Bridging the attitude–behaviour gap in blood donation: A large‐scale knowledge–attitudes–practices study among university students in China

Abstract

Background and Objectives

University students are a vital yet underutilized donor population, constrained by a knowledge–attitude–practice (KAP) gap. We sought to assess the KAP status and identify modifiable determinants of blood donation among university students in a major Chinese city to inform targeted recruitment strategies.

Materials and Methods

A cross‐sectional online survey was conducted from September to November 2023, involving students from all 17 universities in Wuxi, China. The questionnaire covered socio‐demographic data, blood donation knowledge, attitudes and influencing factors. Data were analysed using descriptive statistics, chi‐square tests and multivariable logistic regression.

Results

Among 8537 valid respondents, 917 (10.7%) reported a history of blood donation. Attitudes were generally positive, but knowledge gaps were evident, particularly regarding operational details. Despite this positive outlook, a pronounced attitude–behaviour gap was observed: 59.8% expressed willingness to donate, yet only 10.7% had actually done so. Multivariable analysis identified political/league membership, student leadership roles and club participation as significant positive predictors of donation behaviour. Notably, while experiential factors (sense of glory/honour, donation environment, staff competence) were significant predictors, perceived physical eligibility was not, implying that for non‐donors, this commonly cited barrier may be overridden by positive experiential factors.

Conclusion

A significant gap between willingness to donate (59.8%) and reported donation behaviour (10.7%) exists among students, shaped by distinct factors for donors and non‐donors. Bridging this gap requires a strategic shift from general awareness campaigns to tailored, experiential interventions that address specific cognitive and experiential barriers and reinforce positive donor experiences.

Highlights.

• This large‐scale study (n = 8537) in China quantifies a significant gap between willingness to donate (59.8%) and reported donation behaviour (10.7%) among university students.

• Social engagement (political/league membership, leadership roles, club activity) is a strong independent predictor of blood donation behaviour. Drivers for donors are experiential (environment, staff competence, sense of glory/honour), whereas the key barrier for non‐donors is perceived physical eligibility concerns.

• Evidence‐based, differentiated strategies are proposed to convert intention into action, focusing on targeted education, service optimization and peer‐network partnerships.

INTRODUCTION

Voluntary non‐remunerated blood donation (VNRBD) is universally recognized as the cornerstone of a safe and sufficient blood supply. Despite annual increases in national blood collection volumes in China, the blood service system continues to face persistent challenges, including seasonal shortages, regional imbalances and an over‐reliance on fixed, repeat donors from specific demographics [1]. Within this context, university students emerge as a strategically vital donor population. They are typically in good health, have higher education levels (which may correlate with better health literacy and altruistic values) and represent a cohort with the potential to be cultivated into long‐term regular donors, thus ensuring sustainability [2]. However, the potential of this demographic is frequently unrealized due to a well‐documented global ‘attitude–behaviour gap’ [3, 4]. Our study quantified a substantial local instance: 59.8% of respondents expressed willingness to donate, yet only 10.7% reported having done so. This finding aligns with the prevalent ‘attitude–behaviour gap’ noted among Chinese university students, although the lower donation rate observed here may indicate regional variations or local challenges in converting intent to action. Elucidating the complex factors that facilitate or hinder this intention–behaviour translation is therefore crucial.

Previous research has identified a multifactorial web influencing student donation behaviour. Key elements include the level of accurate knowledge about donation procedures and health impacts, perceived barriers (such as fear of needles, concerns about weakness or time constraints), the powerful influence of social norms and peer behaviour and the practical convenience of donation services [5, 6]. While informative, many existing studies are limited by relatively small sample sizes, a focus on single universities or a lack of comprehensive comparative analysis that delineates the distinct psychological and practical landscapes of donors versus non‐donors within the same population. To address these gaps, we conducted a large‐scale, city‐wide investigation in Wuxi, China. This study aimed to assess the knowledge–attitudes–practices (KAP) status, identify predictors of donation (distinguishing between donor motivators and non‐donor barriers) and propose evidence‐based, tailored recruitment strategies for this population. The findings aim to inform local recruitment and contribute to the global understanding of youth donor mobilization.

