Results from the blood donor competence, autonomy, and relatedness enhancement (Blood Donor CARE) randomized trial

Abstract

Background:

This study aimed to promote competence, autonomy, and relatedness among first-time whole blood donors to enhance intrinsic motivation and increase retention.

Study Design and Methods:

Using a full factorial design, first-time donors (N=2,002) were randomly assigned to a no-treatment control condition or to one of seven intervention conditions designed to promote donation competence, autonomy, relatedness, a combination of two (e.g., competence and autonomy), or all three constructs. Participants completed donor motivation measures before the intervention and six weeks later, and subsequent donation attempts were assessed for one year.

Results:

There was no significant group difference in the frequency of donation attempts or in the number of days to return. Significant effects of group were observed for 10 of the 12 motivation measures, although follow-up analyses revealed significant differences from the control group were restricted to interventions that included an autonomy component. Path analyses confirmed direct associations between interventions involving autonomy and donor motivation, and indirect mediation of donation attempts via stronger donation intentions and lower donation anxiety.

Conclusion:

Among young, first-time, whole blood donors, brief interventions that include support for donor autonomy were associated with direct effects on donor motivation and indirect, but small, effects on subsequent donation behavior.

In the US, first-time blood donors contribute a significant portion of the annual blood supply¹ with many of these donations coming from high school and college blood drives. Unfortunately, long-term retention of young, first-time donors in the donor pool is a significant challenge.^2–4 For some, this may reflect an inability to transition from external forms of motivation (e.g., peer-pressure) that initiated and maintained donation behavior while in school. Others may have significant internal motivation to continue giving, but this motivation may not translate into donation behavior in the face of potentially competing internal barriers (e.g., donation-related fears) and external barriers once they leave school (e.g., less convenient access to a donation site; conflicting work schedule).

Self-Determination Theory⁵ proposes that people are more likely to persist with behaviors that are internally versus externally motivated, and that innate psychological needs for competence, autonomy, and relatedness are essential to foster the development of internal motivation.^5–7 Competence refers to the need to experience a sense of self-efficacy or ability to attain valued outcomes within one’s environment. In the blood donation context, for example, a negative experience (e.g., a vasovagal reaction) may reduce one’s self-efficacy as a donor, whereas a positive experience (e.g., overcoming fear) may promote self-efficacy and contribute to a greater intrinsic interest in engaging in future donations. Autonomy refers to having volitional control over one’s behavior and the freedom to choose activities that are consistent with one’s personal goals and sense of self. Provision of external incentives (e.g., T-shirts) and pressure to re-donate on a pre-determined schedule (e.g., every 8 weeks) may reduce a blood donor’s sense of autonomy,^8–13 whereas a non-judgmental assessment of each donor’s unique reasons for giving, as well as how donation behavior fits in with their broader goals and values, would promote autonomy. Relatedness refers to the desire to be connected to others. One’s perception of social or relational support facilitates expression of intrinsic motivation, and therefore in the blood donation context those who feel a greater sense of belonging or connection to a larger community of donors will be more likely to internalize and integrate the values of that group.^10,14 In sum, according to Self-Determination Theory, efforts to promote basic needs for competence, autonomy and relatedness should enhance the development of intrinsic motivation and thereby increase the likelihood of future donation.

In line with this notion, the current study was designed to promote these three psychological needs among first-time blood donors in order to assess their effect on retention as a donor. To promote donor competence, or sense of self-efficacy as someone who can donate blood, we created a website to provide recent first-time donors with empirically-validated strategies to address common donor concerns related to pain, fear, and potential syncopal reactions. Prior evidence suggests exposure to these coping materials significantly increases both self-efficacy and donation intention among non-donors and donors alike.^15–18 To promote donor autonomy we used a post-donation intervention informed by both motivational interviewing and action planning principles.^19,20 In two prior studies we demonstrated that a post-donation motivational interview designed to promote autonomy with respect to future donation planning significantly increased donation attempts among both experienced²¹ and first-time²² donors. Finally, to promote donor relatedness, we created a Facebook group with the goal of increasing feelings of connectedness and affiliation with other donors by promoting social interaction around themes of blood donation.^23–28

Using a full factorial design, we randomized first-time blood donors to either a control condition or to one of seven intervention conditions designed to promote donation competence, autonomy, relatedness, a combination of two (e.g., competence and autonomy), or all three constructs. Participants completed measures of blood donor motivation before and after the intervention, and subsequent donation attempts were assessed for one year from their next date of eligibility. We hypothesized that, compared to the control condition, each of the interventions would increase donor motivation, and particularly internal motivation for giving, resulting in an increased likelihood of making a subsequent donation attempt.

