A Large-Scale Retrospective Study on Thrombocytopenia Associated with Beta-Lactam Antibiotics

Abstract

While the ability for beta-lactams (BL) to induce thrombocytopenia (TCP) is well understood, their association is not well quantified in the general population. Despite this, when platelets drop in the clinical setting, BL are frequently substituted for alternative antibiotics, leading to suboptimal outcomes. Here, we present a large-scale, retrospective study on the association of TCP and BL when compared to alternative non beta-lactam (nBL) therapy. All adult inpatients who received at least one antibiotic between 2008 and 2021 were included. Incidence of TCP in the 30 days following antibiotic administration was compared across patients receiving exclusively BLs vs nBLs as well as with each antibiotic subclass permutation following propensity score matching. There is a mild, though statistically significant increase in TCP risk for BL when compared to alternative nBL therapy. Risks and benefits should be considered prior to switching off BL therapy if clinically indicated.

Figure 1

Flow diagram of inclusion and exclusion criteria.

Figure 2

Antimicrobial use and associated risk of thrombocytopenia. TCP = thrombocytopenia. Red line indicates an odds ratio of 1.00. * The number of patients receiving anti-tubercular, monobactams, and oxazolidinones exclusively during a hospital admission was too low for there to be a discernable difference in chemotherapy administration by outcome and thus removed as a controlled feature.