MATERIALS AND METHODS

Study design and participants

A descriptive, cross‐sectional survey was conducted over a 3‐month period from September to November 2023. The target population was students enrolled at all 17 universities and colleges within Wuxi City, Jiangsu Province, China. This ensured representation across a wide spectrum of academic disciplines (e.g., engineering, sciences, medicine, humanities) and academic years. A non‐probability convenience sampling method was employed for practical reach. The online questionnaire was disseminated through multiple channels to maximize participation, including official university student union networks, class social media groups (e.g., WeChat, Tencent QQ) and collaborations with faculty advisors. A total of 8547 initial responses were collected via the ‘Questionnaire Star’ platform (https://www.wjx.cn). After a rigorous validation process, 10 responses were excluded because of apparent patterns of random answering (e.g., straight‐lining), logical inconsistencies or being substantially incomplete. Consequently, 8537 valid questionnaires were retained for final analysis, representing a questionnaire validity rate of 99.88%.

Questionnaire development and data collection tool

The research instrument was a structured, self‐administered questionnaire developed through a multi‐stage process. First, an extensive review of national and international literature on blood donor motivation and KAP studies informed the initial item pool. Second, draft questionnaires were reviewed and refined by a panel of experts from the Wuxi Blood Center, including transfusion medicine specialists and donor recruitment professionals, to ensure content validity, relevance and cultural appropriateness. A pilot test was conducted with a small group of 50 students—not included in the main study—to assess clarity, comprehension and time required for completion; minor adjustments to wording were made based on feedback. The final questionnaire comprised four distinct sections.

1. Section A: Socio‐demographic characteristics. This section collected 10 items, including gender, age, academic major, academic year, political/league membership (categorized as member of the Communist Party of China [CPC], Communist Youth League [CYL] or non‐affiliated students), whether the respondent held any student leadership position (e.g., class monitor, student union officer), frequency of participation in campus club/society activities (‘frequent’, ‘occasional’, ‘never’), hometown location (urban/rural), family structure (only‐child vs. multi‐child) and, crucially, self‐reported history of blood donation (yes/no).

2. Section B: Blood donation knowledge assessment. This section contained nine items designed to gauge objective understanding. Topics included the health impact of donation on the donor, minimum donation intervals, basic human physiology (e.g., total blood volume), differences between whole‐blood and apheresis donation, eligibility criteria (age, weight) and pre‐ and post‐donation testing procedures. Responses were captured on a symmetric 5‐point Likert scale ranging from ‘very unknowledgeable’ (1) to ‘very knowledgeable’ (5).

3. Section C: Blood donation attitude assessment. This section contained six items measuring subjective perceptions and intentions. Items assessed beliefs about the health benefits of donation, its social value and role in community spirit, willingness to donate in the future, willingness to donate again (for donors) or encourage others to donate and overall willingness to participate in public welfare activities. A 5‐point Likert scale from ‘strongly disagree’ (1) to ‘strongly agree’ (5) was used.

4. Section D: Influencing factors and preferences. This section investigated the decision‐making calculus. It included nine items on factors potentially influencing the donation decision (e.g., ‘My perceived physical eligibility (i.e., concern about being eligible) determines whether I donate’, ‘The donation environment determines whether I donate’, ‘Sense of glory/honour after donation influences me’, ‘Staff competence influences me’, ‘Influence of online news/public opinion’). These were rated on a 5‐point agreement scale. Two related but distinct constructs were measured: ‘self physical condition’ assessed familiarity with eligibility criteria (knowledge‐related), while ‘perceived physical eligibility (concern)’ captured concerns about meeting health requirements (attitudinal barrier). Additionally, multiple‐choice questions assessed preferred channels for receiving donation information (e.g., social media, campus posters, lectures), types of incentives considered appealing (e.g., practical gifts, honorary certificates, small cash subsidies for transportation) and awareness of specific national and local blood donation policies.

