|Home | About | Journals | Submit | Contact Us | Français|
The ageing population and recent migration flows may negatively affect the blood supply in the long term, increasing the importance of targeted recruitment and retention strategies to address donors. This review sought to identify individual, network and contextual characteristics related to blood donor status and behaviour, to systematically discuss differences between study results, and to identify possible factors to target in recruitment and retention efforts.
The systematic review was conducted in accordance with a predefined PROSPERO protocol (CRD42016039591). After quality assessments by multiple independent raters, a final set of 66 peer-reviewed papers, published between October 2009 and January 2017, were included for review.
Individual and contextual characteristics of blood donor status and behaviour were categorised into five main lines of research: donor demographics, motivations and barriers, adverse reactions and deferral, contextual factors, and blood centre factors. Results on donor demographics, motivations and barriers, and contextual factors were inconclusive, differing between studies, countries, and sample characteristics. Adverse reactions and deferral were negatively related to blood donor behaviour. Blood centre factors play an important role in donor management, e.g., providing information, reminders, and (non-)monetary rewards. No studies were found on network characteristics of (non-)donors.
Although individual and contextual characteristics strongly relate to blood donor status and behaviour, mechanisms underlying these relations have not been studied sufficiently. We want to stress the importance of longitudinal studies in donor behaviour, exploring the role of life events and network characteristics within blood donor careers. Increased understanding of donor behaviour will assist policy makers of blood collection agencies, with the ultimate goal of safeguarding a sufficient and matching blood supply.
In the Netherlands, approximately 2.5% of the population are registered as whole blood or plasma donors and account for 721,000 donations per year, providing about 25 whole blood units per 1,000 inhabitants. However, the number of donors in the Netherlands has been decreasing from more than 400,000 donors in 2010 to about 340,000 donors in 20151. Although this does not pose a short-term threat to the blood supply, due to an even larger decrease in blood demand influenced by advanced surgery techniques and a more restrictive transfusion policy2–4, certain demographic developments may negatively affect the blood supply in the long term.
First, men in their 50s and 60s are overrepresented in the Dutch donor pool5. Within one to two decades, these men will no longer be eligible to donate and a new generation of blood donors needs to be available. However, recruiting and retaining young donors is difficult6.
Second, due to recent migration flows, the diversity of the population is growing and with it, the diversity of patients in need of specific blood and tissue types. Consequently, new and more donors with specific characteristics (e.g., male, ethnic minority) need to be recruited to safeguard a sufficient and matching blood supply.
The Netherlands is not the only country facing these developments7. Hence, recruitment and retention of blood donors is an important study topic. Over the last 40 years, researchers have studied donor behaviour, trying to characterise the “typical blood donor”8–10. However, findings are inconclusive, with results changing over time and varying within and between countries.
This systematic review provides an overview of the great variety of results on characteristics of blood donor status (e.g., first-time, novice, experienced) and behaviour (e.g., donation frequency, return behaviour). We update previous systematic reviews11,12 and extend them by exploring the role of external factors in donating blood. For example, through studies showing how contextual and blood centre factors are related to willingness to donate and actual donor behaviour13,14.
Furthermore, we try to shed light on donor careers. If human behaviour were static, all donors who ever started donating would continue to do so for the rest of their lives. In contrast, after people sign up as donor, some do not return after their first donation, while others continue to give blood until their maximum eligible age. These individual behavioural sequences and corresponding donor statuses are what we define as donor careers. We explore how donor careers play a role in previous study results.
From a practical perspective, this review is helpful in assisting policy makers of blood collection agencies. By presenting and comparing recent findings, blood collection agencies can design and implement evidence-based recruitment and retention campaigns to address (non-)donors in the most effective ways, convince first-time donors to become regular donors, and to guarantee a safe and sufficient blood supply in the future.
The main goal of this systematic review is to answer the following question: what individual, contextual and network characteristics are related to blood donor status and blood donor behaviour, and do these relationships change over time?
This review is conducted in accordance with a predefined PROSPERO protocol (CRD42016039591)15. For this review we searched for studies on individual, network and contextual characteristics of blood donor status and behaviour.