MATERIALS AND METHODS

Recruitment and Participants

Participants were recruited from the database of New York Blood Center (NYBC; New York, NY, USA). Eligibility requirements included whole blood donors identified as being first-time donors with NYBC in the previous week, 16 to 24 years old at the time of their donation, and deemed eligible to donate again. In addition, those contacted had to indicate that they had or were willing to establish a Facebook account, and that they were willing to be randomly assigned to an intervention group. Participants were excluded if they reported having donated blood with another blood collection agency or more than once with NYBC. Invitations to participate in the study were emailed to 23,064 individuals between 5/27/2016 and 6/10/2019, and enrollment closed on 6/19/2019 when a sufficient sample had been recruited.

Procedure

Study invitations were sent by email in the week following eligible donations. Interested donors (and the parents of minor-age donors) were directed to the study website where a full study description was available. Donors who provided informed consent (or parental informed consent and assent for minor-age donors) were then linked to the online baseline assessment materials. Study data were collected and managed using Research Electronic Data Capture (REDCap) tools^29,30 hosted at Ohio University. All interactions with participants were via email, telephone, and social media.

At the completion of the baseline survey, respondents were automatically randomized by the REDCap system. Using a computer-generated random number sequence created by the study statistician with an equal allocation ratio and block sizes of 8, respondents were randomly assigned within gender to one of eight parallel groups: 1) Treatment-as-Usual Control, 2) Competence, 3) Autonomy, 4) Relatedness, 5) Competence + Autonomy (C+A), 6) Competence + Relatedness (C+R), 7) Autonomy + Relatedness (A+R), or 8) Competence + Autonomy + Relatedness (C+A+R).

Participants in the control group received an email informing them that they would be contacted again to complete a second survey in approximately six weeks. Participants in the intervention groups received an email notification describing their group assignment and providing information regarding how to complete their next step in the study. For those assigned to more than one intervention, administration occurred in a set order (Competence, Autonomy, and then Relatedness) and email instructions for the next assigned step were sent within a few days of completing the prior assignment. Although participants could not be blinded to their group assignment due to the nature of the interventions, there was no potential bias by an examiner because the baseline and post-intervention assessments were completed online. Follow-up data from the donor records was collected by blinded blood center staff. Finally, interventionists were unaware whether a participant was assigned to any additional intervention and moderators of the Facebook group were unaware whether participants had also been assigned to review the website or complete an interview call.

Control.

Participants in the control group received the standard NYBC communications for first-time donors, including: 1) a thank you email (approximately 4 days post-donation), 2) delivery of a donor card (approximately 2 weeks post-donation), and 3) an email reminding them of their eligibility to give blood again (8 weeks post-donation). These communications were also sent to participants in the intervention conditions.

Competence intervention.

Participants assigned to the competence intervention were asked to review a donor coping website.¹⁵ Using a combination of text, videos, and interactive features, the website is designed to address common donor fears and offers advice on empirically-validated strategies to reduce fear, pain, and syncopal reactions. Participants were required to view the website material and respond to some individualized questions online to complete this intervention. Adherence was confirmed based on website recording of page views and completion of all questions.

Autonomy intervention.

Participants assigned to the autonomy intervention were asked to participate in a telephone-based motivational interview. The scripted call encouraged donors to reflect on their motivations for giving and how the act of donating is consistent with their broader life goals and values, with the opportunity for the discussion to be tailored to address unique issues raised by each donor.³¹ Topics covered included individual motivations for giving, the relationship between the donor’s personal goals/values and their decision to donate, and developing individualized action and coping plans to deal with anticipated barriers. Interviewers were trained undergraduate and graduate students and calls were recorded (with participant permission) for ongoing interviewer supervision and to ensure treatment fidelity. On average, these calls lasted 14.2 (SD = 5.3) minutes.

Relatedness group.

Participants assigned to the relatedness group were invited to join a closed Facebook group for one month. The group was moderated by members of the study team who posted donation relevant images, graphics, messages, polls and links to encourage discussion of donation, social interaction around donation experiences, and feelings of affiliation among group members.³² Moderator posts were uploaded twice daily using a rotating 30-day schedule so that all participants were exposed to the same material during their assignment to the Facebook group. Participants were required to actively accept the group invitation to count as having completed this intervention, and they had access to the group materials for 30 days.

Participants were required to complete an assigned intervention before they could proceed to the next intervention, and had to complete all of their assigned interventions within six weeks in order to be eligible to complete the post-intervention assessment. Two reminders (at 3- and 6-days after instructions for each step were issued) were given via email and voice or text message to encourage participants to complete their next step in the study. Participants who were eligible to complete the post-intervention assessment were emailed a link to the survey six weeks after completing their baseline survey. To encourage participation and retention, participants who completed their assigned intervention(s) and the post-intervention survey received a check for US$100.