For analysis, responses to the nine knowledge items and six attitude items on the 5‐point Likert scales were summed to create total knowledge and attitude scores, respectively. Higher scores indicated greater knowledge or more positive attitudes. For specific comparative analyses (e.g., chi‐square tests), the total knowledge and attitude scores were dichotomized at the sample median into ‘high/adequate’ and ‘low/inadequate’ groups, whereas primary analyses used the continuous scores. The internal consistency of the multi‐item scales was acceptable, with Cronbach's alpha values of 0.78 (knowledge) and 0.81 (attitude). Given the exploratory aim of this study and to facilitate clear comparisons between ‘higher’ and ‘lower’ knowledge/attitude groups within our specific student cohort, the sample median was used for dichotomization in the absence of a universally accepted cut‐off.

Sample size calculation

The minimum sample size was calculated a priori using the standard formula for estimating a single population proportion, with a margin of error of 5% and an assumed proportion of 0.5. This yielded a base sample of 384. A design effect of 1.5 was applied for the multi‐university design, and a 10% non‐response rate was anticipated. The final minimum required sample size was thus calculated as (384 × 1.5)/0.9 = 640. Our achieved sample of 8537 comfortably exceeded this threshold, providing adequate power for the planned analyses.

Data analysis

All statistical analyses were performed using SPSS Statistics version 22.0. Descriptive statistics were used to summarize socio‐demographic characteristics and the distribution of responses for KAP items, presented as frequencies and percentages for categorical data. The study population was dichotomized into ‘donors’ (those with a history of blood donation) and ‘non‐donors’ (those without). Bivariate analyses comparing these two groups were conducted using appropriate statistical tests based on variable types. Continuous variables (total knowledge and attitude scores) were compared using the independent‐samples t‐test (or the Mann–Whitney U test if normality assumptions were violated), while categorical variables (socio‐demographics and influencing factors) were compared using the Pearson chi‐square test.

To identify socio‐demographic and behavioural predictors of donation behaviour, variables from Section A (socio‐demographics) that showed an association with donor status at a significance level of p < 0.1 in univariate analyses were entered into a multivariable binary logistic regression model (Model 1, Table 1). This model aimed to examine the gross association of these structural and participatory factors with donation, acknowledging that their effects might be partially mediated by knowledge and attitudes, which are examined in separate models. The model employed the backward stepwise elimination method (Wald statistic) to arrive at a parsimonious final model. The results are presented as regression coefficients (B), standard errors (SEs), Wald statistics (Wals), p‐values and odds ratios (ORs). The model's goodness of fit was assessed using the Hosmer–Lemeshow test. To better understand the specific impact of factors influencing the donation decision, an additional multivariable logistic regression analysis was performed. In this model, selected factors from Section D of the questionnaire (perceived physical eligibility, sense of glory/honour, donation environment and staff competence) were entered as independent variables, with donation status (yes/no) as the dependent variable, controlling for gender and age. This analysis quantified the adjusted association (adjusted odds ratio [aOR]) between each specific factor and the likelihood of being a blood donor. For all final inferences, a two‐tailed p‐value <0.05 was considered statistically significant.

TABLE 1.

Multivariable logistic regression analysis of factors associated with blood donation behaviour.

Variable	B	SE	Wals	p	OR
Constant	0.839	0.216	15.094	<0.001	2.315
Gender	0.576	0.088	42.618	<0.001	1.779
Masses (ref.)			26.165	<0.001
Political/league membership (CPC member)	0.162	0.225	0.517	0.472	1.176
Political/league membership (CYL member)	0.557	0.224	6.186	0.013	1.746
Student leader	0.555	0.084	44.077	<0.001	1.743
Club participation (ref.)			44.288	<0.001
Club participation (frequent)	0.418	0.077	29.749	<0.001	1.520
Club participation (occasional)	0.823	0.156	27.711	<0.001	2.278

Note: The reference category for political/league membership is ‘masses’ (non‐affiliated students). The reference category for club participation is ‘never’. This model examines the gross associations; the observed odds ratios for social engagement factors may encompass both their direct effects and indirect effects mediated through variables like knowledge and attitudes.

Abbreviations: CPC, Communist Party of China; CYL, Communist Youth League; OR, odds ratio.

Ethics statement

The study protocol, including the informed consent procedure and questionnaire, was reviewed and approved by the Institutional Review Board (IRB) of the Wuxi Blood Center prior to data collection. The first page of the online survey presented a detailed participant information sheet outlining the study's purpose, voluntary nature, anonymity, data usage and the right to withdraw. Electronic informed consent was obtained from all participants by requiring them to select an ‘I agree to participate’ button before they could proceed to the questionnaire. No personally identifiable information was collected, ensuring participant anonymity and confidentiality throughout the research process.

