It is not all anti‐policy attitudes: Viewing blood donation as a means to belong predicts blood donation among men who have sex with men

Abstract

Objective

To examine whether specific attitudes towards blood donation vary across MSM demographic groups; and preliminarily test whether these attitudes predict prior donation behaviour.

Background

Insights into blood donors' attitudes facilitate effective donor management, especially as policies for men who have sex with men (MSM) become more inclusive. While attitudes towards donating blood can predict donor decisions, it is unclear whether attitudinal factors specific to MSM groups will be important considerations for blood service providers engaging with this newly eligible donor group.

Methods

Data were drawn from a large (N = 3157) online cross‐sectional survey of MSM in New Zealand. New items were developed to assess three attitudinal constructs: negative attitudes towards the MSM deferral policy, favourable attitudes towards blood services, and the tendency to view blood donation as a form of civic participation and belonging. Univariate analyses descriptively characterised demographic factors associated with those various attitudes, and multivariate logistic regressions were used to predict self‐reported donation histories.

Results

Negative views of deferral policies and blood services varied across sample characteristics in univariate analyses, but did not systematically differ between donors and nondonors. However, a perception of blood donation as a means of civic belonging was consistently associated with greater donation, including 1.46 times the odds of having ever donated (p < 0.001) and 1.40 times the odds of having donated recently under the current 3‐month deferral policy (p = 0.017).

Conclusion

These findings highlight the importance of recognising broader civic‐related attitudes in MSM donor research and engagement strategies.

1. INTRODUCTION

Understanding the attitudes, motivations, and beliefs of blood donors (and potential donors) is important to helping blood services maintain a safe and sufficient blood supply. Blood services need to nurture and expand their donor bases, as most countries' blood supplies rely on the generosity of a small number of non‐remunerated community members. In New Zealand (NZ), the national blood service has recently committed to finding ways to make policy more inclusive for MSM, ¹ while progressive shifts to assess donor risk at an individual level (rather than using a blanket, time‐based behavioural criterion) are already underway in multiple countries. ² , ³ , ⁴

Engaging prospective MSM donors may prove challenging for blood services given the negative publicity towards deferral policies. ⁵ , ⁶ Although empirical work in this area is only just beginning, it seems likely that the decision to donate or not is influenced by men's attitudes towards policies, services, and blood donation in general. It is unclear, however, what the stance of newly eligible MSM will be. To maximise blood supply, blood services need to understand how attitudes towards their services and deferral policies may vary between and within such newly eligible populations, as well as assess how such attitudes may predict patterns of donation behaviour. In contributing to research in this area, the current report assesses whether attitudes regarding blood donation services and policies vary across demographic groups of MSM, and whether attitudes predict donation experience and recency.

Attitudes, an affectively imbued appraisal of objects, ideas, or behaviours, ⁷ are important in blood donation contexts for several reasons. First, potential demographic predictors of donation, ⁸ such as age, ethnicity, education, are static and comparatively immutable; although they can help identify target groups, they offer limited opportunities for interventions aimed at maintaining or expanding the donor pool. By contrast, attitudes are comparatively fluid and can be modified through education and communication strategies. For example, blood services will often use promotional messages thought to elicit a positive attitudinal response towards donating blood to recruit new donors and to motivate existing donors to return (e.g., “Be a hero, donate blood. Save lives!”). ⁹ Viewed in this light, attitudinal research is vital for blood donor management because it can help identify specific and malleable factors that can be targeted to enhance recruitment and retention efforts.

Attitudes have long been recognised as likely influential in blood donation behaviour. According to one common framework, health behaviours (such as donating blood) are influenced by intentions that are shaped, in part, by positive or negative evaluations of the behaviour itself. ¹⁰ For example, research across diverse populations, professions, and donation contexts ¹¹ , ¹² , ¹³ , ¹⁴ , ¹⁵ has consistently shown that a positive attitude towards blood donation (e.g., seeing blood donation as rewarding) is associated with greater donation intentions and behaviour. Further, several cross‐sectional surveys ¹⁶ , ¹⁷ , ¹⁸ , ¹⁹ and reviews of blood donation motivations and barriers have suggested that attitudes towards the blood service influence the decision to donate. ²⁰ , ²¹ , ²² Of note, cynicism regarding blood services was especially influential in donors' decision to not return to donate, while nondonors were deterred from engaging with services with a poor reputation. ²⁰ , ²¹ , ²² Consistent with this emergent picture, experimental studies imply that negative attitudes towards blood services causally contribute to a decrease in willingness to donate. ²³ , ²⁴ However, while these prior studies are important to understanding the possible role of attitudes in patterns of blood donation, current attitudinal research is limited in some important ways.