The NYBC donor database was used to track subsequent donation behavior. All donation attempts in the 421 days following each participant’s initial donation (to allow for a one-year follow-up after an eight-week waiting period for whole blood donation) were retrieved for the 2,002 participants who had completed the baseline assessment and were assigned to a study group. The last follow-up window closed on 7/31/2020.

These procedures were approved by the Ohio University and NYBC Institutional Review Boards. The study was registered with ClinicalTrials.gov (NCT02717338) and additional study detail is available from a published protocol.³³

Baseline and post-intervention assessment measures

Table 1 describes the measures completed at baseline and post-intervention. All scales except Donor Autonomy³⁴ and Decisional Balance³⁵ were scored by calculating the mean across all items. Donor Autonomy was calculated by weighting the six individual regulatory style scores from the Blood Donor Identity Survey [i.e., (−3 × Non-regulated) + (−2 × External Regulation) + (−1 × Introjected Regulation) + (1 × Identified Regulation) + (2 × Integrated Regulation) + (3 × Intrinsic Regulation)], with higher scores reflecting greater autonomous motivation.³¹ Decisional Balance was computed by calculating the mean pros and cons subscale scores for each participant, and then subtracting the cons subscale score from the pros subscale score; hence, scores above zero reflect a heavier weighting of pros to cons.

Table 1.

Assessment measures completed at both baseline and post-intervention.

Measure	Description	Sample item	Item Anchors	Cronbach’s α (baseline/post)
Donor Competence	Confidence or self-efficacy in one’s ability to engage in blood donation and to cope with barriers that may arise (6 items).33	“I am confident that I can cope with any concerns I may have about donating blood.”	1 (Not at all) to 7 (Extremely)	.90/.91
Donor Autonomy	Degree of autonomous motivation to donate blood assessed using the Blood Donor Identity Survey (18 items).34	“Donating blood is consistent with my life goals.”	1 (Not at all true) to 7 (Very true)	.76/.77
Donor Relatedness	Perceptions of belonging to or a sense of connection with the blood donation community (9 items).32	“I feel accepted by other blood donors.”	1 (Not at all) to 7 (Extremely)	.88/.88
Attitude	Thoughts and feelings about blood donation (6 items).42	“For me, donating blood within the next year would be…”	1 (Unpleasant) to 7 (Pleasant)	.81/.81
Subjective Norm	Perceived social pressure regarding blood donation (6 items).42	“The people who are most important to me think I should give blood in the next year.”	1 (Disagree) to 7 (Agree)	.89/.92
Perceived Behavioral Control	Confidence and perceived ability to donate blood within the next year (6 items).42	“How confident are you that you will be able to donate blood within the next year?”	1 (Not very confident) to 7 (Very confident)	.84/.86
Intention	Intention to donate blood again in the next year (5 items).42	“I plan to donate blood in the next year.”	1 (Disagree) to 7 (Agree)	.97/.97
Moral Norm	Perception of the moral correctness of donating blood (3 items).31	“I feel a moral obligation to give blood.”	1 (Strongly disagree) to 7 (Strongly agree)	.80/.80
Anticipated Regret	Expectation of negative feelings if do not donate within the next year (3 items).31	“If I do not donate blood within the next year, I will regret it.”	1 (Very unlikely) to 7 (Very likely)	.95/.95
Anxiety	Level of nervousness, tension, and anxiety with respect to donating blood (3 items).31	“If I donate blood I will feel nervous.”	1 (Not at all) to 7 (Very much)	.96/.96
Decisional Balance	Relative importance of perceived pros and cons of donating blood (12 items).35	“I may be helping somebody in my community.”	1 (Not at all important) to 5 (Extremely important)	Pros .86/.87 Cons .74/.76
Commitment	Level of commitment to engage in future blood donation assessed using the Blood Donation Ambivalence Survey (3 items).43	“I am sure that I will donate blood in the future.”	1 (Not at all true) to 7 (Very true)	.78/.82

In addition to the above measures, at baseline participants provided basic demographics [e.g., age, self-reported gender (female, male, transgender), race/ethnicity, donation history] and rated their recent donation in terms of vasovagal symptoms using the Blood Donation Reactions Inventory,³⁶ pain associated with the blood sample finger stick and the donation needle using 0 (“no pain”) to 100 (“worst possible pain”) scales, and overall satisfaction with the donation experience using a 0 (“not at all satisfied”) to 100 (“completely satisfied”) scale.³⁷

Statistical Analysis

Target sample size was calculated based on the minimum sample required to examine intervention effects on donor retention and potential mediators of intervention effects. Based on sample size calculations previously reported,³³ a minimum of 1,944 donors at baseline and follow-up (and 243 donors per group) were needed to achieve at least 80% power at α = 0.05 across the specified hypotheses. To confirm that random assignment resulted in groups that did not differ at baseline, a series of univariate ANOVAs was conducted for continuous variables and Chi-square tests for categorical variables.

Analysis of the effect of the interventions on donation attempts in the follow-up period was conducted with two approaches. First, to investigate whether there was a difference in the distribution of participants who did and did not return among the intervention groups, a Fisher’s exact test was conducted using Monte Carlo simulation with 30,000 replications. Second, a Cox proportional hazards survival analysis was conducted with number of days to first return attempt as the dependent variable (right censored) and intervention group as the independent variable. The survival analysis was conducted first without potential moderators, and then again while examining potential individual moderators of age, gender, race, ethnicity, vasovagal symptoms during the recent donation, donation needle pain, and overall satisfaction with the recent blood donation experience.

Next, to investigate the effect of the interventions on assessment survey measures, 12 separate ANCOVAs (one for each outcome) were conducted with the post intervention score as the dependent variable, the baseline score as the covariate, and the intervention group as the independent variable. Significant effects were followed up with post-hoc comparisons using a Holm-Bonferroni adjustment³⁸ to investigate whether any of the interventions differed significantly from the control condition.

Finally, a path analysis was conducted to examine the role of each of the assessment measures as potential indirect mediators of the relationship between the interventions and subsequent donation behavior. The model used donation attempt (yes/no) as the behavioral outcome, and baseline assessment measures were included as covariates on the post-intervention measure. To maximize the sample size included in the path analysis, the model was fitted using a maximum likelihood estimation method with robust standard errors (estimation method MLR in Mplus), and a χ2 difference test was conducted after removing or adding a path to ensure improved fit.

Analyses were conducted using SAS for Windows Version 9.4 (SAS Institute, Cary, NC) and Mplus Version 8.4, and p < 0.05 was used to denote statistical significance.

RESULTS

Sample Characteristics

Figure 1 provides an overview of the flow of participants from initial recruitment until the end of the follow-up period. Of those invited, 3,907 visited the study website and informed consent was provided by 3,532; subsequently 2,598 of these individuals completed the baseline assessment survey. Of these, 596 were excluded from the study for various reasons, including 572 with a prior donation history, 2 over the age of eligibility, 1 with no match in NYBC donor records, 5 who did not donate whole blood, and 16 who were not randomized due to clerical error.

Figure 1.

Participant flow diagram. Interventions are coded as C (competence), A (autonomy), or R (relatedness).

A total of 2,002 recent first-time whole blood donors were randomly assigned to one of seven intervention groups or a no-intervention control group (Table 2). The groups did not differ significantly in terms of gender, age, race, ethnicity, or average baseline assessment measure scores (all p>.10). In addition, the groups did not differ significantly in terms of their first donation experience ratings (i.e., vasovagal symptoms, pain, satisfaction).

Table 2.

Baseline demographics and mean (SD) assessment measure scores for each group.

Measure	Control (n=250)	C (n=250)	A (n=248)	R (n=260)	C+A (n=251)	C+R (n=250)	A+R (n=248)	C+A+R (n=245)
Gender
• Female	62.8%	64.0%	65.3%	66.5%	65.3%	69.6%	66.1%	62.1%
• Male	36.8%	35.2%	33.9%	33.1%	33.9%	30.0%	33.5%	36.7%
• Transgender	0.4%	0.8%	0.8%	0.4%	0.8%	0.4%	0.4%	1.2%
Age	19.4 (2.4)	19.4 (2.5)	19.4 (2.5)	19.5 (2.5)	19.2 (2.4)	19.4 (2.5)	19.3 (2.4)	19.5 (2.6)
Race
• AI/AN	0.0%	2.0%	0.8%	1.2%	1.2%	0.4%	0.8%	0.8%
• A	20.8%	14.4%	16.5%	17.3%	15.5%	20.0%	16.1%	19.6%
• B/AA	8.8%	8.4%	11.3%	10.4%	7.6%	8.8%	9.3%	7.8%
• H/PI	0.4%	0.0%	0.8%	0.8%	0.8%	0.4%	1.2%	0.0%
• M/O	19.6%	16.4%	16.9%	16.2%	16.3%	21.6%	17.7%	18.0%
• W	50.4%	58.8%	53.6%	54.2%	58.6%	48.8%	54.8%	54.1%
Ethnicity
• Hispanic	24.0%	26.8%	23.4%	22.7%	27.5%	24.8%	24.2%	22.9%
• Not Hispanic	76.0%	73.2%	76.6%	77.3%	72.5%	75.2%	75.8%	77.1%
Donor Competence	5.8 (1.2)	6.0 (1.0)	5.9 (1.0)	5.9 (1.1)	5.9 (1.0)	5.8 (1.1)	5.9 (1.1)	6.0 (1.0)
Donor Autonomy	11.7 (10.2)	11.6 (8.6)	12.1 (10.0)	11.8 (10.2)	12.7 (9.5)	12.1 (10.2)	11.1 (9.6)	11.6 (9.1)
Donor Relatedness	4.9 (1.1)	4.9 (1.2)	4.9 (1.3)	4.9 (1.2)	4.8 (1.2)	4.9 (1.2)	4.9 (1.2)	4.9 (1.2)
Attitude	5.9 (1.0)	5.9 (1.0)	5.9 (1.0)	5.9 (1.0)	6.0 (0.9)	5.9 (1.1)	5.8 (0.9)	5.9 (1.0)
Subjective Norm	4.6 (1.6)	4.5 (1.5)	4.5 (1.6)	4.4 (1.6)	4.7 (1.6)	4.5 (1.6)	4.6 (1.6)	4.5 (1.6)
Perceived Behavioral Control	6.2 (1.0)	6.2 (1.0)	6.3 (1.0)	6.3 (0.9)	6.3 (0.8)	6.2 (0.9)	6.2 (0.9)	6.2 (0.9)
Intention	6.1 (1.4)	6.2 (1.2)	6.2 (1.3)	6.1 (1.3)	6.2 (1.1)	6.0 (1.4)	6.1 (1.2)	6.2 (1.2)
Moral Norm	4.7 (1.6)	4.6 (1.5)	4.7 (1.6)	4.7 (1.6)	4.7 (1.5)	4.7 (1.5)	4.8 (1.4)	4.8 (1.5)
Anticipated Regret	4.3 (2.0)	4.5 (2.0)	4.5 (1.9)	4.4 (2.0)	4.5 (1.9)	4.4 (1.9)	4.6 (1.7)	4.6 (1.9)
Anxiety	2.9 (1.9)	3.1 (1.9)	3.1 (1.9)	3.2 (1.9)	3.1 (2.0)	3.2 (2.0)	3.4 (2.1)	3.1 (2.0)
Decisional Balance	2.4 (1.1)	2.3 (0.9)	2.4 (1.0)	2.4 (0.9)	2.3 (1.0)	2.3 (1.1)	2.3 (1.0)	2.3 (1.0)
Commitment	5.6 (1.4)	5.7 (1.4)	5.6 (1.4)	5.6 (1.4)	5.6 (1.4)	5.6 (1.4)	5.7 (1.3)	5.7 (1.3)

Note: Interventions are coded as Competence (C), Autonomy (A), or Relatedness (R). Race is coded as AI/AN = American Indian or Alaskan Native; A = Asian or Asian-American; B/AA = Black or African-American; H/PI = Hawaiian or other Pacific Islander; M/O = More than one race or Other; W = White.

Intervention effects on donation behavior

Across the entire sample, 48.0% made at least one donation attempt over the 14-month follow-up period since their initial donation. Results of the asymptotic Fisher’s exact test indicated that there was no significant difference in the distribution of donation attempts by intervention group (p = 0.98; 95% CI 0.9784– 0.9816). The percentage of donors in each group who attempted a donation in the follow-up interval was: Control = 46.4%, Competence = 47.2%, Autonomy = 48.8%, Relatedness = 45.4%, C+A = 49.0%, C+R = 48.4%, A+R = 49.6%, and C+A+R = 49.0%.

Results of the Cox proportional hazards survival analysis indicated that there was no significant difference in the survival rate among groups, Wald χ2(7) = 1.2247, p = 0.9903. This supports the non-parametric Kaplan-Meier/Product-Limit estimate which compared the mean of survival time among groups, Kaplan-Meier log-rank χ2(7) = 1.2297, p = 0.9902. The mean (Standard Error) number of days to return in each group was: Control = 295.5 (8.8), Competence = 292.5 (7.9), Autonomy = 300.6 (8.7), Relatedness = 289.4 (8.3), C+A = 300.8 (9.0), C+R = 300.0 (9.1), A+R = 294.8 (8.9), and C+A+R = 295.1 (9.3). Similarly, there was no main effect of intervention group nor any group by moderator interactions when the Cox proportional hazards survival analysis was repeated with potential individual moderators of age, gender, race, ethnicity, or recent donation vasovagal symptoms, donation needle pain, or satisfaction.

Intervention effects on assessment measures

Results of the ANCOVAs of post-intervention scores on each of the dependent measures revealed significant effects of intervention group on competence, F(7,1737) = 3.65, p = 0.0007, autonomy, F(7,1737) = 5.62 p < 0.0001, attitude, F(7,1737) = 4.45, p < 0.0001, perceived behavioral control, F(7,1737) = 2.16, p = 0.0346, intention, F(7,1737) = 3.75, p = 0.0005, moral norm, F(7,1737) = 7.22, p < 0.0001, anticipated regret, F(7,1737) = 2.93, p = 0.0047, anxiety, F(7,1737) = 5.64, p < 0.0001, decisional balance, F(7,1737) = 3.89, p = 0.0003, and commitment, F(7,1737) = 7.01, p < 0.0001. No significant group effects were observed for donor relatedness, F(7,1737) = 1.30, p = 0.2484, and subjective norm, F(7,1737) = 1.70, p = 0.1053.

Table 3 summarizes the results of the follow-up analyses, which reveal a pattern of significant differences from controls for interventions that included autonomy alone or in combination. In contrast, neither the competence intervention alone, the relatedness intervention alone, nor the combination of competence and relatedness was associated with any significant differences relative to controls.

Table 3.

Post-intervention unadjusted means (SD) on assessment measures for each group. Shaded cells indicate a significant difference from Control on post-hoc comparisons.

Measure	Control (n=235)	C (n=235)	A (n=226)	R (n=219)	C+A (n=218)	C+R (n=208)	A+R (n=209)	C+A+R (n=196)
Donor Competence	5.9 (1.1)	6.0 (1.1)	6.0 (0.9)	5.9 (1.1)	6.2 (0.9)	5.8 (1.1)	6.2 (0.9)	6.1 (0.9)
Donor Autonomy	11.7 (9.9)	11.3 (9.4)	13.8 (9.2)	12.1 (10.5)	15.0 (9.1)	11.9 (10.6)	12.7 (9.2)	12.9 (9.0)
Donor Relatedness	4.9 (1.2)	4.9 (1.2)	4.9 (1.2)	5.0 (1.2)	5.1 (1.1)	5.0 (1.2)	5.1 (1.2)	5.0 (1.1)
Attitude	5.9 (1.0)	6.0 (0.9)	6.0 (1.0)	5.9 (1.0)	6.1 (0.9)	5.8 (1.1)	6.0 (0.9)	6.1 (0.8)
Subjective Norm	4.6 (1.6)	4.6 (1.7)	4.6 (1.7)	4.5 (1.7)	4.9 (1.5)	4.6 (1.6)	4.9 (1.5)	4.6 (1.6)
Perceived Behavioral Control	6.2 (0.9)	6.2 (1.0)	6.3 (0.9)	6.3 (0.8)	6.4 (0.7)	6.1 (1.1)	6.3 (0.9)	6.2 (0.9)
Intention	6.1 (1.2)	6.3 (1.2)	6.3 (1.0)	6.2 (1.2)	6.4 (0.9)	6.0 (1.4)	6.4 (1.0)	6.3 (1.1)
Moral Norm	5.0 (1.4)	4.8 (1.4)	5.0 (1.4)	4.9 (1.5)	5.3 (1.3)	4.9 (1.5)	5.2 (1.3)	5.2 (1.4)
Anticipated Regret	4.7 (1.8)	4.8 (1.8)	4.8 (1.7)	4.7 (1.8)	5.1 (1.7)	4.7 (1.8)	5.2 (1.6)	5.0 (1.8)
Anxiety	3.1 (1.9)	3.2 (1.8)	2.9 (1.7)	3.2 (1.9)	2.8 (1.9)	3.0 (1.9)	3.1 (1.9)	2.7 (1.8)
Decisional Balance	2.3 (1.1)	2.4 (1.0)	2.4 (1.0)	2.3 (1.1)	2.6 (1.0)	2.3 (1.1)	2.4 (1.0)	2.4 (1.1)
Commitment	5.7 (1.4)	5.7 (1.4)	5.8 (1.3)	5.7 (1.3)	6.1 (1.0)	5.6 (1.4)	6.0 (1.0)	6.0 (1.2)

Note: Interventions are coded as Competence (C), Autonomy (A), or Relatedness (R).

Path analyses of the intervention, assessment, and behavior relationships

Path analyses were conducted to examine the donor motivation measures as potential indirect mediators of the relationship between the interventions and subsequent donation behavior. The results of the first model, which used donation attempt (yes/no) as the outcome, not surprisingly confirmed the associations noted in the aforementioned ANCOVAs, and also demonstrated additional significant paths from the autonomy intervention alone or in combination with the competence and/or relatedness interventions (Figure 2; Table 4). The path analysis also demonstrated that a donation attempt was more likely among those with stronger post-intervention intention (p < 0.001) and less likely among those with greater post-intervention anxiety (p = 0.025).

Figure 2.