RESULTS

This section presents the findings of the study, beginning with the socio‐demographic profile of participants and their donation status, followed by an assessment of their KAP, and concluding with the factors associated with donation behaviour.

Participant characteristics

The socio‐demographic profile of the 8537 respondents is summarized as follows. The sample was predominantly male (72.9%, n = 6221) and overwhelmingly fell within the typical university age range of 18–25 years (94.3%). In terms of political/league affiliation, the majority (76.5%, n = 6533) identified as non‐affiliated students (‘the masses’), while 23.5% (n = 2004) were members of either the CPC or the CYL. A notable proportion of students held formal student leadership roles (21.6%, n = 1848) and participated frequently in extracurricular club activities (32.2%, n = 2751). The most critical finding from this baseline description was the low prevalence of donation experience: only 10.7% (n = 917) of the total sample reported having ever donated blood, establishing a clear donor (n = 917) versus non‐donor (n = 7620) dichotomy for subsequent analysis. Initial bivariate comparisons (chi‐square tests) revealed significant differences (p < 0.001) between donors and non‐donors across several key variables, including gender, political/league membership, leadership role and club participation frequency.

KAP status

The assessment revealed a landscape of mixed knowledge, generally positive attitudes, and a pronounced attitude–behaviour gap. As shown in Figure 1, knowledge levels were uneven, with particular gaps in operational details, while attitudes were generally favourable. A pronounced attitude–behaviour gap was evident, with willingness to donate (59.8%) far exceeding the actual donation rate (10.7%) (Figure 2).

FIGURE 1.

Knowledge and attitudes regarding blood donation among university students in Wuxi, China (N = 8537). (a) Knowledge level on specific topics. This panel depicts the percentage of students demonstrating correct knowledge across nine key domains related to blood donation. The figure illustrates significant knowledge gaps, particularly regarding operational details such as donation sites and specific procedures. (b) Attitude level on specific topics. This panel illustrates student attitudes across dimensions related to blood donation willingness and perception. Attitudes were generally favourable, although willingness for long‐term commitment was notably lower.

FIGURE 2.

The attitude–behaviour gap in blood donation: willingness to donate versus actual donation behaviour. This figure compares the proportion of students who expressed willingness to donate blood with those who had actually donated. A pronounced attitude–behaviour gap was evident, with willingness to donate (59.8%) far exceeding the actual donation rate (10.7%).

Factors independently associated with donation behaviour

The results of a multivariable logistic regression analysis examining the gross (total) associations of socio‐demographic and behavioural factors with donation history are presented in Table 1. This model identified political/league membership (CYL), student leadership roles and frequent club participation as strong independent predictors. A separate multivariable logistic regression analysis of specific influencing factors (controlling for gender and age) showed that ‘sense of glory/honour’, ‘donation environment’ and ‘staff competence’ were significant positive predictors of donation behaviour, whereas ‘perceived physical eligibility’ was not (Table 3). This finding is further explored in section Differential influencing factors: Donors versus non‐donors section. An interesting pattern was observed for club participation. Compared to non‐participants, both ‘occasional’ and ‘frequent’ participation were significant positive predictors. Notably, ‘occasional’ participation showed a stronger association (aOR = 2.28) than ‘frequent’ participation (aOR = 1.52).

TABLE 3.

Multivariable logistic regression analysis of influencing factors on donation decision.

Variable	B	SE	Wals	p	OR
Constant	3.572	0.442	65.391	<0.001	35.587
Gender (male)	−0.401	0.088	20.957	<0.001	0.669
Age	−1.122	0.212	28.065	<0.001	0.325
Self physical condition	0.018	0.086	0.044	0.834	1.018
Sense of glory/honour	0.173	0.055	10.050	0.002	1.189
Donation environment	0.120	0.055	4.701	0.030	1.127
Staff competence	0.186	0.057	10.644	0.001	1.205
Perceived physical eligibility (concern)	0.095	0.089	1.141	0.286	1.099
Influence of online news	−0.109	0.063	3.020	0.082	0.897

Note: The model controls for gender and age. Although ‘perceived physical eligibility (concern)’ was not a significant independent predictor in this multivariable model, its high salience as a self‐reported barrier among non‐donors (Table 2) suggests it may function as a prevalent, pre‐conscious concern that is overridden by positive experiential factors.

