JMIR Publications

ABSTRACT

Background:

The employment of residual blood donor samples has been instrumental in conducting SARS-CoV-2 seroepidemiological studies across countries. However, the use of blood donor samples can introduce bias, known as the 'healthy donor effect', because blood donors are generally healthier than the general population. This study aims to refine the interpretation of blood donor SARS-CoV-2 seroepidemiological studies.

Objective:

This study evaluates potential biases in SARS-CoV-2 seroprevalence studies using blood donor samples. We examine the association between blood donation experience and COVID-19 diagnoses to assess the healthy donor effect — the tendency for blood donors to be healthier than the general population. Understanding this bias is vital for accurately interpreting seroepidemiological data.

Methods:

A web-based survey involving approximately 10,000 Japanese residents, aged 16-69, was conducted from 14-28 December 2023 to clarify the association between blood donation experience and COVID-19 diagnosis. A logistic regression was used adjusting for demographics, social economic status, COVID-19 vaccination status, and comorbidities.

Results:

Among 10,781 participants, 3,583 (33.2%) reported a history of COVID-19 diagnosis, while 5,015 (46.5%) indicated they had donated blood at least once in their lifetime. Blood donors had 1.62 (95% Confidence Interval: 1.48–1.78) times higher odds of SARS-CoV-2 infection than non-donors.

Conclusions:

Higher COVID-19 diagnosis rates among blood donors may result from increased community involvement and more frequent testing, indicative of an 'inverse' healthy donor effect. This pattern implies that, in terms of SARS-CoV-2 infection, blood donors might not represent the healthiest population. Consequently, seroprevalence studies using blood donor samples could overestimate the seroprevalence of SARS-CoV-2 within the general Japanese population.