Studies matching our search terms were collected using Google Scholar, PubMed, ScienceDirect, and Web of Science. Core keywords of the search were: (blood) AND (donor OR donation) AND (motivation OR attitude OR behaviour OR recruitment OR retention OR altruism OR centre OR network OR life event). To make sure we collected all possible relevant literature for review, we conducted a manual search in some of the most relevant journals on prosocial behaviour and blood transfusion published between October 2009 and January 2017 (e.g., Transfusion, Vox Sanguinis, Voluntary and Nonprofit Sector Quarterly).
We built on and extended two earlier systematic reviews11,12 and included studies published after October 2009, the last month of inclusion in one of these comparable reviews11, to provide an up-to-date review without replicating former systematic analyses. We decided not to shift our inclusion date further to February 2012 (last month of inclusion in the most recent systematic review)12 because this would have required us to exclude 20 relevant papers related to blood donor status and behaviour, not discussed in either of the former reviews.
Additional details on the search strategy and review process can be found in the PROSPERO protocol15.
As a first step in the study selection process, papers that matched our search terms were assessed on title and abstract, based on six inclusion criteria: (i) published in English, German or Dutch; (ii) published in a peer-reviewed journal; (iii) published after October 2009; (iv) conducted in a Western country; (v) used quantitative methods; and (vi) used blood donor status or blood donor behaviour as an outcome measure. Of the 399 studies retrieved, 307 were rated as clearly ineligible.
Thereafter, two reviewers (TWP, EFK) independently read and evaluated the full text of the 88 remaining studies. Again, the selection was based on the six inclusion criteria. Disagreements on inclusion of specific studies were resolved by discussion. As a result of this critical evaluation, another 26 studies were excluded from review.
As a final step, we conducted a quality control of the 66 studies included using a combination of four open-access critical appraisal tools for quantitative research: Critical Appraisal Skills Programme (CASP UK)16, Strengthening the Reporting of Observational Studies in Epidemiology (STROBE)17, Standard Quality Assessment Criteria (Qualsyst)18, and the Critical Review Form19. These tools have been designed by epidemiologists, methodologists, and statisticians to improve both the quality of reporting on individual studies and the critical evaluation of study reports. Each appraisal tool has its own strengths and focus, but none of them incorporated a rating scale of all study characteristics relevant to our review. We, therefore, combined questions from each appraisal tool to carefully rate the included studies and all of the subparts (see Appendix A, Table I for the individual scores on these items for each study included and Appendix B, Table II for the items of the developed review form).
The quality of all studies was assessed by one rater (TWP), while four other raters (E-MM, RB, WdK, EFK) divided the studies among them. Weighted Cohen’s kappa (κw) showed a moderate to good agreement between the raters, κw=0.608 (95% confidence interval: 0.384–0.832), p<0.000. Major disagreements on the inclusion of specific papers were resolved by discussion to define the final set of 66 included studies (Figure 1).
Most of the reviewed studies stem from Northern Europe (n=25) and North America (n=24), while the remainder were conducted in Southern Europe, Australia and New Zealand. The characteristics of these studies varied, with sample sizes ranging from 190 to 2.1 million, consisting of a minimum of 24% to a maximum of 80% male participants. Study designs and methods also differed, including field experiments, randomised controlled trials and descriptive studies based on registry data (Appendix C, Table III). We systematically summarise and discuss the studies’ results, classified into two main categories: individual and contextual characteristics of donor status and behaviour (Figure 2).
More than half (n=35) of the studies reported on socio-demographic characteristics of donors and nondonors, including sex, age, race and ethnicity, religion, education, employment, income, and demographic transitions.
The results on sex differences among donors and non-donors are mixed. Nine of the 21 studies reported that men were more likely to be donors than women20–28, ten studies reported the opposite5,7,29–36, while two found no sex differences37,38. After a first donation, men were more likely to return than women23,32,33,39–42. One study showed that this relationship was present in the short term (6-month follow-up)7, while another study only found a long-term difference (25-month follow-up)43. Men donated more frequently6,7,21,31–33,44,45 and were more often multigallon donors (more than 10 lifetime donations) compared to women46.
