|Home | About | Journals | Submit | Contact Us | Français|
The aim of the study was to test the relative efficacy of action planning and reward distribution to promote retention of whole blood donors.
A sample of 7,399 donors was randomised to one of three interventions: “action planning” (n=2,585); “reward” (n=2,397); and “thank you” (n=2,417). Participants in the action planning condition were invited to write the date of their next donation on a post-it note before taking it home at the end of the donation process. Participants in the “reward” condition were given an anti-theft credit card sleeve at the end of the donation process. The “thank you” intervention is the usual condition at the end of the process; it was considered the control condition. The dependent variable was the proportion of donors who registered to give blood at six months.
Overall, 4,444 (60.06%) donors registered to give blood at six months. There were no differences between the three interventions in the proportions of donors who registered to give blood (d.f. 2, chi-square=3.72, p<0.15). However, gender modified the effect of the intervention (d.f. 2, chi-square=6.57, p<0.0375); more women registered in the “thank you” condition than in the other two.
The results suggest that action planning and the distribution of a reward have no motivational effect on the return to give blood. Nonetheless, women appear to respond more negatively to these interventions at the end of the donation process.
Blood agencies have to maintain a sufficient, safe supply of blood and blood products in order to meet the needs of the populations they serve. They are confronted with difficulties such as the aging population and the need to improve the cost-effectiveness ratio for donor retention. Overall, this creates a context favouring the investigation of more effective strategies to increase the number of individuals who return to give blood. Thus, the aim of the present study was to test the relative effectiveness of interventions to favour the return to register for blood donation among donors who gave blood at fixed sites using a randomised field trial methodology.
Two approaches, action planning and the distribution of a reward, were tested against a control condition in the everyday context of fixed blood donation sites. Study participants, because they had just successfully completed a blood donation, were assumed to have positive intentions towards blood donation, predisposing them favourably towards future donation. Moreover, the interventions selected were not heavy-handed but modest and simple, as well as inexpensive. If effective they could be easily incorporated into the usual procedures of the blood collection agency. They also respected the legal restrictions banning donor remuneration.
Action planning is a behavioural change technique that comes from the action phases model1, which facilitates the passage from intention to action by articulating certain action parameters (what, when, where, how, etc.). This technique helps individuals to decide which steps are required to achieve a given goal (e.g., give blood again). Preparing an action plan is thus a good way to help donors reach their objective regarding giving blood again.
This behavioural change technique relies on the assumption that blood donors who initially register to give blood are positively motivated to do so, making subsequent donation frequency dependent on volition rather than motivation. This assumption is well supported in the scientific literature. For instance, the mean intention score to give blood again in the next six months was 4.5 (on a scale ranging from 1 to 5) among recent blood donors (new and experimented donors)2. Therefore, removing barriers and situations interfering with the adoption or maintenance of the desired behaviour (e.g. repeated blood donation) should help to close the gap observed between donors’ positive predisposition and subsequent behaviour. For instance, if the donor is absorbed in his or her daily activities, he or she might not have or perceive an opportunity to insert blood donation into his or her schedule3. A tool to help scheduling could reasonably be expected to support return donation behaviour.
In the field of health, action planning has been applied in promoting various behaviours including blood donation and has, generally speaking, been effective in bringing about behaviour modification. In their meta-analysis of action planning’s efficacy, Gollwitzer and Sheeran4 observed an average-to-important effect size (d=0.59)5 for health-related behaviours. With regard to blood donation specifically, studies showed a more modest yet still statistically significant effect size (new donors: d=0.08, p<0.01; lapsed donors: d=0.11, p<0.01)6,7. It is worth mentioning that in these latter studies, the effect size had been calculated using analyses on intention-to-treat: all donors were included, regardless of whether they had been exposed to the intervention. Consequently, the effect size was more important when only active participants were included in the analyses (new donors: d=0.41, p<0.001; lapsed donors: d=0.47, p<0.001).
We hypothesized that individuals invited to write the date of their next donation on a specially formulated “post it” reminder and to take it home at the end of the donation process would return to give blood in a higher proportion than the proportion of donors in the other intervention conditions (hypothesis 1).
This behavioural change technique refers to Deci and Ryan’s self-determination theory8, which identifies two types of motivation: intrinsic and extrinsic. Motivation is deemed intrinsic when stemming from internalised codes of conduct or personal values (e.g. altruism), but extrinsic when action is regulated through an external source, for example in order to receive a reward, to please others, or to avoid punishment.
