Quality control of multisource aeroallergen data

Pollen and spore prevalence data from 14 air sampling stations were retrospectively analyzed for accuracy of taxa identification and counting. Pollen identification was excellent and levels calculated by our laboratory closely matched those reported by most stations. Neither identifications nor counts were routinely reliable for fungus spores. While Alternaria was usually counted correctly, Cladosporium was frequently undercounted, and ascospores and basidiospores, which often constituted more than half the total spore load, were either not counted or severely undercounted. Errors resulted from undercounting, poor recognition, and reporting (including naming) errors. Multisite aerobiological studies that include documentation of fungus spore levels must include pretraining for each technician, and carefully documented standardization of counting and reporting procedures.