Description:
We agree with John T Connelly and colleagues1 that assessment of analyte stability - and matrix effects must be part of every assay validation. It is, however, essential - that validation is done with the assay under consideration, as different lipoarabinomannan - (LAM) assays use different reagents that bind to different LAM epitopes with different - stabilities or availabilities. Therefore, results from the research assay described - by Connelly and colleagues,1 showing rapid signal loss, are not generalisable to other - LAM tests.