Background/Case Studies: During the process of instrument validation, correlations between an established method and the new method are typically performed. Although correlation testing is common, the number of correlations performed has been observed to be variable due to a variety of factors. A manufacturer of automated immunohematological test systems evaluated the average number of correlations performed for ABO/RH (type) and antibody detection (screen) assays at varying sizes/types of facilities to evaluate the effect the size of the facility had on the number of correlations performed.
Study
Design/Methods: A retrospective review was conducted on the number of correlations performed at 190 facilities installing automated instruments between 2018 to 2023. Data was gathered either from the number of correlations specified in the validation plan or from final summaries prepared at the conclusion of the validation. The mean of the number of correlations performed and the annual volume for types and screens were compared and stratified by bed size. Hospitals were categorized by number of beds while donor centers and clinics were grouped separately (other). An ANOVA (Analysis of Variance) was performed for each assay to determine if there was a significant trend between the number of correlations and hospital size.
Results/Findings: The average number of correlations performed for small facilities ( < 250 beds) for medium size facilities (250-499 beds), for large size facilities (≥ 500 beds), and donor centers/clinics (listed as Other) are listed in Table 1. Conclusions: For both type and screen assays, the number of correlations was positively associated with hospital bed size (p= < 0.0008). Further analysis revealed that comparisons between small and medium hospitals displayed no such difference. Statistical significance was only observed when large facilities were compared to small or medium ones. This suggested a trend that large facilities may behave fundamentally differently than smaller hospitals in the number of correlations and volume of testing they can perform.
Importance of research: Blood establishments often wish to follow consensus in quality processes. The research here gives concrete data showing what other blood establishments are doing in terms of quantity of correlations for type and screen assays. As more facilities move to automation to combat a shrinking labor pool and more demand to do more with less, leaders are often conflicted over how much correlation to perform.
