Julie Ann Justo, PharmD, MS, BCPS-AQ ID, provides an example of how institution-specific information can be used to generate a customized risk score.
Julie Ann Justo, PharmD, MS, BCPS-AQ ID, clinical assistant professor, University of South Carolina College of Pharmacy, provides an example of how institution-specific information can be used to generate a customized risk score.
Interview transcript (modified for readability):
“The first step in implementing a risk score is figuring out the high-impact initiatives at your hospital. Infectious disease is always very local; in our particular institution we [focus] on gram-negative bacteremia because, for example, if you look at our hospital antibiogram and our Pseudomonas susceptibility rates, cefepime, piperacillin tazobactam, and meropenem—our workhorse carbapenem—all show about 85% susceptibility to Pseudomonas.
We could talk about the reasons why [this is the case] but, we really wanted to figure out in whom certain anti-pseudomonal beta-lactams would work best since none of those [agents] are at 90% or above, [and] ideally for critically-ill patients, we would like to see an agent that was at least 95%.
That was an example of a very practical question [we asked] when we were creating what amounted to sepsis guidelines for gram-negative broad-spectrum selection. We had some national guidelines that would help [create] an antibiogram, [but] nothing there seemed to quite fit the bill, [and so] we were left with customizing [a guideline to fit our institution] and the prediction score was one way [to] do that.”