Abbreviation: OR, odds ratio.

Regression analysis of knowledge and attitude factors

To further elucidate which specific knowledge and attitude components were independently associated with donation behaviour, a multivariable logistic regression analysis was performed (Table 4). After controlling for gender and age, several factors emerged as significant predictors. Among knowledge items, correct understanding of the health impact of donation (OR = 1.263, p < 0.001), appropriate donation volume (OR = 1.295, p = 0.002), donation interval (OR = 1.946, p < 0.001) and screening procedures (OR = 1.351, p = 0.001) were positively associated with being a donor. Conversely, knowledge of basic physiology (OR = 0.641, p < 0.001) and blood type differences (OR = 0.779, p = 0.001) showed negative associations. Regarding attitudes, the intention to repeat donation (OR = 1.493, p < 0.001) and satisfaction with the donation experience (OR = 1.538, p < 0.001) were strong positive predictors. Negative associations were also observed for perceiving donation as beneficial to health (OR = 0.838, p = 0.005), as well as for family subjective norm (OR = 0.851, p = 0.03) and emergency response intention (OR = 0.854, p = 0.036). The negative associations observed for knowledge of basic physiology and blood type differences, along with these counterintuitive attitude findings, are noteworthy and will be discussed in the following section.

TABLE 4.

Multivariable logistic regression analysis of knowledge and attitude factors regarding blood donation.

Variable	B	SE	Wals	p	OR
Constant	2.948	0.445	43.905	<0.001	19.062
Gender (male)	−0.378	0.09	17.611	<0.001	0.685
Age	−1.040	0.212	24.040	<0.001	0.353
Knowledge
Perceived_Knowledge_Health Impact	0.233	0.06	14.877	<0.001	1.263
Perceived_Knowledge_Site Location	0.098	0.051	3.668	0.055	1.103
Perceived_Knowledge_Physiology	−0.445	0.068	42.773	<0.001	0.641
Perceived_Knowledge_Appropriate Volume	0.258	0.085	9.286	0.002	1.295
Perceived_Knowledge_Age Eligibility	−0.092	0.086	1.143	0.285	0.913
Perceived_Knowledge_Donation Interval	0.666	0.080	68.765	<0.001	1.946
Perceived_Knowledge_Blood Type Difference	−0.249	0.078	10.116	0.001	0.779
Perceived_Knowledge_Screening Procedure	0.301	0.088	11.713	0.001	1.351
Perceived_Knowledge_Local Policy	−0.121	0.072	2.857	0.091	0.886
Attitudes
Perceived_Benefit	−0.177	0.063	7.976	0.005	0.838
Intention_Repetition	0.401	0.074	29.549	<0.001	1.493
Subjective_Norm_Family	−0.161	0.075	4.682	0.030	0.851
Intention_Social Promotion	−0.306	0.078	15.42	<0.001	0.736
Intention_Emergency Response	−0.158	0.075	4.375	0.036	0.854
Attitude_Satisfaction	0.430	0.083	27.142	<0.001	1.538

Note: The model controls for gender and age. All knowledge and attitude items were measured on a 5‐point Likert scale.

Abbreviation: OR, odds ratio.

Differential influencing factors: Donors versus non‐donors

A comparative analysis of the factors influencing the donation decision revealed distinct profiles for donors and non‐donors (Table 2). As shown in Table 2, donors and non‐donors were influenced by distinct factors. Donors were predominantly motivated by experiential and psychosocial factors (sense of glory/honour, donation environment and staff competence), whereas non‐donors most frequently cited ‘perceived physical eligibility (concern)’ as a barrier. Factors such as ‘donation point convenience’ showed comparable levels of concern across both groups, indicating they were universal considerations rather than key differentiators.

TABLE 2.

Key factors influencing donation decision: Donors versus non‐donors (% agree/very agree).

Variable	Donors (n = 917)	Non‐donors (n = 7620)
Self physical condition	68.2	73.1
Sense of glory/honour	72.5	56.8
Donation environment	71.3	60.1
Staff competence	65.4	55.2
Perceived physical eligibility (concern)	65.1	70.1
Influence of online news	40.5	46.0

Note: Values are the percentage of respondents who reported ‘agree’ or ‘strongly agree’ on a 5‐point Likert scale. ‘Perceived physical eligibility (concern)’ refers to the subjective worry about meeting health requirements, distinct from objective knowledge.

DISCUSSION

This large‐scale study confirms the existence of a significant attitude–behaviour gap among university students in China, a phenomenon consistent with global observations [7, 8]. The 10.7% donation rate in our cohort falls within the broad 5%–30% global range, aligning with recent multi‐country findings on the variability of actual rates [9]. A recent 16‐country cross‐sectional study involving 12,606 university students found that while attitudes towards blood donation were generally positive, actual donation rates varied widely depending on local blood service infrastructure, cultural norms and awareness campaigns [9]. In Nepal, a study reported that only 5.5% of bachelor‐level students had good knowledge of blood donation, despite 53.2% holding positive attitudes—a pattern closely mirroring the KAP gap identified in our Chinese cohort [10]. Comparisons with studies from diverse regions further contextualize these findings. For example, research in some Western settings often highlights convenience, fear of needles or lack of information as primary barriers, whereas in certain Middle Eastern or African contexts, cultural or faith‐based beliefs or trust in the blood service system may play a more prominent role. Our study, situated in a Chinese metropolitan context, identifies a distinct combination of cognitive barriers (e.g., concerns about physical eligibility) and strong social engagement predictors (e.g., political/league membership, leadership roles), suggesting that the specific constellation of factors driving the attitude–behaviour gap is culturally and structurally embedded.

Our study reveals a critical paradox for intervention design. While non‐donors most frequently self‐reported ‘perceived physical eligibility (concern)’ as a barrier (70.1%, Table 2), this factor was not an independent predictor in the multivariable model (Table 3). Conversely, positive experiential factors (‘sense of glory/honor’, ‘donation environment’, ‘staff competence’) were significant predictors of actual donors [11, 12]. This suggests that for non‐donors, health concerns may function as a salient, pre‐conscious deterrent that is overridden only by a compelling positive expectation of the donation experience itself. Therefore, strategies targeting non‐donors should move beyond generic health assurances and instead combine myth‐busting with active cultivation of a positive anticipated experience—through peer testimonials, service quality transparency and emphasis on psychosocial rewards—to convert apprehension into action.

The robust association between donation behaviour and social engagement indicators (CPC/CYL membership, leadership, club activity) is noteworthy. As illustrated in Table 1, these social‐structural factors are strong predictors of donation behaviour, whereas Table 3 highlights that experiential factors (sense of glory/honour, donation environment, staff competence) are more salient for existing donors and underscore the importance of identity reinforcement, altruistic validation and positive social image in motivating donation behaviour [13, 14, 15]. Although the model in Table 1 does not adjust for potential mediators such as knowledge and attitudes, the strength of the associations implies that social embeddedness itself—likely fostered through exposure to prosocial norms and peer influence—constitutes a powerful contextual pathway to donation. This can be interpreted through the lens of social identity and normative conduct theories [16]. Students embedded in formal organizations or active social networks are more likely to be exposed to and internalize pro‐social norms. They may perceive blood donation as a recognized expression of social responsibility within their peer group, lowering the psychosocial cost of participation [17]. These individuals serve as natural ‘peer ambassadors’, making partnerships with student unions and clubs a strategically efficient recruitment channel. The stronger association with occasional (vs. frequent) club participation further suggests that initial social integration may be a key catalyst for first‐time donation. The distinct driver‐barrier profiles (Tables 2 and 3) clearly delineate intervention targets: experiential and psychosocial factors (environment, staff competence, sense of glory) for donors, and the salient cognitive barrier of perceived physical eligibility concerns for non‐donors [18].

