resistance.  One strategy being endorsed by Antibiotic Stewardship Programs is the adoption of an “antibiotic timeout,” during which the pharmacist reviews the appropriateness of the regimen every 72 hours.  We elected to challenge hospitalists to perform a daily “timeout” to assess antibiotic use. We compared the prevalence of timeout on the teaching and nonteaching services at our institution to better understand the uptake of this practice at a large academic medical center.

Methods:

We reviewed progress notes from a randomly selected group of patients hospitalized on the Internal Medicine service between 1/1/2014 – 1/31/2015 who had received antimicrobials for at least 48 hours.  Patients receiving prophylactic antibiotics, and those with formal Infectious Diseases consults were excluded.   

Documentation of an appropriate antibiotic timeout required the following elements:  antibiotic name, indication for therapy, current day of therapy, and expected therapy end date.  For example, “MRSA sepsis, Vancomycin day #6 of 42” would be considered appropriate, but “MRSA Sepsis, Vancomycin day #6” would not.  Any documentation to suggest that the end date was not yet determined was considered acceptable. 

We recorded the presence or absence of an appropriate timeout in all progress notes written by the Medicine service for a given patient.  In addition, we recorded whether the patient was on a teaching service, and whether the note author was a resident or attending.  Of note, rotating hospitalists from the same group staffed patients on both teaching and nonteaching services.  As a result, notes were divided into 3 categories:  resident notes, attending notes on a teaching service, and attending notes on a non-teaching service.  Between-group differences were analyzed using 2×2 chi-square analyses.

Results:

We reviewed 137 patient charts, and a total of 535 individual progress notes.  Of the 535 notes, 67% were from nonteaching services, and overall 70% of the notes were written by attending physicians.  An appropriate timeout was documented in 25% of all notes. 

Timeouts were present more frequently on teaching services.  34% of notes from patients on teaching services contained a documented timeout, as compared to 17% on nonteaching services (p < 0.05).  Notes authored by residents contained timeout documentation 36% of the time, as compared to 19% in notes authored by attendings (p<0.05).

Conclusions:

Antibiotic timeout documentation was infrequent on the Medicine service.  However, documentation was significantly higher on teaching services, despite the fact that the supervising attending physicians were the same.  Understanding the factors that motivate housestaff to address antibiotic stewardship more consistently would be informative in motivating faculty to improve.