First, prior research has been conducted in general population samples without a specific focus on what may be newly eligible groups of donors – MSM. The historical exclusion of MSM from donation combined with impending policy changes might mean that MSM hold distinct attitudes towards blood services and deferral policies that impact their donation behaviour in particular ways. For instance, some MSM donors whose blood tested positive for a transfusion‐transmitted infection have reported knowingly donating while ineligible, citing disagreement with the deferral policy as a motivating factor, ²⁵ while others did not comply with deferral policies because they had concerns with confidentiality and the potential disclosure of sexuality. ²⁶ , ²⁷ , ²⁸ Such attitudes can impact blood safety and establish MSM as a key population for blood services. In contributing to research in this area, the current paper provides descriptive data characterising attitudes that might be particularly salient among MSM (regarding the MSM deferral policy and the overall blood services) as well as general blood donation attitudes in a large sample (N = 3157) of MSM in NZ.

Second, attitudinal research has tended to focus on the potential links between attitudes and intention to donate, rather than studying donor behaviour. A recent review found that, although attitudes towards the act of donating blood predicted future donation behaviour – mainly via intentionality – intentions were less useful in predicting future blood donation behaviour compared to other types of charitable behaviours. ²⁹ In theory, intentions are influenced by other situational factors ³⁰ (e.g., fear of needles, availability of time, proximity of donor centres) that may intervene and disrupt the attitude–behaviour relationship should the opportunity to donate blood arise. In a similar vein, using intentions as a proxy for donation behaviour might be less useful among minority groups (such as MSM communities), given the barriers such groups have historically faced. ³¹

Conversely, exploring the factors associated with past donation behaviour may be more reliable in predicting future behaviour. While intentions may be important, investigating the direct link between attitudes and behaviour could help identify the attitudinal factors unique to both donors and nondonors. With this information, blood services will have clear targets for strategies aimed at increasing donor interest and turnout. Such attitudes in the general donor population typically include the desire to be conscientious or altruistic and wanting to perform a civic duty. ²¹ , ³² In contributing to this understudied area among the MSM population, the current paper examines the associations between blood donation attitudes and men's history of blood donation behaviour.

Finally, prior studies of attitudes in blood donation contexts have tended to employ sub‐optimal measurement techniques. Inconsistent measurement could explain empirical inconsistencies in terms of whether particular attitudes predict blood donation behaviour ¹⁸ , ³³ , ³⁴ , ³⁵ , ³⁶ , ³⁷ , ³⁸ , ³⁹ , ⁴⁰ , ⁴¹ , ⁴² , ⁴³ , ⁴⁴ or not. ⁵ , ⁶ , ⁴⁵ , ⁴⁶ , ⁴⁷ , ⁴⁸ , ⁴⁹ In addition to indexing attitudes towards donation behaviour and practices, research has also measured attitudes towards services ⁵ , ⁶ , ¹⁸ , ³³ and general motivations to donate. ⁵ , ⁶ , ¹⁸ , ³³ , ³⁷ , ³⁸ , ⁴¹ , ⁴² , ⁴⁴ , ⁴⁵ , ⁴⁶ , ⁴⁷ , ⁴⁸ , ⁴⁹ However, attitudes are often indexed using single items ⁵ , ⁶ , ¹⁸ , ³³ , ⁴² or operationalised in aggregate forms ¹⁸ , ³⁷ , ³⁸ , ³⁹ , ⁴¹ , ⁴³ , ⁴⁴ , ⁴⁵ , ⁴⁶ , ⁴⁷ , ⁴⁸ , ⁴⁹ that obscure what the “attitude” is actually about. Combining items that represent attitudes about different phenomena (e.g., attitudes towards services, behaviours, motivations, and knowledge) obscures their individual impact on donor behaviour, thus reducing the clarity regarding which attitudes blood services should target. In contributing to this research, the psychometric strategy used to construct the attitudinal measures for this report was pre‐registered (see: https://aspredicted.org/ZMY_YKN).

Of particular note, it may be important to consider the meaning associated with blood donation behaviour among MSM. Reviews of donor motivation in general populations have consistently identified civic values as a key influence on the decision to become a donor and donate blood. ¹⁷ , ²¹ , ³² , ⁵⁰ For example, presentations to local blood services often increase during periods of adversity: the 9/11 terrorist attack, ⁵¹ the COVID‐19 pandemic ⁵² and, within the MSM community, in reaction to the Pulse nightclub shooting. ⁵³ Importantly, while blood donation motivations emphasising civic engagement are prosocial in nature, they also reflect a fundamental desire to belong in society. For example, in a cohort of prospective MSM donors in Canada, many viewed the ability to donate blood as an important part of contributing to the wider community and being a “civic minded person”. ⁵⁴ Although robust evidence is lacking, our suspicion here is that similar motivations are likely to prove especially important in historically marginalised groups. Hence, in contributing to research in this area, the current report tests for possible differences between groups of donating and non‐donating MSM in the extent to which they view blood donation as a meaningful expression of civic belonging.

1.1. The Present Study

Using a large sample of MSM recruited in New Zealand, the aim of the current report was twofold: (a) to psychometrically investigate the attitudinal items created to measure MSM's attitudes towards blood donation services, deferral policies, practices; and (b) to examine whether these attitudinal constructs are uniquely associated with past and recent donation behaviour after controlling for demographic and MSM‐specific characteristics.

2. METHODS

2.1. Design

The Sex and Prevention of Transmission Study is a self‐completed, voluntary, confidential, online, cross‐sectional national HIV behavioural surveillance survey in NZ. Participants were recruited between April and August 2022 through advertisements on social media, gay dating mobile applications, national television and print media, via community organisations' mailing lists, and posters in public and gay‐friendly venues across NZ. Consent was obtained prior to the online questionnaire hosted on Salesforce/SurveyTitan. The study was approved by the Health and Disability Ethics Committee (EXP11450). Details of the behavioural surveillance elements of the parent study are reported elsewhere. ⁵⁵

2.2. Participants

People were eligible if they were over 16 years old, lived in NZ, and identified as a gay, bisexual, or non‐heterosexual man (cis or trans) or reported having had sex with another man at least once. For the current analysis, participants were excluded if they did not respond to the first question of the blood donation section of the survey (“Are you interested in donating blood?”). Of the 3838 participants recruited for the study, 3157 individuals were eligible for this analysis.

2.3. Measures

2.3.1. Background variables

Factors collected from the sample included age, ethnicity, region, education level, employment status, disability status, sexual identity, and use of pre‐exposure prophylaxis (PrEP) against the transmission of HIV. For analysis, age was simplified into three categories (<30, 31–44, and 45+ years). Ethnicities were prioritised as per NZ Census Standards in the following hierarchical order: Māori, Pacific, Asian, Other, and NZ European.

2.3.2. MSM social connection

A classification of MSM's social connection was measured using a binary categorical variable in two questions: “How much of your free time is spent with other gay, bisexual, takatāpui, or other MSM?” (A lot of time/Not a lot of time) and “How open are you about your sexual orientation/sexual identity?” (Out to most people/Not out to most people).

2.3.3. Attitudinal measures

In the absence of a previously validated scale suited to the local context, we used prior studies to adapt nine attitudinal items relating to blood donation behaviour and services. To enhance validity, participants were reminded of the current policy which stated: “The current blood donor policy in NZ does not allow men who have sex with men (MSM) to donate blood for 3‐months following anal or oral sex with another man. People on PrEP cannot donate until they have stopped taking it for 3 months, and people living with HIV are not allowed to donate”. All items were scored between 1 and 5 on a Likert scale with the end points anchored (Strongly disagree–Strongly agree).

Three attitudinal constructs were derived based on the component loadings from a principal components analysis (PCA). Although this analysis was informed by our pre‐registered factor structure, the approach did not yield an interpretable solution. In the absence of an established structure for these newly developed items, we adopted a more data‐driven approach (PCA) to maximise construct clarity. This resulted in three interpretable constructs: a positive attitudinal construct regarding policy and services, negative attitudes regarding the deferral policy, and a construct reflecting the extent to which individuals view blood donation as a means of participating in, or belonging to, broader society (i.e., civic belonging). All item wording and factor loadings are shown in Table 1.

TABLE 1.

Rotated component matrix factor loadings and reliabilities for attitudinal measures.

	Factor loadings
	Component 1	Component 2	Component 3
Item 1. It is unrealistic having to abstain from oral/anal sex for 3 months to donate blood	0.858
Item 2. The current blood donor deferral policy is discriminatory	0.692
Item 3. The current deferral policy is unfair towards MSM	0.780
Item 4. I trust the NZ Blood Service to weigh the interests of gay and bisexual men and blood recipients fairly		0.882
Item 5. I support the current deferral policy for MSM in New Zealand	−0.543	0.575
Item 6. The current policy fairly balances gay and bisexual men's interests in donating blood with our increased risk of HIV		0.627
Item 7. Being excluded from blood donation makes me feel less a part of broader society a			0.746
Item 8. Being able to donate blood would enable me to contribute to society			0.837
Item 9. Being excluded from donating blood prevents me from helping others a			0.800
Cronbach's alpha (α)	0.857	0.725	0.790

Note: Extraction method: Principal Component Analysis. Rotation method: Varimax with Kaiser normalisation. Rotations converged in 5 iterations. Showing factor loadings higher than 0.500.

^a These items are reverse scored.

Other than Item 5, factor loadings were specific to one of the three constructs (ranging between 0.627 and 0.882). Item 5 (“I support the current deferral policy for MSM in New Zealand”) cross‐loaded onto the negative attitude construct (Factor Loading = −0.543) and positive attitude construct (Factor Loading = 0.575) and, given face and content validity considerations together with a stronger loading, was loaded on the positive component. Items were aggregated using the average score of three items so that higher scores represent a stronger attitude. Reliability coefficients all exceeded the recommended minimum Cronbach's alpha of 0.60 for small item components. ⁵⁶ The correlations among the three underlying attitudinal components are shown in Table 2. As would be expected, there were moderate relationships among the constructs (ps <0.001) ranging from 0.459 (civic belongingness–positive attitudes) to 0.634 (civic belongingness–negative attitudes).

TABLE 2.

Descriptive statistics and Pearson's correlations between the three attitudinal components.

	1	2	3
1. Negative attitudes towards the MSM deferral policy	–	−0.573	0.634
2. Favourable attitudes towards the blood service		–	0.459
3. Attitudes towards blood donation as a means of civic belonging			–
Mean	4.13	2.40	3.98
Standard deviation	1.09	1.04	1.05

Note: All ps <0.001 (2‐tailed).

Negative attitudinal appraisal of the current MSM deferral policy was measured using three items (e.g., “The current deferral policy is unfair towards MSM”; α = 0.857). Positive attitudes towards the blood service in relation to MSM were measured using three items (e.g., “I trust the NZ Blood Service to weigh the interests of gay and bisexual men and blood recipients fairly”; α = 0.725). The general extent to which individuals viewed blood donation as a means of participating in, or belong to, broader society, and not specific to the MSM experience nor the blood service like the previous two constructs (e.g., “Being able to donate blood would enable me to contribute to society”; α = 0.790).

2.3.4. Previous blood donation behaviour

Blood donation behaviour was assessed via self‐report by two questions: “Have you ever donated blood?” (yes/no) and past donors were asked the year of when they last donated. Participants were classified into two behavioural groups. Firstly, as a function of donation behaviour: donors (reported donating blood at least once; N = 1327) and nondonors (N = 1830). Secondly, as a function of donation recency: recent donors (those who had donated blood in the last two years in 2021 and 2022; N = 136) and people who had not recently donated recently (including those who had never donated blood and those who had donated blood previously but more than two years ago; N = 3021). The shift to a 3‐month deferral period in NZ was implemented in December 2020, thus also making donation recency a behavioural indicator of MSM who donated under the current eligibility rules. This behavioural classification is based on self‐reported donation history which prior studies have shown to be a valid and reliable proxy for actual donor behaviour. ⁵⁷ , ⁵⁸ It was also previously used to classify MSM in the same sample. ⁵⁹

2.4. Analysis

Data were analysed using IBM SPSS Statistics 27. We used a series of one‐way analyses of variance tests and t‐tests to investigate significant differences in participants' responses to the three attitudinal constructs as a function of their demographics, MSM social identity, past donation experience, and recency of their donation behaviour. Type I error rate of 5% was used for all analyses. Due to the skew present in the attitudinal measures, all analyses were conducted on log transformed variables. Tables report the raw mean scores, but F‐ratios, t‐statistics, and p‐values are from the analyses on the log transformed variables. Due to the large number of variables in the analyses, casewise deletions were used for missing values.

Next, two multivariate logistic regressions were conducted including all demographic, MSM‐specific, and attitudinal variables to explore their unique associations with past and recent donation behaviour. All predictors were force‐entered to allow for an exploratory assessment of each variable's unique contribution while accounting for shared variance among the predictor variables. This approach was intended to descriptively examine patterns within the sample, rather than to construct a definitive explanatory model of MSM donor behaviour.

3. RESULTS

3.1. Variation in Attitudes as a Function of the Sample's Background Characteristics

Table 3 provides a summary of mean differences between ratings of each of the three attitudinal constructs as a function of demographic variables. MSM aged under 45, students, those living in urbanised regions, and identifying as NZ European or Māori, on average, reported more negative attitudes towards the deferral policy (ps <0.003). Interestingly, however, these same demographic variables also predicted placing greater civic importance on donating blood (ps <0.011). By contrast, participants aged 45 and over, identifying as Asian, without a university degree, in employment other than full‐time, and being a student (ps <0.005) reported more positive attitudes towards blood services. There were no significant differences in attitude ratings of those who reported having a long‐term disability.

TABLE 3.

Variation in attitudes as a function of sample's background characteristics.

	Attitudes
	Component 1: Negative attitudes towards the policy					Component 2: Favourability of policy and service					Component 3: Attitudes towards blood donation as a means of civic belonging
	N	Mean	SD	F/t‐stat	p‐value	N	Mean	SD	F/t‐stat	p‐value	N	Mean	SD	F/t‐stat	p‐value
Age
<30	1228	4.28	0.95	65.56	<0.001	1227	2.24	0.92	68.07	<0.001	1224	4.06	0.97	25.59	<0.001
31–44	1002	4.26	1.04			1001	2.31	1.00			997	4.06	1.02
45<	602	3.71	1.29			603	2.78	1.13			599	3.71	1.15
Ethnicity
NZ European	2179	4.17	1.07	4.80	0.002	2179	2.36	1.03	7.27	<0.001	2148	3.99	1.04	5.05	0.002
Māori	392	4.19	1.04			393	2.35	1.03			383	4.09	0.98
Asian	275	3.92	1.15			274	2.68	1.03			270	3.81	1.12
Other ethnicities	171	3.98	1.24			170	2.50	1.13			166	3.84	1.16
Region
Urbanised regions	2416	4.19	1.05	21.60	<0.001	2415	2.35	1.01	9.46	0.002	2404	4.01	1.02	7.16	0.008
Non‐urbanised regions	412	3.94	1.23			411	2.51	1.08			412	3.88	1.14
Education
No University degree	1460	4.09	1.10	1.55	0.212	1460	2.47	1.07	5.48	0.004	1415	3.99	1.07	0.03	0.973
University degree	941	4.14	1.08			940	2.35	1.00			937	3.97	1.03
Postgraduate degree	640	4.21	1.08			639	2.33	1.04			639	3.96	1.04
Employment
Employed full‐time	1982	4.19	1.07	10.87	<0.001	1978	2.35	1.01	14.78	<0.001	1973	4.01	1.03	3.81	0.010
Employed part‐time	239	4.03	1.18			239	2.60	1.07			238	3.88	1.05
Student	374	4.26	0.94			375	2.21	0.91			373	4.04	0.96
Other	234	3.81	1.27			235	2.65	1.17			233	3.81	1.16
Long‐term disability
No	2390	4.15	1.07	0.55	0.458	2389	2.40	1.00	2.99	0.084	2381	3.98	1.03	0.00	0.960
Yes	427	4.13	1.15			427	2.25	1.10			425	4.01	1.08

Note: Bold denotes significance level p < 0.05. Mean and SD (standard deviation) are of the raw mean scores. F‐ratios, t‐statistics and p‐values are of the analyses on the log transformed variables. Mean values are an aggregate of items scored between 1 (Strongly disagree) and 5 (Strongly agree) on a Likert scale with the end points anchored.

3.2. Variation in Attitudes as a Function of MSM‐Specific Variables

Table 4 reports the mean differences in attitudinal ratings across variables that are specific to the MSM population. Participants reporting spending a lot of time with other gay, bisexual, takatāpui, or other MSM, and those who were out about their sexuality to most people (i.e., individuals more affiliated with the gay community) reported consistently more negative attitudes towards policy and placed more civic importance on the ability to donate blood (ps <0.001). In contrast, those who reported lower affiliation with the gay community rated the blood services more positively (ps <0.001). Participants who reported a sexual identity other than gay or bisexual reported more negative attitudes towards policy, the lowest positive attitude towards policy and services, and placed the highest importance on blood donation as a civic act (ps <0.001). Lastly, participants taking PrEP reported more negative attitudes towards the policy compared to those who are not on PrEP (p < 0.001), but their ratings did not differ across the other attitudes.

TABLE 4.

Variation in attitudes as a function of MSM‐specific variables.

	Attitudes
	Component 1: Negative attitudes towards the policy					Component 2: Favourability of policy and service					Component 3: Attitudes towards blood donation as a means of civic belonging
	N	Mean	SD	F/t‐stat	p‐value	N	Mean	SD	F/t‐stat	p‐value	N	Mean	SD	F/t‐stat	p‐value
PrEP Use (last 3 months)
No	2153	4.10	1.11	12.49	<0.001	2152	2.41	1.04	2.46	0.117	2116	3.99	1.03	1.23	0.268
Yes	748	4.27	1.01			747	2.34	1.02			739	4.05	1.01
Community attachment
Not a lot of time spent with gay, bisexual, takatāpui, or other MSM people	943	3.92	1.18	44.61	<0.001	942	2.58	1.08	40.87	<0.001	923	3.80	1.12	34.79	<0.001
A lot of time spent with gay, bisexual, takatāpui, or other MSM people	2081	4.23	1.03			2080	2.31	1.01			2053	4.06	1.00
Outness
Not out to most people	1210	3.93	1.16	71.96	<0.001	1206	2.60	1.05	90.57	<0.001	1202	3.78	1.11	74.84	<0.001
Out to most people	1657	4.31	1.00			1659	2.22	0.97			1653	4.13	0.97
Sexual identity
Gay	1884	4.12	1.08	20.46	<0.001	1885	2.43	1.03	21.69	<0.001	1853	3.97	1.05	8.69	<0.001
Bisexual	546	3.94	1.20			544	2.54	1.14			534	3.86	1.10
Other identities	561	4.37	0.93			560	2.15	0.94			554	4.14	0.99

Note: Bold denotes significance level p < 0.05. Mean and SD (standard deviation) are of the raw mean scores. F‐ratios, t‐statistics and p‐values are of the analyses on the log transformed variables. Mean values are an aggregate of items scored between 1 (Strongly disagree) and 5 (Strongly agree) on a Likert scale with the end points anchored.

3.3. Predicting Past Donation Behaviour with Attitudinal Constructs

Table 5 reports the mean differences across the three attitudinal constructs as a function of past blood donation behaviour (donor–nondonor) and recency in donation behaviour (donated in the last two years–did not donate in the last two years). Participants who had ever donated blood did not differ from nondonors in their negative policy attitudes (p = 0.122) or in positive attitudes towards policy and services (p = 0.329). Negative attitudes also did not predict how recently a participant had donated blood (p = 0.080). However, participants without a recent donation reported slightly more positive attitudes towards policy and services relative to more recent donors (p = 0.047). Lastly, both previous donors and recent donors reported viewing blood donation as more important to civic belonging (ps <0.006).

TABLE 5.

Predicting past donation behaviour with attitudinal constructs.

	Attitudes
	Component 1: Negative attitudes towards the policy					Component 2: Favourability of policy and service					Component 3: Attitudes towards blood donation as a means of civic belonging
	N	Mean	SD	t‐stat	p‐value	N	Mean	SD	t‐stat	p‐value	N	Mean	SD	t‐stat	p‐value
Ever donated blood
Never donated	1736	4.17	1.07	2.16	0.142	1735	2.37	1.04	1.67	0.197	1703	3.92	1.08	15.34	<0.001
Donated at least once	1267	4.11	1.10			1266	2.43	1.04			1252	4.07	0.99
Recency in donation behaviour
Had not donated in the last two years	2872	4.14	1.09	0.95	0.329	2870	2.40	1.04	3.94	0.047	2825	3.97	1.05	7.96	0.005
Donated in the last two years	131	4.20	0.95			131	2.24	0.91			130	4.23	0.84

Note: Bold denotes significance level p < 0.05. Mean and SD (standard deviation) are of the raw mean scores. T‐statistics and p‐values are of the analyses on the log transformed variables. Mean values are an aggregate of items scored between 1 (Strongly disagree) and 5 (Strongly agree) on a Likert scale with the end points anchored.

Table 6 presents the multivariate logistic regression models predicting past and recent donation behaviour. The model predicting past donation was significant (χ2 = 144.07, −2 Log Likelihood = 3329.39, Nagelkerke R ²  = 0.099, p < 0.001). Greater civic belongingness was associated with higher odds of having ever donated (OR 1.46, 95% CI 1.30–1.63, p < 0.001), while holding stronger negative attitudes towards the 3‐month deferral policy predicted lower odds of donation (OR 0.82, 95% CI 0.73–0.92, p < 0.001). Belonging to a younger age group and not holding a university degree were associated with lower odds of ever having donated. In contrast, favourable attitudes towards the overall services were not associated with past donation (OR 1.08, 95% CI 0.97–1.19, p = 0.164).

TABLE 6.

Multivariate logistic regression models predicting donation behaviour by recency and ever donor status.

	Model 1: Ever donated blood (Donated at least once versus Never donated)			Model 2: Recency in donation behaviour (Donated in the last 2 years versus Did not donated in the last 2 years)
		95% CI			95% CI
	OR	Lower	Upper	OR	Lower	Upper
Background Characteristics
Age
<30	0.69***	0.57	0.84	1.47	0.89	2.40
31–44	Ref.			Ref.
45<	1.66***	1.30	2.10	0.59	0.26	1.36
Ethnicity
NZ European	Ref.			ref.
Māori	1.00	0.78	1.28	0.58	0.30	1.14
Asian	0.93	0.69	1.25	0.80	0.40	1.60
Other ethnicities	1.17	0.81	1.69	1.11	0.50	2.45
Region
Urbanised regions	Ref.			Ref.
Non‐urbanised regions	1.04	0.87	1.23	1.20	0.79	1.86
Education
No University degree	0.69***	0.57	0.83	0.89	0.57	1.41
University degree	Ref.			Ref.
Postgraduate degree	1.06	0.85	1.32	0.81	0.45	1.44
Employment
Employed full‐time	Ref.			ref.
Employed part‐time	0.89	0.65	1.22	0.56	0.20	1.60
Student	1.18	0.91	1.54	1.93**	1.19	3.14
Other	0.77	0.54	1.09	1.32	0.55	3.19
Long‐term disability
No	Ref.			Ref.
Yes	0.86	0.67	1.10	0.63	0.34	1.17
MSM‐Specific Characteristics
PrEP Use (last 3 months)
No	Ref.			Ref.
Yes	1.08	0.89	1.30	0.33**	0.16	0.67
Community attachment
Not a lot of time spent with LGBT+ people	Ref.			Ref.
A lot of time spent with LGBT+ people	0.99	0.82	1.18	0.49**	0.31	0.76
Outness
Not out to most people	Ref.			Ref.
Out to most people	0.91	0.76	1.09	0.57**	0.38	0.86
Sexual identity
Gay	Ref.			Ref.
Bisexual	1.18	0.94	1.49	2.57***	1.62	4.08
Other identities	0.88	0.70	1.11	1.61	0.93	2.78
Attitudes
Negative attitudes towards the MSM deferral policy	0.82***	0.73	0.92	0.84	0.64	1.10
Favourable attitudes towards the blood service	1.08	0.97	1.19	0.82	0.63	1.07
Attitudes towards blood donation as a means of civic belonging	1.46***	1.30	1.63	1.40*	1.06	1.85

Note: *p < 0.05; **p < 0.01; ***p < 0.001. Model 1 is reporting the odds associated with having ever donated blood at least once (N = 1327) against nondonors (N = 1830). Model 2 is reporting the odds associated with having donated blood more recently in 2022 and 2021 (N = 136) against those who did not donate during this time (N = 3021).

The model predicting recency of donation (i.e., having donated blood under the 3‐month deferral policy) was also significant (χ2 = 119.31, −2 Log Likelihood = 811.97, Nagelkerke R ²  = 0.149, p < 0.001). A greater belief in blood donation as a means to belong to society predicted more recent donation (OR 1.40, 95% CI 1.06–1.85, p = 0.017), while policy‐ (OR 0.84, 95% CI 0.64–1.10, p = 0.205) and service‐related attitudes were not related (OR 0.82, 95% CI 0.63–1.07, p = 0.146). In addition, identifying as bisexual and being a student were associated with higher odds of recent donation, while having greater MSM social connections and being on PrEP were associated with lower odds of having donated recently.