Path analysis of the relationships among donation intervention, assessment measures, and subsequent donation attempts. Only significant pathways are illustrated, with the exception of paths from baseline assessment measures to the equivalent post-intervention measure, which were all significant at p < 0.001 but are not illustrated for clarity.

Table 4.

Results of path analyses of the relationships among donation intervention, assessment measures, and subsequent donation attempts (yes/no).

Path		Coefficient	p-value
From	To
Baseline Donor Competence	Donor Competence (Post)	0.600	< 0.001
Competence + Autonomy		0.168	0.002
Autonomy + Relatedness		0.224	< 0.001
Competence + Autonomy + Relatedness		0.157	0.004
Baseline Donor Autonomy	Donor Autonomy (Post)	0.744	< 0.001
Autonomy		1.890	< 0.001
Competence + Autonomy		2.285	< 0.001
Autonomy + Relatedness		1.744	< 0.001
Competence + Autonomy + Relatedness		1.673	0.001
Baseline Donor Relatedness	Donor Relatedness (Post)	0.670	< 0.001
Competence + Autonomy		0.128	0.036
Baseline Attitude	Attitude (Post)	0.723	< 0.001
Competence + Autonomy		0.166	< 0.001
Autonomy + Relatedness		0.093	0.041
Competence + Autonomy + Relatedness		0.190	< 0.001
Baseline Subjective Norm	Subjective Norm (Post)	0.728	< 0.001
Competence + Autonomy		0.217	0.012
Autonomy + Relatedness		0.178	0.026
Baseline Perceived Behavioral Control	Perceived Behavioral Control (Post)	0.626	< 0.001
Competence + Autonomy		0.098	0.043
Competence + Relatedness		−0.133	0.032
Baseline Intention	Intention (Post)	0.694	< 0.001
Autonomy		0.164	0.001
Competence + Autonomy		0.165	0.007
Autonomy + Relatedness		0.176	0.001
Competence + Autonomy + Relatedness		0.183	0.005
Baseline Moral Norm	Moral Norm (Post)	0.629	< 0.001
Autonomy		0.176	0.015
Competence + Autonomy		0.430	< 0.001
Autonomy + Relatedness		0.332	< 0.001
Competence + Autonomy + Relatedness		0.329	< 0.001
Baseline Anticipated Regret	Anticipated Regret (Post)	0.648	< 0.001
Competence + Autonomy		0.273	0.004
Autonomy + Relatedness		0.345	< 0.001
Competence + Autonomy + Relatedness		0.214	0.027
Baseline Anxiety	Anxiety (Post)	0.700	< 0.001
Autonomy		−0.251	0.009
Competence + Autonomy		−0.417	< 0.001
Competence + Relatedness		−0.191	0.047
Autonomy + Relatedness		−0.356	< 0.001
Competence + Autonomy + Relatedness		−0.464	< 0.001
Baseline Decisional Balance	Decisional Balance (Post)	0.661	< 0.001
Competence + Autonomy		0.236	< 0.001
Autonomy + Relatedness		0.119	0.058
Baseline Commitment	Commitment (Post)	0.536	< 0.001
Autonomy		0.176	0.031
Competence + Autonomy		0.440	< 0.001
Autonomy + Relatedness		0.340	< 0.001
Competence + Autonomy + Relatedness		0.302	< 0.001
Intention Post	Return	0.476	< 0.001
Anxiety Post		−0.061	0.025

To illustrate the magnitude of the intervention effect on donor return via post-intervention intention, a graph of the predicted probability of return was created for the four groups that differed significantly from controls on baseline-adjusted post-intervention intention scores (i.e., Autonomy; C+A; A+R; C+A+R). For each of these significant interventions, estimates of post-intervention intention were calculated at baseline intention values of 1 through 7 (i.e., the full range of intention scores), and then the predicted probability of return was calculated. Each of the four significant interventions was associated with a higher predicted probability of donor return as compared to the control group, with approximately 2% more returns predicted among those with the highest baseline intentions (Figure 3).

Figure 3.

Predicted probability of a subsequent donation attempt for the four interventions (i.e., Autonomy; Competence + Autonomy; Autonomy + Relatedness; Competence + Autonomy + Relatedness) that differed significantly from controls on baseline-adjusted post-intervention intention. The intervention curves are all higher than control, but they cannot be distinguished due to their overlap.

DISCUSSION

With just under half of the sample attempting a subsequent donation over the 14-month follow-up interval, analyses of the proportion of returnees in each group clearly demonstrated that there was no direct effect of the interventions on donor return behavior. In contrast, the intervention groups had significant effects on 10 of the 12 donor motivation variables, with the autonomy intervention having the most reliable effect on these outcomes, particularly when combined with the competence or relatedness interventions, or both. In fact, each of these intervention conditions was associated with higher levels of donor autonomy and lower levels of donation-related anxiety relative to the control condition. Hence, the autonomy intervention appears to have bolstered a sense of internal motivation for giving blood while helping to address fear or anxiety about the donation process. It is also worth noting that although follow-up analyses of intention did not yield significant individual group differences relative to controls using adjusted p-values, stronger donation intentions were observed at the p < 0.05 level for all intervention groups that included autonomy alone or in combination. Interestingly, the two motivation variables where there was no significant effect of group were subjective norm and relatedness, which implies that exposure to the interventions did not influence the perceived social pressure for donating blood or the participants’ sense of connection to other donors. It is not surprising that these outcome variables were not influenced by the content of the competence or autonomy interventions, but they do suggest that the brief exposure to the Facebook group was not sufficient to promote a stronger sense of community with, or accountability to, other donors. This may be because the Facebook group was not actively moderated to encourage interaction; rather, it included a standardized set of rotating posts with some that encouraged social interaction among members. The clear pattern of effects in favor of interventions involving autonomy, with adjunctive but not main effects of the competence and relatedness interventions, supports the notion that the observed effects are not merely based upon providing these new donors with greater attention or engagement. Indeed, if that were the case then one might expect the largest benefits to have accrued from the Facebook intervention due to its explicit focus on bringing donors together for support.

In contrast to the absence of a direct effect of the interventions on donor behavior, results of the path analysis revealed that the interventions did have an indirect effect on donation behavior via donation intention and anxiety. That is, groups that included the autonomy intervention alone or in combination were associated with both higher donation intentions and lower donation anxiety. In turn, higher intentions and lower anxiety were associated with an increased likelihood of a repeat donation attempt during the 14-month follow-up interval. Although these indirect effects were statistically significant, they represented only about two percent more predicted donation attempts.

Taken together, the analyses of both direct and indirect effects indicate that the autonomy intervention had the most consistent influence on donor motivation. While this intervention was significantly related to decreased anxiety and increased autonomy and intention (at p<.05) on its own, in combination with other interventions it was associated with significant effects on additional donor motivation variables including competence, attitude, moral norm, decisional balance, and commitment. Thus, the autonomy intervention, which included a combination of reflection on personal motivation for giving and development of individualized action and coping plans to deal with anticipated barriers,^31,39,40 may be the most promising approach to bolstering donor motivation, but it also appears to benefit when supplemented by additional efforts to support novice donors’ ability to cope with donation-related concerns (e.g., fear, pain, and syncopal reactions) and to connect with others.

Although the observed motivational effects were small, they must be considered in the context of strong baseline motivation levels; on average, participants entered the study with scores of at least 6 on 1–7 scales of donation intention, competence, and perceived behavioral control. Thus, “ceiling effects” undoubtedly limited the opportunity to observe larger changes in these measures. Further, while these findings provide support for the importance of fostering autonomy to increase donor motivation and behavior, the results are not as robust as some earlier reports. One potential reason for this discrepancy is that increases in donation intention^21,31,41 and behavior^21,22 in prior studies were observed among experienced donors, suggesting that the interventions may have a greater impact among those who have developed a stronger habit or identity as a donor. At the same time, there are likely to be inherent limits in the ability of any intervention to further promote a particularly strong habit. This is supported by recent evidence that an online motivational interview increased donation intention, but not behavior, among highly experienced donors (averaging 35 prior donations) who already returned at a rate of more than 80% in an equivalent 14-month follow-up interval.⁴¹ A related but distinct difference between the current study and prior reports is the fact that the present sample included high school age donors. High school students not only have less donation experience, but their donation behavior is likely to be influenced by different situational factors than older donors. For example, high school donors are in an environment where there can be significant peer pressure to donate, unique incentives such as an excused absence from class, and scheduled in-school blood drives. Additionally, retention of these young donors may be negatively affected when they leave school and are required to be more active in seeking a donation opportunity. Including younger donors in the study also presented a methodological challenge in accurately tracking return behavior, as graduating students may be more likely to move during the follow-up interval as compared to prior samples of older donors. Finally, it is important to note that the opportunity to attempt another donation was curtailed for a portion of the sample as the window for their follow-up assessment overlapped with the onset of COVID-19 pandemic restrictions. Specifically, with the cancellation of in-person classes for high schools and colleges in the New York area in March of 2020, the opportunity to provide an additional donation at school was shortened for 17.3% of the sample (i.e., those recruited after January 2019), and this may have led to lower retention than would otherwise have been seen.

In sum, among young, first-time, whole blood donors, brief interventions that support autonomy are associated with direct effects on donor motivation and indirect effects on donation behavior. Although the observed effect on retention was small during this study’s 14-month follow-up, those whose internal motivation was enhanced by the interventions may choose to donate again, possibly even years later.