The World Health Organization urges nations to rely on voluntary, non-remunerated donors for their blood supply9. Some agencies or nations eschew any form of incentive and go so far as to make compensating donors illegal10. However, many blood agencies do currently use rewards (e.g. items of appreciation, lottery tickets or symbolic incentives such as a pin or medal) to entice donors and meet blood demands.
In the literature on blood donation, reports have been mixed regarding the efficacy of using incentives to prompt people to give blood11,12. Nonetheless, success with such an external motivational approach has been reported, particularly with infrequent or lapsed donors13. Consequently, certain authors have suggested reconsidering the use of such action-inducing tools in the domain of blood donation14. It was also recommended that rewards should be non-monetary as well as congruent with the act of donation, and that discretion should be used when offering them11.
According to the self-determination theory, receiving a promotional object (reward) following completion of a contingent task for a free-choice behaviour such as giving blood is likely to have a deleterious effect on intrinsic motivation15. In blood donation centres (e.g. Louisiana’s The Blood Center and Florida’s One Blood), when a reward is used, donors know to expect it; this is named task-contingent rewarding. However, where remuneration is illegal and using incentives is ethically questionable, advertising and promoting rewards is eschewed and thus the potential use of tangible rewards is quite limited. Indeed, in this latter context, in order to enter the blood donation facility in the first place, it is most likely that donors are already intrinsically motivated. Thus no effect on subsequent behaviour is expected, and offering a reward as a strategy to achieve donor retention is questionable.
We hypothesised that individuals who receive an anti-theft credit card sleeve at the end of the donation process will return to give blood in a proportion similar to the proportion of donors in the “thank you” control condition (hypothesis 2). This reward was chosen because it met several practical criteria (very low cost, small, non-perishable, easy to handle and distribute) while respecting legal restrictions regarding donor remuneration.
Héma-Québec has the mandate of collecting, processing and distributing labile blood components in the Canadian province of Quebec (population: 8 million). The majority (over 85%) of whole blood collections are obtained through mobile drives organised at intervals throughout the territory, while the rest are obtained at four fixed blood donation sites, sometimes on appointment. These four fixed sites are located in urban areas and are specifically designed to facilitate the return of blood donors because they are open seven days a week and allow donors to make appointments. Consequently, there is always an opportunity to give blood, notwithstanding the required 56-day deferral following donation.
To be included in the study, donors had to give blood at one of the four centres located either in Montreal or Quebec City, over a period of 57 days between the end of June and the end of August, 2014. This period of time prevented blood donors from being exposed more than once to the intervention and also, for practical reasons (e.g. training of staff) allowed each intervention to be offered at random several times. This procedure also had the advantage of lowering bias as it was possible to offer the same intervention on different days of the week and to rotate the staff to interventions.
The population targeted in this study comprised only allogeneic donors aged 18 years and over. A total of 9,677 individuals successfully completed a donation at one of the fixed sites during the study period. Donors were excluded if they were apheresis donors (e.g., plasma donors) (n=2,278) (see Figure 1). Thus, 7,399 donors were eligible for the study. The project was approved by Héma-Québec’s research ethics committee.
All interventions were offered on site after the donation and subsequent rest period had been completed; donors were unaware of the interventions prior to completing their donation. This moment was chosen in order to minimise the interventions’ impact on the blood donation process and avoid prolonging donation duration, a key performance metric. Moreover, in service organisations, it is acknowledged that the last contact leaves a durable impression so this moment appeared propitious16.
The interventions were delivered by regular “volunteers” working with the blood agency at the four fixed sites. The blood agency staff explained the nature of the interventions to the volunteers in a training session beforehand. In order to facilitate the interventions’ delivery, and to cope with logistical issues (e.g. limited counter space dedicated to project material), only one of the three interventions was performed on each given day, at each site. Microsoft Excel 2010 (Microsoft Corp., Redmond, WA, USA) was used to achieve randomisation when assigning interventions to days and to sites. A total of 226 intervention days were conducted, and the number of times each intervention was offered was the following: “action planning”, 75 times; “reward”, 70 times; and the control “thank you”, 81 times.
Participants in the control condition were simply thanked for their donation before being offered a beverage and healthy snack, as this is the usual procedure. This was done for all participants. In addition to the usual thank you, participants in the “action planning” condition were invited to indicate the date of their next blood donation, with the assistance of a calendar, on a specially designed post-it bearing the logo of the blood agency. The post-it also included priming words such as “solidarity” and “pride” on top before providing a space allowing blood donors to plan a date for their next donation. In order to facilitate the donors’ task, a telephone number enabling them to book an appointment was printed on the post-it. Finally, they were invited to take this post-it home with them. Participants in the “reward” condition were given an anti-theft credit card protector sleeve, an object of minimal value bearing the logo of the blood agency.
