Investigation of prolonged PCR positivity in COVID-19 patients

Authors:
Open Access
Publisher website
Google Scholar
Abstract
This study aimed to investigate prolonged severe acute respiratory syndrome coronavirus-2 polymerase chain reaction (PCR) positivity in hospitalized COVID-19 patients and evaluate the diagnostic performance of rapid antigen tests (RAgT) compared to real-time reverse transcription-PCR (RT-PCR). A prospective single-center study included 82 patients with prolonged PCR positivity (≥28 days). Serial RT-PCR and RAgT were performed at days 7 to 10, 14 to 20, and 28 to 30 post-diagnosis. Clinical data, comorbidities, and laboratory parameters (inflammatory markers, hematological, and biochemical profiles) were analyzed. Statistical analyses included sensitivity, specificity, Cohen Kappa, and receiver operating characteristic curve assessment. RAgT demonstrated 100% sensitivity and 74% specificity relative to RT-PCR, with strong agreement (Cohen κ = 0.803, P* < .001). Inflammatory markers were elevated: C-reactive protein (median 3.55 mg/L), ferritin (510 µg/L), and D-dimer (1400 ng/mL). Comorbidities were present in 62% of patients, primarily hypertension (17%) and diabetes mellitus (15%). Female patients exhibited higher inflammatory markers (C-reactive protein, fibrinogen, and lactate dehydrogenase) and anemia prevalence. Receiver operating characteristic analysis revealed excellent discriminative performance (area under the curve = 0.99). RAgT showed high sensitivity in detecting severe acute respiratory syndrome coronavirus-2 among patients with prolonged PCR positivity, suggesting potential utility in differentiating active infection from residual viral RNA. Persistent PCR positivity may correlate with noninfectious viral shedding, and RAgT could complement RT-PCR in clinical decision-making, particularly in resource-limited settings. Further multicenter studies are needed to validate these findings and assess the impact of viral variants.

This publication has 22 references indexed in Scilit: