Comparison between RT-PCR and TMA for SARS-CoV-2 detection and proposal of a diagnostic algorithm based on cut-off values.

INTRODUCTION

SARS-CoV-2 caused a pandemic with more than 778 million cases and 7 million deaths. The high diagnostic demand forced laboratories to incorporate new techniques, such as transcription-mediated amplification (TMA). This study compares TMA with RT-PCR in routine clinical samples and evaluates the relationship between relative light units (RLU) and RT-PCR results to define possible cut-off ranges.

METHODS

A prospective study was carried out over 1 year in the microbiology laboratory of Arnau de Vilanova University Hospital in Lleida. All samples for SARS-CoV-2 were processed using TMA (Aptima Hologic), and a portion was analyzed in parallel with RT-PCR (LightMix®, GeneXpert®, or Cobas®). Concordance between techniques, sensitivity, and specificity were calculated. RT-PCR was considered the reference technique.

RESULTS

A total of 15,156 samples were analyzed with both techniques. Concordance was substantial (κ = 0.692), with a sensitivity of 97.89% and a specificity of 64.74%. Of the results, 11.28% corresponded to false positives and 8.44% to false negatives for TMA. RLU values between 350 and 559, considered negative by the TMA system, were positive by RT-PCR in 20.56%. In intermediate RLU values (560-899), the probability of having a negative RT-PCR was around 50%, dropping to 3.78% for values between 1000 and 1099.

CONCLUSIONS

These results support the use of TMA as a screening method, provided the positivity threshold is lowered to 350 RLU and positive cases are confirmed by RT-PCR.