Donors were not informed of the study, as informing them of the research goals and providing explanations about the nature of the interventions would most likely have had an impact on their degree of motivation to return to donate blood and thus been a potential source of significant bias.
An objective measure of donation behaviour was obtained for each participant, namely their registration at one of the fixed sites, as recorded within the donor information system. Registration was used as the key dependent variable, because it includes both individuals who successfully donated blood and those who ended up being deferred. It takes into consideration the fact that people acted towards the goal of giving blood.
All analyses were done with SAS version 9.1 (SAS Institute Inc, Cary, NC, USA). For each of the three intervention groups, we calculated and compared descriptive statistics on demographic characteristics to verify the randomisation process. We applied the omnibus procedure GENMOD to analyse the effect of experimental conditions on the proportion of donors who presented to donate, with the “thank you” condition labelled as the control condition. Since gender, age and past donation behaviour are known to be associated with blood donation, we adjusted for these variables in all analyses. Tests for modification of intervention effects by gender, age, and past behaviour were conducted via moderated regression analysis. The categorical groupings used for age (3 categories: 18–34 years; 35–49 years; and ≥50 years) and past behaviour (3 categories: 1 only; 2–4; ≥5) are those used internally by the blood bank for data presentation and operational decision-making.
The final sample (n=7,399) consisted of 3,748 women (51.66%) and 3,651 men (49.34%); their mean age was 38.85 years (SD=16.16). The participants comprised first-time donors (n=761) and repeat donors (n=6,638) and the mean number of previous donations was 19.16 (SD=25.15). In the sample, 4,444 donors (60.06%) registered to give blood again; of these, 83.66% did so at one of the fixed sites.
The demographic and personal characteristics of the three experimental groups are shown in Table I. We found no statistically significant differences between the three groups for any of the variables, suggesting that randomisation was successful. However, each of these variables was strongly and significantly associated with the outcome variable, which was subsequent registration to give blood (gender, d.f. 1, χ2=129.98, p<0.0001; age, d.f. 2, χ2=188.79, p<0.0001; and past blood donations, d.f. 2, χ2=781.13, p<0.0001).
The GENMOD omnibus test for the effect of the intervention conditions on the registration at the targeted blood drive did not reach the level of statistical significance (d.f. 2, χ2=3.72, p<0.155). These findings did not justify more focused contrasts to assess pairwise differences between conditions.
The GENMOD procedure was used to test interaction effects between the interventions and gender, age and past behaviour. Gender significantly interacted with the interventions (d.f. 2, χ2= 6.57, p<0.05). As shown in Table II, the stronger effect of the “thank you” intervention compared with the “reward” or “action planning” interventions was limited to women. Compared to the “reward” condition, the values for women were β=0.1026, SE=0.0373, p<0.01 whereas the values for men were β=−0.0164, SE=0.0288, p=0.57. The values for women and men in comparison with the “action planning” condition were respectively, β=−0.0812, SE=0.0361, p<0.05 and β=−0.0069, SE=0.0288, p=0.81. There were no interaction effects for age and past donation behaviour.
Six months after exposure to the three different interventions, there were no statistically significant differences between the groups with regard to the proportion of donors who registered to give blood. A difference was observed between genders only with regard to the “thank you” intervention (the control group). More women registered to donate blood again in this condition than in the other two conditions; there were no differences between conditions for men.
We had hypothesised that our action planning intervention would have a positive impact on return behaviour (hypothesis 1). In the literature, action planning has been reported as an effective strategy to help people to do what they plan to do and accomplish a given task4. This did not happen in the present study. In fact, when planning was reported as an efficient strategy to increase return to give blood, the interventions clearly referred to specific barriers to action6,7. As such in these studies, coping planning was the strategy adopted; it was not action planning. The important distinction between the two dimensions of planning has been documented by Sniehotta and his colleagues17. In their research, they showed that there was no significant behavioural change when action planning was the sole intervention. Behavioural change was recorded with the combined use of action and coping planning. Based on their findings, they suggested that action planning might not be sufficient to change behaviour, and it needs to be augmented by other volitional variables, such as coping planning. Therefore, in summary, interventions focusing on coping planning may provide a better basis for future field-based research, especially in the domain of blood donation.
