Rapid Diagnostics for Gram-Negative Bloodstream Infections
Kimberly Claeys, PharmD, discusses the need for more data on rapid diagnostics for gram-negative bloodstream infections.
Segment Description: Kimberly Claeys, PharmD, an assistant professor of pharmacy practice at Maryland School of Pharmacy, discusses the need for more data on rapid diagnostics for gram-negative bloodstream infections.
Claeys: Rapid diagnostics are critical to improving outcomes in bloodstream infections in a variety of ways. We already know from previous literature that having stewardship involved in rapid diagnostics improves patient outcomes. It can decrease length-of-stay, improve mortality. The thing is most of that data is driven by gram-positive bloodstream infections. We really don't know our role for gram-negative bloodstream infections as much.
The thing that I thought was surprising or unexpected that we observed was that we actually still saw a change in antibiotic time even without active antimicrobial stewardship intervention. Based on previous studies we assumed that without an active antimicrobial stewardship intervention we weren't going to see that big of an impact. But seeing a decrease of about 20 hours to optimal therapy even without stewardship involvement was surprising. Granted it was primarily driven by our ID consult teams and it was all escalation, but I'm still surprised at how big the difference was.
With our current study we found that we were doing really well with optimal escalation of therapy without active antimicrobial stewardship involvement. I think stewardship is going to be key to help with de-escalation. Again, gram-negatives are tricky it's not like gram-positives where the panel tends to be more complete for the rapid diagnostic. The rapid diagnostic in these cases doesn't always provide complete data so making decisions off of it can be more challenging. But having our stewardship team go through and validate how the rapid diagnostic can be used based on your own institutional susceptibility profile is really key to helping actually use this for de-escalation. Realistically, de-escalation is difficult in the setting of incomplete data that these rapid diagnostics provide, but it’s not impossible.