Studies on the relationship between age and donor status and behaviour also showed mixed results. Two studies indicated that the likelihood of donating increased with age30,47, while three others stated that younger people were more likely to donate31,37,38. Four studies found a nonlinear association, with older people being more likely to donate until a certain age, after which the propensity decreased5,7,25,28. One study found no relationship between age and blood donation20. Younger people were more likely to be first-time donors than older people23,41,42 and repeat donors were older than first-time donors7,41,42. The return rate of older donors was higher than that of younger donors6,40,43. In the United States, where the minimum eligible donor age is 16, the highest return rates were found for donors between the age of 16 and 1843. Older people were more likely to be frequent givers and multigallon donors, compared to younger people6,21,30,31.
Regarding race and ethnicity, six studies from the United States, Great Britain and the Netherlands found higher rates of donors among (non-Hispanic) whites from the United States, and people with a British or Dutch ethnic background, respectively, compared to other race and ethnic groups5,7,31,33,34,48. Besides, (non-Hispanic) whites donated more frequently and were more likely to return than African-Americans, Asians, and Hispanics31,33.
Three studies examined the relation between religion and donor status. No relation was found in Spain20, while only limited evidence (positive relation for Catholic men aged 35–44) was found in the United States48. In contrast, another study from the United States found a positive relation for both organisational and subjective dimensions of religion (respectively church attendance and involvement in religious groups, and importance of faith in daily life)50.
With regard to education, five studies found higher education to be related to a higher propensity to donate5,20,30,39,48. Five others found an inverted U-shape relationship25,28,29,34,37, while one study found no relationship between education and donation38. Men with medium or higher education were more likely to have donated blood compared to men with a lower educational level, which did not hold for women47.
In two studies, donors did not differ from non-donors in terms of their employment status37,38, while one study showed that unemployed people were more likely to be non-donors39. Regarding donor status, Gemelli et al.42 showed that first-time donors were more likely to be students than returning donors, while the group of returning donors had higher numbers of retired people, professionals and tradespeople compared to the group of first-time donors.
Two studies found no relationship between personal income and donating38,51, one study found a negative relationship34, and two others concluded that people with a high (family) income were more likely to be donors than people with a low (family) income37,48.
Among donors in the Netherlands, the proportion of those either married or never married was larger than in the general population5. Married people were also more likely to be donors in Canada29. In contrast, being married lowered the chance of being a donor in Germany and the United States34,38. In Spain, people who were divorced or widowed had a 50 percent higher chance of donating blood20. Gillum and Masters48 found that being married was positively related to the likelihood of being a donor for men, but not for women. Having children lowered the chance of being a donor38.
Motivations and barriers to donating blood have been widely studied, mainly along three lines of research: self-reported motivations and barriers, the Theory of Planned Behaviour (TPB)52, and (mechanisms of) altruism.
Five studies examined self-reported motivations21,27,29,36,39, and five investigated self-reported barriers to donate blood6,27,30,39,53 (Figure 3). Certain motivations to donate differed between members of socio-demographic groups21,29. Members of socio-demographic groups who are more likely to go through life events that might affect blood donation (e.g., studying, pregnancy) were more likely to cite motivations and barriers to donate associated with these events29,30,53.
Six studies used the TPB to predict donor behaviour. Only the intention to donate was robustly correlated with donor behaviour. Other variables in the TPB model (self-efficacy, subjective and moral norm, affective and cognitive attitude, and role identity) explained little if any variance when intention was included54–56.
Dutch donors were marked by high levels of intention, attitudes and self-efficacy36. Multigallon donors scored higher on self-efficacy, affective attitude and self-identity than occasional donors46. Affective attitude was positively related to return behaviour, while pressure to donate showed a negative relationship. Higher levels of self-efficacy, cognitive attitudes, affective attitudes and subjective norms were associated with lower levels of dropout57.
