On our Year in Review webinar, presenters Stephanie Hughes, PharmD, BCIDP, Clinical Applications Specialist, and Jon Faldasz, PharmD, BCPS, our Senior Director of Product & Customer Experience, shared recent product enhancements for InsightRX Nova and InsightRX Apollo.
Accounting for the Loading Dose Effect: Moving the Needle on Early Target Attainment
With precision dosing, we are always striving to perfect our predictions of the patient’s pre- steady-state and steady-state exposures, as seeing the predictions in both phases provides the most complete picture. After discovering that the loading dose still impacts concentration levels later on in the course of therapy, during steady-state predictions, we are now incorporating loading doses into steady-state predictions within the InsightRX Nova dashboard.
Many clinicians think about the impact of loading doses in the beginning phases of therapy, before the creation of steady-state predictions. Without a loading dose, concentration levels take a while to build; a patient might require eight doses to reach the target steady-state concentration for the area under the curve (AUC). When a loading dose is added, however, the equation changes.
Our pharmacists noticed that even after eight maintenance doses, the leftover effect of the loading dose was still impacting the AUC for many patients, such as those with reduced renal function. The new enhancement incorporates the loading dose into the reference table predictions, and recalculates predictions to a more accurate representation of pre-steady-state and steady-state exposures that reflect the seamless incorporation of the patient’s loading dose into AUC-24ss.
While this is a simple innovation, it’s important to eradicate as many assumptions as possible to achieve the most accurate prediction. Just as we assess all doses at the exact same time as they’re given—and do not simply assume a perfect Q8 interval—we no longer have to assume that a loading dose is not likely to affect the patient after several doses of a regimen are given.
In fact, the latest vancomycin therapeutic drug monitoring (TDM) guidelines discuss better incorporation of the loading dose predictions as one of the benefits of Bayesian dosing software programs. By front-loading doses before transitioning to a maintenance dosing regimen, clinicians can rapidly achieve target concentrations within the first 24 to 48 hours, which is especially important for critically ill patients with serious infections.
With better loading dose strategies and interventions, clinicians may be able to move the needle for early and overall target attainment. We find that leveraging InsightRX Nova to conduct a vancomycin loading dose intervention—either for all patients or only for those with more serious infections—enables a significant impact on both the pre-steady-state and post-steady-state periods.
To Cap or Not to Cap Creatinine Clearance: A Model Analysis Provides Answers
Our second enhancement stems from our pharmacist users’ uncertainty about whether to use creatinine clearance (CrCl) caps in various models. A common practice for adults receiving vancomycin is to cap CrCl above 150 ml/min. InsightRX Nova also lets users configure CrCl caps for other models, if they believe it will be a good practice for their patients. Our pharmacist clients often ask if using CrCl caps will lead to better predictions or outcomes.
To answer those questions, our Data Science team performed an analysis of 4,759 patients across 22 hospital systems who were clinically dosed with vancomycin using two models: the Thomson model (JAC 2009) and the Carreno model (AAC 2017). They found that the Thomson model performs better with a CrCl cap at 150 ml/min rather than no cap. On the other hand, the Carreno model, which is used on Obesity Class 3 patients with a BMI of 40 or above, performs better uncapped.
The full report of the analysis is available in the InsightRX Nova 1.37 release notes. The next version of the platform will include a modified Thomson model with an implied cap of 150 ml/min as an option for user configuration.