Abstract
Importance Blood donors are at increased risk for iron deficiency. Current hemoglobin (Hb) monitoring policies are inadequate and iron management strategies are warranted.
Objective To determine the effects of ferritin-guided donation intervals for whole blood donors on Hb and ferritin levels, Hb deferral, iron deficiency (ferritin < 15 ng/mL) prevalence, donor return, iron deficiency-related symptoms, and the blood supply.
Design, setting and participants Between November 2017 and November 2019 a ferritin-guided donation interval policy was implemented and evaluated through a stepped-wedge cluster-randomized controlled trial in all blood collection centers in the Netherlands. During the study period, 412,888 donors visited the blood bank.
Interventions Ferritin was measured in new donors and at every 5th whole blood donation. Subsequent donation intervals were extended to six months if ferritin is 15 ≤ 30 ng/mL and to twelve months if ferritin is <15 ng/mL.
Main outcomes and Measures Primary outcomes are ferritin and Hb levels, iron deficiency, Hb deferral, and donor return. Secondary outcomes are self-reported iron deficiency-related symptoms. Yearly donor and donation statistics were used to assess the impact on the blood supply.
Results We measured 36,099 donors, median age 43 years and 52% female, making 37,621 donations during the study period. Ferritin-guided donation intervals were associated with increased log-transformed ferritin levels at all time points in the trial, up to 0.24 log10 ng/mL as compared to baseline (95% CI 0.22 – 0.27, p <0.001). Hb increased as well, up to 0.37 g/dL (95% CI 0.31 – 0.43, p <0.001). Decreased odds of 85% (95% CI OR 0.10 – 0.21, p <0.001) for iron deficiency and 77% (95% CI OR 0.11 – 0.41, p <0.001) for Hb deferral were reported compared to baseline. Odds of donor return decreased over the course of the trial, with as much as 40% (95% CI OR 0. 51 – 0.67, p <0.001). We found no evidence for improved self-reported iron deficiency-related symptoms after implementation of the new policy. More invitations had to be sent to donors to obtain one donation and the average donation frequency per year decreased.
Conclusion and Relevance Ferritin-guided donation intervals were significantly associated with increasing overall Hb and ferritin levels, thereby decreasing the prevalence of iron deficiency and Hb deferrals in whole blood donors. Ferritin-guided donation intervals seem beneficial for Hb levels and iron maintenance, but additional efforts are required to retain donors and to remedy self-reported iron deficiency-related symptoms.
Trial Registration This trial has been registered in the Dutch trial registry (NTR6738) on September 29th, 2017 (https://trialsearch.who.int/Trial2.aspx?TrialID=NTR6738).
Question Are ferritin-guided donation intervals effective in diminishing iron deficiency and low hemoglobin in whole blood donors?
Findings In this stepped-wedge cluster randomized trial with 412,888 donors and 1,643,700 donations, ferritin-guided donation intervals were associated with a significant improvement in hemoglobin and ferritin levels and decreased donor return rates. Data did not suggest improved self-reported health outcomes.
Meaning Ferritin-guided donation intervals, in addition to hemoglobin monitoring, as part of the donor iron management strategy of blood establishments effectively reduce the prevalence of iron deficiency and anemia in whole blood donors, but additional efforts to retain donors are required.
Introduction
Annually, more than 100 million whole blood donations are made worldwide to provide life-saving treatment for patients requiring transfusion.1 Whole blood donors that donate repeatedly are at risk of developing iron deficiency and subsequent anemia.2 Male donors lose approximately 8% of their total iron stores with each whole blood donation, whereas menstruating women lose on average 81% of their total iron stores.3–5 Iron stores can be monitored through measuring ferritin. Iron deficiency, ferritin levels <15 ng/ml, presents in approximately 15.0% of female and 9.4% of male Dutch repeat donors.6
There is no consensus on iron management strategies for donor health, but key approaches include extended donation intervals and iron supplementation.7–9 Concerns about negative perceptions, side effects, and low compliance may favor extended donation intervals. 9,10 In previous studies, extended donation intervals were associated with lower risk of Hb deferrals.11,12 The INTERVAL trial showed that longer donation intervals led to significantly higher Hb and ferritin levels.12 Most blood establishments monitor Hb to protect donor health, prevent anemia and ensure sufficient Hb content in the product, deferring blood donors with Hb levels below 13.5 g/dL (males) or 12.5 g/dL (females).7–9,13–15 However, iron deficiency can present without anemia, especially in repeat donors. As Hb levels do not reflect iron stores, the current standard is insufficient to maintain proper iron reserves. The World Health Organization (WHO) advises additional monitoring of ferritin to detect iron deficiency and monitor donor health.16
In 2017, Sanquin Blood Bank in The Netherlands implemented a nationwide policy for iron monitoring of whole blood donors. In addition to the Hb measurement at the pre-donation screening, ferritin is measured at the first visit of new donors and after every fifth whole blood donation of repeat donors. If ferritin is between 15 and 30 or below 15 ng/mL, the subsequent donation interval is extended to six or twelve months, respectively. A stepped-wedge approach was used to evaluate the effects of this ferritin-guided donation interval policy for whole blood donors on (1) Hb and ferritin levels, (2) iron deficiency, (3) Hb deferral, (4) donor return rates, (5) self-reported iron deficiency-related symptoms, and (6) the blood supply.
