Performance of procalcitonin point-of-care tests in predicting bacterial causes of acute febrile illness in Northwest Ethiopia

Objective
Antimicrobial resistance in sub-Saharan Africa is fueled by empiric antibiotic use due to the absence of diagnostic testing for bacterial infections. We aimed to assess the performance of qualitative, semiquantitative, and quantitative procalcitonin (PCT) rapid point-of-care (POCT) tests in identifying bacterial causes of fever among acute febrile patients.

Methods
A total of 181 frozen serum samples from adults (aged over 15 years) were assessed via qualitative, semiquantitative, and quantitative procalcitonin tests. The agreement between the test methods was evaluated via kappa statistics. A composite reference test was created by combining the results from all the PCT methods, due to the absence of a reliable gold standard to verify bacterial infections and the unavailability of standard PCT references. Each test’s diagnostic accuracy was compared with that of the reference test. Bacterial infections were determined on the basis of blood culture and qPCR for Borrelia and Rickettsia species.

Results
All the PCT tests had substantial to near-perfect agreement with each other (kappa = 0.69–0.81) and compared with the composite reference standard (kappa = 0.68–0.97). All three PCT test methods demonstrated excellent diagnostic accuracy, with an AUC ranging from 0.91 to 0.99; (95% CI 0.86–0.98 p < 0.001), in predicting bacterial infections, although the sensitivity and agreement decreased significantly at higher PCT concentrations, measured by semiquantitative and quantitative tests.

Conclusions
Point-of-care PCT tests may be helpful tools for distinguishing bacterial from nonbacterial causes of acute febrile illness in low-resource settings, though further studies with larger sample sizes are warranted to validate these findings and optimize test interpretation across varying PCT levels.