BACKGROUND: The objectives of this study are to estimate the prevalence of iron deficiency (ID) among French whole-blood (WB) donors to identify factors associated with ID and to generate decision trees. STUDY DESIGN AND METHODS: A prospective National multicentre study was performed on WB donors from March 11, to April 5th, 2019. Samples were selected randomly to perform serum ferritin. ID was defined as ferritin value under 26 ng/ml. All results were stratified by sex. Factors associated with ID were analysed using multivariate logistic regression model. CART algorithm was used for decision trees. RESULTS: Eleven thousand two hundred fifty eight WB donors were included. ID was more frequent in women (39·5%) than in men (18·0%). Among 7200 repeated donors, women below 50 yo had a higher risk (OR = 2·37; [1·97-2·85] IC95) than those above 50 yo. Factors associated with ID were: haemoglobin level under the threshold at donation n-1 except for women and n-2 donation; a low mean corpuscular haemoglobin at n-1 and n-2 donations; a shorter interval since n-1 donation and between n-1 and n-2 donations except for women; and women who had given three or four times in the last year. CART algorithm defined high risk of ID subgroups within three populations of donors, new female donors, repeated male donors and repeated female donors. In these identified subgroups, prevalence of ID was up to 72·1%. CONCLUSIONS: Our study showed the high prevalence of ID among French WB donors, identified well-known and new factors associated with ID and defined algorithms predicting ID in three populations.
BACKGROUND: The objectives of this study are to estimate the prevalence of iron deficiency (ID) among French whole-blood (WB) donors to identify factors associated with ID and to generate decision trees. STUDY DESIGN AND METHODS: A prospective National multicentre study was performed on WB donors from March 11, to April 5th, 2019. Samples were selected randomly to perform serum ferritin. ID was defined as ferritin value under 26 ng/ml. All results were stratified by sex. Factors associated with ID were analysed using multivariate logistic regression model. CART algorithm was used for decision trees. RESULTS: Eleven thousand two hundred fifty eight WB donors were included. ID was more frequent in women (39·5%) than in men (18·0%). Among 7200 repeated donors, women below 50 yo had a higher risk (OR = 2·37; [1·97-2·85] IC95) than those above 50 yo. Factors associated with ID were: haemoglobin level under the threshold at donation n-1 except for women and n-2 donation; a low mean corpuscular haemoglobin at n-1 and n-2 donations; a shorter interval since n-1 donation and between n-1 and n-2 donations except for women; and women who had given three or four times in the last year. CART algorithm defined high risk of ID subgroups within three populations of donors, new female donors, repeated male donors and repeated female donors. In these identified subgroups, prevalence of ID was up to 72·1%. CONCLUSIONS: Our study showed the high prevalence of ID among French WB donors, identified well-known and new factors associated with ID and defined algorithms predicting ID in three populations.
Authors: Sara Moazzen; Maike G Sweegers; Mart Janssen; Boris M Hogema; Trynke Hoekstra; Katja Van den Hurk Journal: J Clin Med Date: 2022-06-21 Impact factor: 4.964