Methods
Design
Ferritin measurement IN Donors—Effectiveness of iron Monitoring (FIND’EM) is a stepped-wedge cluster randomized trial, described in detail by Sweegers et al.17 We conducted cross-sectional ferritin measurements before, during, and after implementing a new policy for all donors visiting a blood collection center during pre-selected weeks. We measured ferritin in stored plasma samples from donors who consented to use their leftover donations for medical research. Additionally, optional questionnaires were offered to donors. The trial was powered on 4ng/mL change in the ferritin level outcome over six months, assuming a SD of 54 ng/mL, ICC of 0.10 and 150 visits per donation center.17 The policy and implementation strategy was approved by Sanquin’s Medical and Ethical Advisory Board. The trial was approved by Sanquin’s Board of Directors.
Randomization
The 138 fixed and mobile blood collection centers in the Netherlands are divided over 29 staffing clusters that we randomly assigned to four randomization groups. The stepped-wedge design involved random and sequential rollout of ferritin-guided donation intervals in all 29 clusters by randomization group from November 2017 to November 2019. The allocation was not concealed and the planned timing of the rollout was revealed before the study. Donors were notified of the new policy through letters upon implementation at their regular donation center.
Procedures
As part of routine procedures, donors undergo a health screening to assess their eligibility to donate by measuring Hb levels and blood pressure and assessing recent travel, sexual and medical history prior to every donation. In the Netherlands, donors also make a pre-donation screening visit prior to their first donation. In accordance with European legislation, Hb cut-offs for donor eligibility are >13.5 g/dL for males and >12.5 g/dL for females.18 Donors with low Hb (capillary measurement prior to donation with HemoCue 201, Angelholm, Sweden) are deferred for 3 months. If eligible, males can donate every 56 days up to five times a year, females every 122 days up to three times a year, independent of menopausal status.
The intervention entails extended donation intervals based on ferritin measurements. The cut-offs for extended intervals were based on definitions for iron deficiency (<15 ng/mL, 12 months deferral) and reduced iron levels (≤30 ng/mL, 6 months deferral) by Alvarez-Ossorio et al.9 The six month interval is based on the results from Schotten et al., suggesting intervals of at least 180 days are required to fully replete iron stores after donation.8 A twelve month interval is indicated for iron deficient donors to replete to higher levels than pre-donation. In whole blood donors, ferritin is measured at the pre-donation screening visit before the first donation and of each fifth whole blood donation. If ferritin levels are >30 ng/mL, donors are eligible to donate at regular donation intervals. If ferritin levels are ≤30 ng/mL but ≥15 ng/mL, repeat donors are deferred for six months; new donors are allowed to donate, but ferritin is measured again at this first donation, after which they follow the repeat donor policy. If ferritin levels are <15 ng/mL, new and repeat donors are deferred from donating for twelve months. Ferritin levels are measured again in the donation after this extended interval. Donor physicians may additionally advise donors to visit their general practitioner and/or donate at a lower frequency.17
Outcome assessment
Before the trial, we established four measurement points: 1) one week before implementation (baseline), 2) midway through implementation, 3) one week after full implementation, and 4) one year after full implementation of ferritin-guided donation intervals. Primary outcomes included Hb and ferritin levels, Hb deferral, iron deficiency, and donor return rates, while secondary outcomes encompassed iron deficiency-related symptoms and the impact on the blood supply. Pseudonymized data on age, sex, Hb levels, and deferrals were extracted from the blood bank database (eProgesa software application; MAK-SYSTEM, Paris, France) for donors who visited a blood collection center during a measurement week and consented to their information and leftover materials being used for research. Donor return was defined as attempting a donation within six months after becoming eligible to donate again, accounting for minimum donation intervals and possible deferrals. We measured ferritin levels on the Architect i2000sr (Abbott Diagnostics) in plasma samples collected at donation in K3EDTA tubes (VACUETTE®, Greiner Bio-One, Kremsmünster, Austria) and stored at −30⁰C for three to 24 months.19
New and repeat donors in fixed blood collection centers during measurement week 2 and 3 that provided informed consent were asked to complete questionnaires. These questionnaires assess physical and mental wellbeing (SF-36), fatigue (Checklist Individual Strength), donation-related symptoms, restless legs syndrome (RLS) and pica, cognitive functioning (Cognitive Failure Questionnaire), and warm glow.12,17,20–23 Questionnaires were completed using Qualtrics XM® software (Provo, UT, USA) during or shortly after the donation.17
We aimed to assess the impact of the new policy on the blood supply by comparing operational data from the blood bank during and one year before and after the study period. We gathered data on whole blood donations, donors and donations invitations from the blood bank database (eProgesa software application; MAK-SYSTEM, Paris, France). Available donors were defined as not being deferred for medical (iron-related or other), travel or lifestyle factors; new donors as showing up for the pre-donation screening; and lapsed donors when they deregistered or did not respond to 5 subsequent invitations. Demands for each year were based on hospital requests and converted to blood units.
Data analyses
Descriptive statistics are provided for the donors measured during measurement week 1. Using donors measured during all four weeks, we conducted complete case analysis and used ordinary least squares regression in case of continuous outcomes and logistic regression for binary outcomes. After diagnostic tests of linearity and errors assumptions, a log base 10 transformation was applied to improve linearity of ferritin. We conducted robust regression to analyze cognitive functioning, physical health, mental health and fatigue, as the normality assumption for logistic regression was violated. Measurements after implementation were grouped in periods of 6 months and compared to the control condition. The control condition was defined as no implementation of the new policy or implementation less than 6 months ago. Analyses were adjusted for age, weight and height, and stratified by sex and menopausal status. Donor return analyses were adjusted for return being during a COVID-19 lockdown. Null hypotheses are that there are no associations between implementation of the new policy and the outcomes. A two-sided p-value of <0.05 is considered statistically significant and a Bonferroni correction is applied. Statistical analyses were conducted using the R language and environment for statistical computing (4.1.1) and the lme4 (1.1.33), car (3.1.2), lmtest (0.9.40), sandwich (3.0.2), olsrr (0.5.3), and robustbase (0.95.1) packages.24–30 HTML files of the scripts and results produced in Quarto are available on GitHub through https://github.com/Sanquin/FINDEM.31
Deviations from protocol
As the number of donors with low ferritin was higher than anticipated after implementation in the first group, further implementation was executed in smaller steps to mitigate the impact on the blood supply.17 Additional ferritin measurements were done on samples from all donations one year after the third measurement round (between November 23rd and 27th 2020). This fourth time point enables evaluation of long-term results. We furthermore deviated from the protocol by using ordinary least squares regression instead of linear mixed models. In the linear mixed models for analysis of the primary outcomes with random effects for the donation center clusters, the ICC ranged from 0.00 to 0.05; R2 did not improve with random effects; and residual variance exceeded the variance between clusters. This indicates that the random effects of the donation center clusters were of minimal importance and justifies the use of linear regression models. We conducted sensitivity analyses for donor return, stratifying by ferritin level, reason for deferral, and donation center type. Lastly, we added a sixth element (impact on the blood supply) to the research question and accordingly adjusted the methods.
Results
A total of 412,888 whole blood donors made 1,634,700 donations during the study period. We performed measurements in samples of 37,621 donations by 36,099 donors (Figure 1). Table 1 shows the characteristics of donors by sex and randomization group in measurement week 1, in Table S1 they are compared to measurements from week 4. Overall, across the randomization groups, donor characteristics were well balanced and comparable with the general donor population (Table 1).32 Characteristics of the 7,573 participants in the questionnaire in measurement week 2 are provided in Table S2.
Figure 2 shows that the prevalence of iron deficiency and Hb deferral decreased after implementation of the policy. Ferritin-guided donation intervals are associated with significantly higher ferritin and Hb levels and less iron deficiency and Hb deferrals (Figure 3a). Compared to pre-implementation, the intervention was associated with up to 0.24 ng/mL higher log10 ferritin and 0.37 g/dL higher Hb at 36 months after implementation in males (Figure 3a). In premenopausal and postmenopausal females, we also observe positive, but smaller, associations with an increasing trend over time. Moreover, the odds ratio of 0.19 indicates that males are 81% less likely to be iron deficient at 36 months after implementation than before the intervention (Figure 3a). In postmenopausal females, odds ratios also decrease over the course of the trial, with iron deficiency being 87% less likely at 36 months after implementation compared to pre-implementation (Figure 3a). Hb deferral in males was up to 77% less likely as time since implementation increased, compared to pre-implementation. Similarly, in premenopausal females, Hb deferral is up to 44% less likely at 30-35 months after implementation than at pre-implementation (Figure 3a). In postmenopausal females, Hb deferral is up to 51% less likely at 36 months after implementation than at pre-implementation.
After adjustment for return being during a COVID-19 lockdown, odds for donor return within six months are lower at each time point after implementation compared to pre-implementation (Figure 3a). Generally, the negative association with donor return was stronger in males. At 36 months after implementation, males were 42% less likely to return for donation (Figure 3a). For premenopausal and postmenopausal females, we do not observe a consistent decrease in odds over time (Figure 3a). Tables S4-6 show that the differences in odds between groups of ferritin levels, reasons for deferral and donation center types are mostly not statistically significant.
There were no consistent significant patterns for pica, fatigue, cognitive functioning, the SF-36, and warm glow after implementation of the new policy (Figure 3b). However, males and premenopausal and postmenopausal females were 49%, 127%, and 127% more likely to report RLS at 24 months after implementation than before implementation.
Table 2 shows that during the implementation of the policy, the donation target and number of donations decreased, but the number of invitations needed to obtain one donation increased. The average donation frequency of donors decreased over the years and subsequently the number of donors that donated increased. The number of available donors increased and the proportion that donated fluctuated between 80 and 90% over the years. The number of simultaneously unavailable donors stabilized at around 40,000 donors per year. While the percentage of low ferritin deferrals increased as a result of the implemented policy, the percentage of low-Hb deferrals decreased. The number of new donors at the blood bank was higher in 2019 and 2020 than in other years, while the number of lapsed donors in 2019 was exceptionally low.
Discussion
This stepped-wedge cluster-randomized trial in whole blood donors shows that ferritin and Hb levels are significantly higher and iron deficiency and Hb deferral significantly lower after the implementation of ferritin-guided donation intervals compared to the prior standard of Hb monitoring only. Conversely, we found that the odds of donor return within six months decreased substantially with this policy. Most iron deficiency-related symptoms remained unaffected, except for an increase in reports of RLS. The continuity of the blood supply was maintained, but additional efforts were required to account for the decreased donor availability.
Ferritin-guided donation intervals were associated with significantly less iron deficiency and higher ferritin and Hb levels compared to pre-implementation, consistent with findings from donor population simulations assessing ferritin testing’s impact.33 The weaker effects observed in premenopausal females could be due to their already less frequent donations before the policy, driven by longer donation intervals and more frequent low-Hb deferrals, providing more recovery time.34,35 Additionally, because of more stable iron stores due to the healthy donor effect, combined with lower donation frequency and ferritin levels in this group, may result in less pronounced improvements compared to other groups.36,37
Our finding that the new policy is associated with significantly lower odds of donor return supports previous findings on decreased return rates after deferral, but may additionally be explained by the negative effects of the COVID-19 pandemic on blood donation intentions, as donors measured after full implementation were slated to return during COVID-19 lockdowns.35,38–40 The weaker negative association with donor return in premenopausal compared to other groups may again be explained by their exposure to more frequent deferral through the old Hb monitoring policy.
The increased reporting of RLS seems surprising, but may simply be a result of increased awareness through information provided to donors upon implementation in their donation center. Further investigation shows no statistically significant correlation between RLS and ferritin levels. The lack of improvement we found in self-reported iron deficiency-related symptoms may partially be due to the small number of questionnaire participants before implementation. However, our findings align with previous randomized placebo-controlled trials investigating iron repletion and self-reported donor health.41–43 Moreover, the benevolence and warm glow that drive donors may overshadow negative health consequences.23 More functional and objective measures of iron deficiency-related symptoms like fatigue are thus needed. Positive effects of the steady reduction in iron deficiency and Hb deferral on self-reported iron deficiency-related symptoms may be delayed and require longer follow-up in future studies.
Given the results of this trial regarding donor health and the continued fulfillment of the donation target during and after the implementation of ferritin-guided donation intervals, Sanquin decided to keep the policy in place. Our findings do, however, raise some important policy and practical considerations. First, it provides policymakers with a monitoring option that requires no effort from blood donors, may be more effective in protecting iron status than standard Hb monitoring practices, and does not impose a risk of side effects compared to iron supplementation.10 However, the decreasing donor availability due to decreasing odds of donor return and increased numbers of deferred donors stress the need for intensified donor recruitment and retention efforts.34 This is reflected in the blood bank practice, as more invitations are needed to obtain one donation. Moreover, as donation frequencies decreased after implementation of the policy, more donors were needed to reach donation demands. To increase donor availability, iron supplementation can be considered as an alternative or additional strategy.33,44
To our knowledge, this is the first randomized study that goes beyond the limitations of Hb monitoring by evaluating the effectiveness of ferritin-guided donation intervals in improving Hb and ferritin levels, donor return rates and decreasing iron deficiency, Hb deferrals, iron deficiency-related symptoms. One of the major strengths is the stepped-wedge cluster randomized approach as it allows simultaneous implementation and evaluation of a new policy. However, there are also limitations to this study. The nature of the intervention poses a risk of overestimating the positive effect of ferritin-guided donation intervals. Donors whose donation intervals are extended, may not return frequently, potentially resulting in selection bias. The risk of overestimation is to some extent minimized as the duration of the trial exceeds the length of extended donation intervals. There may even be underestimation of the effect on iron deficiency, as cut-offs for iron deficiency used in this study are lower than suggested in literature.45,46 There may also be undetected delayed effects in this trial, because ferritin levels are measured at every fifth donation, and thus individual donors may not be confronted with the policy immediately upon implementation in their donation center.
The FIND’EM study provides substantial evidence that ferritin-guided donation intervals increase Hb and ferritin levels and thereby effectively prevent iron deficiency and Hb deferral, while efforts are needed to mitigate effects on the blood supply. Future studies may investigate long-term effects on donor health outcomes. Our results may help blood services worldwide to improve their donor iron management policies in order to protect donors from developing iron deficiency.
Declarations
Ethics approval and consent to participate
Sanquin’s Ethical Advisory Board has reviewed and approved the FIND’EM protocol. Only data from donors who gave consent for their information and left-over material to be used for research purposes at the health check were included in the study. Donors who completed questionnaires provided informed consent for their questionnaire data to be matched with data from the donor database.
Competing interests
The authors declare that they have no competing interests.
Funding
The current project is supported by the ‘Product and Process Development Cellular Products Grant’ (PPOC18-15) granted to K. van den Hurk by the Research Programming Committee, Sanquin, Amsterdam, The Netherlands.
Authors’ contributions
KvdH designed the study. KvdH, MS, FP and FQ set up and coordinated data collection. AM, and SR wrote the analysis plan and drafted the manuscript. AM, SR and FP conducted the analyses with support from EH. The manuscript was critically reviewed and approved by all co-authors.
Data Availability
All data produced in the present study are available upon reasonable request to the authors.
Acknowledgements
We would like to acknowledge all blood bank staff for their commitment to the implementation of the policy and their efforts in collecting data for our study. We are thankful to all donors for their contribution to the study and the blood supply.
Footnotes
a.meulenbeld{at}sanquin.nl, s.ramondt{at}vu.nl, m.sweegers{at}nki.nl, f.quee{at}sanquin.nl, f.prinsze{at}sanquin.nl, e.hoogendijk{at}amsterdamumc.nl, dorine.swinkels{at}radboudumc.nl, k.vandenhurk{at}sanquin.nl
Broadened the scope by adding an additional outcome and adjusted methods, results and discussion accordingly. As a result, replaced Table 2 with new outcomes. Updated figure 3 and the supplemental tables.