In studying altruism as a motivation to donating blood, Bolle and Otto51 found no difference in the level of altruism between donors and non-donors (i.e., total amount of money donated to a charitable cause after filling in an online questionnaire served as a measure of their level of altruism). Evans and Ferguson58 proposed a refinement of the general altruism concept, arguing that there are five theoretically distinct dimensions of altruism: impure altruism, kinship, self-regarding motives, reluctant altruism, and egalitarian warm-glow. Donors consistently scored higher than non-donors on feelings of warm-glow and reluctant altruism, but not on other forms of altruism59.
Regarding the donor career, cognitive and behavioural motives (e.g., intentions, self-efficacy and habit formation) showed associations in all stages of the donor career. For first-time and novice donors, reluctant altruism was a distinguishing factor, while experienced donors were marked by warm-glow and pure altruism. Impure altruism was higher among first-time donors than novice and experienced donors60.
No differences in levels of susceptibility to social influence between donors and non-donors appeared in one study61. Among participants who were aware of the need for blood, those who were asked to give blood were more likely to donate in the upcoming blood drive than those who were not asked to make an active decision62.
As the self-reported barriers indicated, negative donation experiences and deferral might be reasons to lapse or stop donating6,30,53. Nine studies explored the role of adverse reactions (e.g., fainting, needle reactions) and deferral (e.g., low haemoglobin, travelling abroad) on donor status and behaviour.
Donors who experienced an adverse reaction showed lower return rates than donors who did not experience an adverse event42,57,63. This stopping risk increased with the severity of the reaction64,65 and had a higher impact on first-time donors than repeat donors41,64–66. One study found that vasovagal reactions and fatigue, but not needle reactions, were negatively related to return rates35. Mixed results were found on the relationship between both age64,65 and sex35,41,57,64, and return rates after an adverse reaction.
Donors who were temporarily deferred were less likely to return (especially for first-time donors and longer deferral periods) and had lower donation frequencies after deferral42,67,68. Age and education were positively related to return after deferral67.
Besides individual characteristics, context also plays a role in blood donor behaviour. For example, it was found that children raised in a “blood donor family” were more likely to become donors themselves69. Here we discuss contextual characteristics by differentiating between person-related factors (i.e., urbanisation, community characteristics, collective life events) and blood centre factors.
In Spain and the United States, no differences were found in likelihood to donate between people from rural and urban areas20,48. In contrast, German municipalities with a larger population reported lower donation rates70. Although there were no differences in donation frequency between urban and rural areas in the United Kingdom, people from London donated less than those in other regions7. In contrast, Canadian, Greek and Serbian donors from metropolitan areas either showed higher propensities to donate28, or higher donation frequencies31,39 than donors from non-metropolitan areas. In Switzerland, living in an urban area was associated with higher chances of becoming an inactive donor40.
In a German study, communities with a higher percentage of people aged 30 and above, a lower percentage of non-German inhabitants and a lower percentage of unemployed people had higher donor rates70. There were no differences in propensity to donate between people living in low, middle or high-income regions20,70. In Canada, communities with a higher proportion of singles and a lower proportion of children had higher annual donation rates. In contrast, communities with a higher proportion of educated people and higher immigrant rates had lower annual donation rates44.
As mentioned before, different motivations might be important for people in deciding to donate blood. One way these motivations can be triggered is by the occurrence of a life event. One study examined the effect of a collective, traumatic life event on donor behaviour71. In the weeks after the terror attacks on September 11th 2001, the number of first-time donors was almost three times higher than in September 2000. However, the return rate of these first-time donors did not differ. Women and older people were more likely to become loyal donors compared to men and younger people.
Blood collection agencies play an important role in donor management, for example by providing information about donating blood, reminding donors about their next donation opportunity, and offering monetary or symbolic rewards.
The effectiveness of information and reminders provided by blood collection agencies was tested in seven experimental studies. People who read a short educational brochure on blood donation (e.g., information on the need for blood and the donation process) were more likely to sign up for an upcoming blood drive compared to people who read a standard blood bank brochure or a brochure unrelated to blood donation72. This effect was also present when donation-anxiety was heightened (i.e., in the presence of a mobile blood collection unit)73.