With regards to the use of a reward, our study focused on the use of a non-task-contingent reward, the reward being given as a “token of appreciation”. This was determined in accordance with the legal necessity to eschew any type of donor remuneration and the accompanying policy of not advertising non-monetary incentives, a stance consistent with the strictly voluntary nature of blood donation advocated by the WHO9. Other research has shown that while cash payments do diminish the supply of blood donors, through a motivational crowding out effect, this is not necessarily the case for reward tools such as non task-contingent tokens of appreciation, in-kind rewards or payments that can be donated to charity, which can have a positive impact on donor supply11,18,19. Our research did not show a statistically significant impact on donor return for those who were offered tokens of appreciation, confirming our hypothesis in this regard (hypothesis 2). Consequently, our results do not support the use of this type of strategy in retaining blood donors, at least not in a context in which advertising is not permitted and the target population consists of voluntary donors who are probably intrinsically motivated to perform a pro-social action.
The two interventions were compared to the standard, baseline approach whereby donors are only thanked for their blood donation. Thanking donors is generally accepted as a basic element of good customer service. Gratitude or acknowledgement from staff is identified among the interactions that can contribute to producing positive feelings of affection and provide donors with behavioural confirmation, elements which help sustain ongoing donations11,20.
Interestingly, our results showed that among female donors the proportions who returned to give blood in the action planning and reward conditions were lower than the proportion observed in the standard thank you condition. This suggests that the other two interventions had a dampening effect on female donors’ propensity to re-enlist. This interpretation is also supported by our observation that after completion of the present study, the proportions of male and female donors who returned to give blood (men: 66.7%; women: 55.8%) in a follow-up period of 6 months were similar to those observed in the control (thank you) condition of our study (see Table II; men: 66.4%; women: 57.0%). Regarding the use of a reward tool, Mellström and Johannesson18 found a significant motivational crowding out effect in women when compensation for donation was introduced, but not for men: the introduction of a monetary payment reduced the supply of female blood donors by almost half and was perceived as de-valuing their donation and perceived contribution. The authors linked this result to the idea that monetary payments may reduce intrinsic motivation to behave altruistically. For instance, it was documented that compared to men, women show an increased perceived altruistic value to giving blood and are more likely to donate blood for humanitarian reasons21. Thus, female donors in our study may have responded negatively to the reward intervention because they are more prone to being intrinsically motivated.
However, with regard to the action planning intervention, it is difficult to identify a specific reason for the dampening effect on returning to give blood among women. Previous testing of the moderating effect of gender on action planning (i.e., implementation intentions) did not show differences in responses between men and women6. In that study both men and women returned to give blood more often than the donors in the control condition. Nonetheless, a possible explanation is the recent observation that “perceived pressure to donate” was negatively associated with return to donate blood22. It is thus possible that compared to men, women are more prone to associate the action planning approach as an undue pressure to donate again and consequently they react negatively. Additional studies will be needed to test this specific outcome for the use of action planning compared to coping planning.
Lastly, some may point out that motivation levels may differ between donors who choose to give blood at fixed sites and those who donate at mobile blood drives. It is possible that conducting the study at Héma-Québec’s fixed donation sites may have attenuated our capacity to detect a difference between interventions, as a result of potential differences in donor motivation. While this does not diminish the internal validity of our study’s conclusions, it may warrant caution regarding generalisation of the results to a mobile blood drive setting. This provides another opportunity for further research, although the action-planning intervention tested here in the context of fixed sites would be inappropriate for mobile blood drives given that donors may not know the time and date of the next donation opportunity.
The results of this study indicate that action planning and the distribution of a non-behaviour-contingent token of appreciation have no motivational effect on the return to give blood. However, women appear to respond more negatively to these behavioural change approaches. In order to promote donor retention in areas in which the WHO’s recommendations for entirely voluntary blood donation are adopted, further research involving coping planning instead of action planning may be warranted, taking into account gender differences.
Furthermore, bearing in mind these recommendations and despite this study’s negative results, it may be beneficial to conduct field-based research in order to test similar inexpensive interventions in lower income countries in which rates of replacement and first-time donors are higher.
Funding and resources
This work was supported by internal funds from Héma-Québec.
All Authors made substantial contributions to the research design, or the acquisition, analysis or interpretation of data. They all contributed to drafting the paper and revising it critically, and have approved the submitted and final versions.
The Authors declare no conflicts of interest.