The negative associations between donation history and knowledge of basic physiology/blood type differences, as well as the belief that donation benefits health, require careful interpretation. A more detailed understanding of blood physiology may inadvertently heighten risk perceptions, while believing in health benefits might reflect an ‘optimistic bias’—acknowledged as a general fact but not internalized as personal motivation among non‐donors. These counterintuitive findings highlight the complexity of health cognition: merely increasing technical knowledge or emphasizing health benefits may be counterproductive if not paired with strategies addressing affective fears and building positive service expectations.

To bridge the identified gap, we propose a dual‐track strategy that leverages the power of peer networks. The recruitment track for non‐donors should employ targeted, myth‐busting communication campaigns—delivered via trusted peers and digital platforms—to convert health‐related anxieties into perceived capability. The retention track for existing donors must focus on service excellence, through staff training, environment optimization and meaningful post‐donation recognition, to foster repeat donation. Crucially, both tracks should be implemented in partnership with student political bodies, unions and clubs. Empowering these organizations to lead group drives and peer‐to‐peer messaging can dramatically enhance the reach and credibility of both recruitment and retention efforts.

The limitations of this study should be acknowledged. First, the cross‐sectional design captures a snapshot in time and precludes the establishment of causal relationships. Second, while the sample is large and from multiple institutions, it is confined to one city (Wuxi), which may limit the generalizability of findings. Third, the use of a non‐probability convenience sampling method, despite its efficiency in reaching a large multi‐campus sample, means the findings may not be fully representative of all university students in Wuxi or broader contexts. Future studies employing random sampling strategies are recommended to confirm the generalizability of these associations. Fourth, reliance on self‐reported data, particularly for donation history, is subject to recall and social desirability bias. This common methodological limitation of cross‐sectional surveys means that our findings on donation prevalence and associated factors should be interpreted with caution regarding potential inaccuracies in self‐reported behaviour. Fifth, the questionnaire assessed stated willingness rather than measuring actual future behaviour. Sixth, the dichotomization of continuous scores using a sample median split, while a practical data‐driven approach in the absence of a gold standard, may have reduced statistical power and limited direct comparability with studies using different thresholds.

Despite these limitations, our findings offer a transferable framework for understanding and addressing the donor recruitment gap among students. First, the pronounced attitude–behaviour gap is a common challenge, underscoring the universal difficulty in converting intention into action. Second, the strong predictive power of social engagement indicators suggests that pro‐social peer networks (e.g., student clubs, unions) are potent leverage points for recruitment across diverse settings, although the relative importance of specific factors (e.g., political affiliation) may vary culturally. Most importantly, the divergent driver‐barrier profile between donors and non‐donors provides a universally applicable strategic lens: optimizing the service experience is crucial for retention, while converting non‐donors requires addressing their primary cognitive barriers (e.g., health concerns) while simultaneously building positive expectations. Implementing this approach elsewhere would first require local diagnostic surveys to identify context‐specific predictors and barriers. Future research should address the limitations of this study. The following directions are prioritized. First, linking survey data with verifiable donation records would help mitigate self‐report bias. Second, longitudinal studies are needed to establish causal pathways. Third, multi‐centre studies across diverse socio‐cultural contexts are essential to test the generalizability of the findings. Fourth, designing and evaluating theory‐informed intervention trials—for example, comparing peer‐led education campaigns against standard promotion—using objectively measured donation outcomes is a crucial next step [19, 20].

In conclusion, this study provides an empirical map of factors influencing blood donation among Chinese university students, highlighting that the challenge is not a lack of goodwill but a multifaceted intention–action gap. Bridging this gap requires differentiated strategies: addressing cognitive barriers for non‐donors while nurturing experiential motivators for existing donors, delivered through strategic partnerships with campus organizations. Unlocking the potential of this donor pool demands an evidence‐based approach that treats students as individuals with varying pathways to becoming lifelong contributors.

CONFLICT OF INTEREST STATEMENT

The authors declare no conflicts of interest.

Zhao L, Jin J, Yao J, Chen L. Bridging the attitude–behaviour gap in blood donation: A large‐scale knowledge–attitudes–practices study among university students in China. Vox Sang. 2026;121:950–958.

DATA AVAILABILITY STATEMENT

The data that support the findings of this study are available from the corresponding author upon reasonable request.