4. DISCUSSION

As blood donation policies globally become increasingly inclusive, identifying how attitudinal characteristics vary across different segments MSM becomes increasingly important. In the absence of an existing measure and using a preregistered psychometric strategy, we developed face valid indices of attitudes towards blood donation for MSM: a negative attitude towards deferral policy, a positive attitude towards blood services, and an attitude regarding blood donation as a form of civic belonging. In terms of associations with background factors, univariate analyses showed that MSM with donation experience and those who had recently donated rated blood donation behaviour as more important to civic belonging. However, recent donors did not differ from other donor groups in their negative views regarding policy, and blood services were rated as slightly less positive among men who had donated recently. In multivariate models, the perception of blood donation as a meaningful form of civic belonging emerged as the most consistent attitudinal predictor of past donation behaviour. Below, the implications of these findings are discussed in line with the key contributions to the study: characterising attitudes within a large MSM sample, investigating how these attitudes were associated with donation behaviour, and broadening the attitudinal constructs that may be relevant to MSM's donation behaviour.

In relation to the first contribution, we established that attitudinal ratings were consistently patterned across some MSM demographic groups. Prior research has highlighted the prevalence of negative donation attitudes among MSM, ⁵ , ⁶ including within a subset of MSM who have subsequently tested positive for infections post‐donation, ²⁵ but it has been unclear whether these blood donation attitudes varied within MSM sociodemographic subgroups. Our analyses showed that younger MSM, those connected to the wider gay community, urban residents, and individuals of NZ European or Māori backgrounds were more likely to view blood donation as an important expression of civic belonging, but concurrently held more negative views towards deferral policies. Perhaps surprisingly, positive attitudes towards blood services were lowest among more educated individuals (despite non‐significant differences in their attitudes towards deferral policies and civic belonging), perhaps indicating that educational attainment might independently predict perceptions of service quality. While it is not within the scope of this report to speculate as to why this might be, variations in attitudes across demographic groups challenge the assumption that all MSM share homogeneous views regarding blood donation. Rather, they begin to show how different demographic factors can intersect to predict MSM's attitudes towards blood donation and, potentially, their behaviour. In turn, this pattern implies that a one‐size‐fits‐all approach regarding blood donation behaviour among MSM communities may not be appropriate given clear attitudinal diversity, especially as some attitudes appear more important to the prediction of donation behaviour.

A second contribution and area of focus for the current study lies in using a more robust index of attitudes to predict donation behaviour. As noted, prior studies have frequently studied intentions and have typically found that positive attitudes towards blood donation organisations are associated with greater intent to donate. ²³ , ²⁴ In the current study, however, our data showed more positive views predicted less donation in the last two years at the univariate level, while negative attitudes only predicted having ever donated in a multivariate model. This may suggest that the attitudinal predictors of intention are different from those predicting donor behaviour, that there might be differences across samples, or that the particular attitudinal variables we operationalised are responsible. While the attitudinal constructs used in this study were grounded in a pre‐registered strategy, the final structure diverged from initial expectations based on exploratory analyses. As these items were newly developed, we were unable to formally assess convergent or divergent validity due to the absence of external validation measures in the parent study. However, the three groupings demonstrated strong face validity, moderate intercorrelations, and meaningful associations with past donation behaviour in multivariate models. With this, we believe that the predictive utility supports their use as a reasonable foundation for future refinement and validation.

In comparison to the scattered links positive and negative attitudes had with donor behaviour, our analyses suggest that viewing blood donation through the lens of civic belongingness is the only consistent attitudinal predictor of donation behaviour in our sample. Such a pattern is broadly consistent with findings in general population samples showing that many donors are motivated to give blood as a means of benefiting the broader community and contributing to society. ¹⁷ , ²¹ , ³² , ⁵⁰ However, while qualitative research has previously identified this motivation among prospective MSM donors, ⁵⁴ our findings extend this pattern to quantitative data as well as demonstrating that a broad sense of civic belonging can differentiate past donors from nondonors.

These findings have important implications as blood services begin to consider managing newly eligible MSM donors. Firstly, blood services should maximise engagement by using strategies that highlight blood donation as a meaningful act of societal participation and inclusion. Having a unified or consistent message in blood donation campaigns may promote a sense of fairness and inclusivity among all donors, including MSM, as well as potentially reducing the need for specialised materials and targeted recruitment strategies aimed at different populations. Our results thus indicate that the same messages would resonate with both MSM and non‐MSM groups. Second, negative attitudes may pose a barrier to future recruitment of those who had never donated blood before. While targeting individuals who value donating blood as a part of civic membership could optimise engagement efforts, these individuals also hold strong negative attitudes towards the current 3‐month deferral policy. As such, specific tailoring in messaging might need to be made in communications with prospective MSM donors.

One way to address negative attitudes could be to highlight the fairness and inclusivity of the new policy for MSM, which may help to address barriers to donation. Given that our sample (donation history notwithstanding) held negative attitudes towards the time‐based deferral policy, emphasising the differences of the new individualised approaches could be beneficial to donor engagement. Policy communications should clearly state that, unlike blanket time‐based deferral rules, the new individual risk assessment approach allows MSM in long‐term monogamous relationships to donate. ² , ³ , ⁴ Acknowledging the improvement in risk management (e.g., these individuals pose little risk of having recently acquired undiagnosed HIV) should be clearly communicated to MSM. Testament to the importance of policy communication, emerging research is already exploring how to effectively communicate policy changes related to individualised risk assessments. ⁶⁰ Emphasising such points in future recruitment messages could be crucial for engaging MSM, especially those who may have concerns that discourage their participation in blood donation.

Considering the final contribution of this paper, our findings highlight the need for greater specificity in the measurement of blood donation‐related attitudes. Aggregate scores of ostensibly related attitudinal constructs should be avoided, given that positive attitudes towards various aspects of blood donation were not consistently associated with increased behavioural engagement. Put another way, our data tend to indicate that broad attitudes in which the “content” is mixed or unclear may be of less use in predicting blood donation. So, for example, it seems unlikely that the importance of civic belongingness would have emerged if we had used general positive and negative attitude constructs. In turn, this measurement specificity implies that blood services should prioritise reinforcing the societal meaning of donation over improving general perceptions of the blood service. Treating these attitudes as independent constructs will be important to identifying those relevant to donation behaviour, which will improve recruitment messaging, interventions, and communication strategies.

4.1. Recommendations for future research

Future research on attitudes towards blood donation should prioritise several key study design directions. As with prior cross‐sectional work, causality remains unclear; we are unable to determine whether specific attitudes should be targeted to change behaviour. Improvements in understanding appear likely with the development and implementation of experimental designs (e.g., targeted messaging) that incorporate behavioural proxies (e.g., using sham blood donation appointments) or follow a cohort over time to track actual behaviour. Such studies are needed to ameliorate the issues associated with reliance on intentions as a proxy in blood donation contexts, establish causality, and build on the understanding of the relationship between blood donation attitudes and actual behaviour.

4.2. Strengths and limitations

A strength of our analysis is the use of NZ's HIV behavioural surveillance survey, encompassing a large and diverse group of NZ MSM in terms of age, ethnicity, residence, and sexual identification. This increases the likelihood that the findings are relevant and applicable to a wide range of MSM. One key limitation to consider is that we relied on self‐reported donation behaviour without linkage to administrative records. Blood donation is generally viewed more positively both in the general population, ²¹ , ³² and among MSM, ⁵⁴ and may therefore be subject to socially desirable responding or overreporting. However, prior research has shown high concordance between self‐reported and donation behaviour, with one study reporting 98% accuracy in recall among verified donors. ⁵⁸ Combined with the anonymous nature of the survey, this likely mitigated some risk of misreporting, and thus supporting the validity of self‐report as a behavioural proxy in this context.

5. CONCLUSION

In light of the increasing liberalisation of blood donation policies for MSM, blood services need to understand the attitudes and behaviours of this emerging donor base. While negative views towards deferral policies and blood services were present and varied within our sample, they were not as consistent in predicting reports of past or recent donation as was a perception of blood donation as a meaningful expression of civic belonging. Further investigation is needed to understand the nuanced effects of attitudes towards policies and services on donation behaviour in the MSM context. Given the complexity and the need for specificity in measuring attitudes, as well as the limitations of using intention proxies, more experimental or longitudinal studies are necessary to establish causal relationships. Finally, as policies evolve, rigorous research practices will be necessary to provide specific attitudinal targets for blood services' recruitment and retention strategies in the new context of eligible MSM donors.

AUTHOR CONTRIBUTIONS

Koson Tony Sriamporn and Nathan S. Consedine conceptualised the paper and led the development of the attitudinal items. Peter J. W. Saxton designed the Sex and Prevention of Transmission Study behavioural surveillance program and collected the data. Koson Tony Sriamporn conducted the literature review, analysed the data, and drafted the initial manuscript. All authors contributed to the re‐drafting and evaluation of the final manuscript for publication.

FUNDING INFORMATION

This work was funded by the Health Research Council of New Zealand (Project reference: 20/887).

CONFLICT OF INTEREST STATEMENT

The authors have no conflicts of interest to disclose.

PARTICIPANT CONSENT

Written informed consent was obtained from each participant for study participation and data publication.

Sriamporn KT, Saxton PJW, Consedine NS. It is not all anti‐policy attitudes: Viewing blood donation as a means to belong predicts blood donation among men who have sex with men. Transfusion Medicine. 2025;35(5):424‐436. doi: 10.1111/tme.70012

DATA AVAILABILITY STATEMENT

The data that support this study cannot be publicly shared due to ethical or privacy reasons and may be shared upon reasonable request to the corresponding author if appropriate.