Background: Patients on immunosuppressive medications may not exhibit the systemic inflammatory response syndrome (SIRS) in the setting of bacterial infection. Our study examines the relationship between serum PCT levels and the odds of manifesting SIRS and BSI in patients on immunosuppressive medications and examines whether this relationship is altered from patients who are not on these medications. The diagnostic performance of Procalcitonin (PCT) detecting BSI in patients on immunosuppressive agents is compared to that in non-immunosuppressed patients. Methods: We tested the association between BSI, serum PCT levels, contemporaneous SIRS scores using logisitic regression in a dataset of 4279 patients. The diagnostic performance of these variables for detecting BSI was assessed. Results: In patients on immunosuppressive medications, multivariate logistic regression models demonstrate that while the serum PCT level is associated with BSI (OR: 2.48, p <.05) — the SIRS score is not. At any given serum PCT level, patients on immunosuppressive agents have lower odds of exhibiting SIRS despite having the same odds of having BSI as non-immunosuppressed patients. PCT (AUC: 0.68) performs better than SIRS (AUC: 0.52) in detecting the presence of BSI in patients on immunosuppressive medications. The diagnostic performance of PCT for detecting BSI in immunosuppressed patients is not significantly different from the non-immunosuppressed cohort. Conclusions: As PCT levels rise, patients on immunosuppressive agents are less likely to mount a SIRS response, despite having a high probability of BSI. PCT might prove helpful in this setting as immunosuppressive agents do not alter the diagnostic performance of serum PCT in detecting BSI.
Bibliographical noteFunding Information:
We would like to thank Zohara Cohen MS, Justin Dale for help with the technical aspects of data management. Research reported in this publication was supported by the National Center for Advancing Translational Sciences of the National Institutes of Health Award Number UL1TR000114 . The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health.
© 2018 Elsevier B.V.
PubMed: MeSH publication types
- Journal Article