Both first-time and active donors who were reminded to donate by telephone were more likely to donate than (first-time) donors who were not reminded22,74. Overall, men and older people were most likely to donate after a call75. First-time donors were especially likely to donate when they received both an informational brochure and a telephone call reminder76. Among repeat donors, the combination of a telephone call and an email reminder had a positive effect on return rates of men, but not women77.
Moreover, not only the presence and content of the promotional and educational material of the blood collection agency matters, the framing of this material might also influence donor behaviour78,79. First, Moussaoui et al.79 found that the return rate of lapsed donors did not vary when they received a donation invitation framed with a “save lives” message or a neutral invitation. Chou and Murnighan78 even found that donors where more likely to donate at an upcoming blood drive when they received a loss-framed message (i.e., “help prevent someone from dying!”), than a gain-framed message (i.e., “help save someone’s life!”).
Canadian donors mentioned absence of a nearby blood drive as an important reason for reduced donation frequencies30. Within German communities, there was a relationship between available donation sites and donation rates for mobile sites but not for fixed sites70. The relationship had an inverted U-shape, with the positive association decreasing as the number of sessions rose.
The effects of monetary rewards have been tested in experimental settings. When Italian donors were rewarded with a day’s paid leave, employed donors made, on average, one donation extra per year (most likely on Mondays and Fridays to extend their weekends) compared to self-employed or unemployed donors24. Donation frequencies increased with the monetary value of incentives offered. Furthermore, donors, especially younger ones, were more likely to donate in places where higher rewards were offered80.
Regarding non-monetary rewards, the number of donors and frequency of donations in the United States increased when symbolic incentives were offered (e.g., t-shirts, coupons, mugs) and when their perceived value increased. If another donation site close to the donors’ standard centre offered an incentive, donors were more likely to move to that centre and adjust their timing to receive the incentive81. When comparing the influence of private and public symbolic rewards (respectively receiving a medal and being mentioned in the local newspaper) on donation frequency in Italy, Lacetera and Macis45 found that donors only increased their frequency when the thresholds for the public rewards were within reach. There was no decrease in donation frequency after these quotas were reached.
First-time donors satisfied with the overall donation experience were more likely to return to donate than those who were (moderately) unsatisfied22. However, only 1% of lapsed donors reported that dissatisfaction with the personnel was a barrier to donating30. Satisfaction with medical personnel was lower for younger donors28.
For men, but not for women, increased waiting time at the donation site decreased return rates82. Among active donors, 28% reduced their donation frequency because of waiting time, while 23% of the donors mentioned it as a reason for lapsing30.
There might be several other ways in which blood collection agencies can influence donor behaviour. A post-donation telephone interview (e.g., identifying motivations, making a donation plan) increased the likelihood of a subsequent donation, but not the donation frequency34. Van Dongen et al.66 also showed the importance of donation planning, as for the third donation decision only planning failure was a significant predictor of non-return behaviour.
The goal of this review was to identify individual, network and contextual characteristics that relate to blood donor status and behaviour, and to systematically discuss differences between study results. We found empirical evidence on five main lines of research: donor demographics, motivations and barriers, adverse reactions and deferral, contextual factors, and blood centre factors.
Demographic characteristics are strongly related to donor status and behaviour. However, the results vary considerably between studies, countries, and sample characteristics. There is no general profile in terms of certain socio-demographic features that is characteristic of first-time, loyal, frequent, or non-donors. Individual (non-)donor behaviour cannot be fully understood without taking into account the widely ranging cultural and historical contexts on national and regional levels. Previous research has suggested that donor profiles vary between blood collection regimes because donors are recruited using different strategies83.
Self-reported barriers were quite consistent between studies. However, barriers to donate blood varied between members of different socio-demographic groups. Regarding motivations, we can conclude that blood donation is not just a purely altruistic act. Motivations to donate blood are dynamic and multidimensional, and include both self- and other-regarding motives. These findings have implications for blood collection agencies, as more tailored recruitment and retention campaigns might be able to address barriers and motivations for (non-)donors from specific socio-demographic groups more effectively.
