Reinfection rate differing between different studies and populations could partly or entirely be the result of differing error rates in lab processes. Just a 0.1% contamination rate between samples in the lab or hospital would more than explain most of the results.