Adverse reactions and deferral are negatively related to donor behaviour, especially for first-time donors. There might be a relation with donors’ age and sex, but these results are inconclusive and understudied.
Regarding contextual factors, we cannot draw any strong conclusions. Results on urbanisation and community characteristics are mixed, with no clear differences to be found between or within countries. Furthermore, we recognise culture as an important contextual factor84,85, but none of the studies investigated its role in relation to blood donor status or behaviour.
Blood collection agencies play an important role in blood donor behaviour. Providing information and reminders were effective ways of boosting attendance rates. Experimental studies on (non-)monetary rewards also showed promising results. However, since all studies were performed in two USA and Italian cities by the same research group, more research is needed to draw conclusions on the generalisability of the results. Some other blood centre factors play a role as well (e.g., decreasing waiting time, planning future donations), but too few studies have investigated these factors to conclude on their effectiveness.
No studies were found on network characteristics of donors and non-donors. Although some studies included parental and partner status5,20,38,48, these relations could not be attributed exclusively to (social) networks, but also represent demographic transitions.
Systematic reviews are limited by the quality of the available studies and, more specifically, the representativeness and comparability of findings. Several of the studies included in this review failed to describe basic characteristics of the sample (e.g., mean age, percentage of men and women), while others relied on non-random samples of university studies, making it difficult to generalise the study results and draw reliable conclusions.
With regard to comparability, a variety of concepts were used to study the same topic (e.g., community characteristics, self-reported barriers), while others used different definitions of donor status, making it difficult to compare findings across studies. In order to enable international comparisons, we would recommend the use of DOMAINE definitions86 to characterise groups of donors and their behaviour.
Despite the limitations, this review can serve as a basis for future research. First, we want to emphasise the importance of donor careers. Most research on donor behaviour and motivations used cross-sectional methods without taking into account that people and their behaviour might change. However, Ferguson et al.60 showed how altruistic motives to donate blood differed between first-time, novice and experienced donors. We encourage the use of dynamic approaches and methods, following individual donors across several years to investigate motivational changes.
Moreover, if behaviour and motivations change over time, it will be interesting to explore how, when and why these changes take place. One possibility might be the occurrence of a life event. Collective events seem to have an effect on donating blood for the first time71, but based on self-reported barriers30,53, we can assume that individual events have effects as well. For example, health-related events might increase the awareness of need for blood or feelings of social responsibility, making it more likely that a person starts donating or increases the frequency of donating. Other life events can influence the network characteristics of donors (e.g., moving to another city), which affect the propensity to donate due to decreased network influences.
Second, it would be worthwhile to further explore network characteristics to discover how family ties and peer pressure influence individual donor behaviour. Bani and Strepparava21 found that around 22% of the respondents were influenced by family and friends in their decision to donate, while Pedersen et al.69 suggested that familial and heritable influences could be even stronger, extending beyond the donors’ own awareness.
Finally, we suggest paying attention to the broader level of contextual and blood centre factors. These are important from a practical point of view because they can be influenced by blood collection agencies (e.g., providing information and reminders proved to have a positive influence on donor behaviour). Current research can be improved by modelling blood centre factors in hierarchical (“multilevel”) models to investigate the role of contextual characteristics (e.g., regional differences within countries) in these blood collection agency strategies. They can also be tested in field experiments, which allow for stronger causal inferences.
In summary, this systematic review has provided an overview of the recent literature on individual and contextual characteristics related to blood donor status and behaviour. If the great diversity of the results have one thing in common, it is that blood donor behaviour cannot be understood from one set of (non-) donor characteristics, as we have already stressed the importance of cultural and historical contexts in individual behaviour. Research on donor behaviour should try to explore the interrelationships between the individual, contextual, and network levels (e.g., multilevel designs and longitudinal studies), which could help us to better understand donor behaviour, and further assist blood collection agencies in designing tailored recruitment and retention strategies. We hope that this will contribute to safeguarding a sufficient and matching blood supply in the future.
This project was funded by Sanquin Research, PPOC15-32.
The Authors declare no conflicts of interest.