This cross-sectional survey, which originated from a registered randomized control trial, was carried out in Jiangsu province of China during the period from December 2023 to June 2024. The study has been approved by the ethics committee of the first affiliated hospital with Nanjing Medical University (No. 2021-094). The survey primarily relied on a designed questionnaire. A pilot study involving 20 participants was carried out to refine the survey instrument, and a 100% response rate was attained. The questionnaire was further revised, and its reliability and construct validity were verified.

The inclusion criteria for blood donors were: (a) aged 18–60 years; (b) voluntary whole blood donation; (c) weight ≥50 kg; (d) negative for HIV, hepatitis B, hepatitis C, and syphilis screening by rapid assay; (e) fully understanding the study's purpose and method, willing to participate, and signing written informed consent. The exclusion criteria were: (a) aged <18 or >60 years; (b) pregnancy; (c) using antibiotics within 48 h before blood donation, taking immunological drugs, or undergoing immunosuppression therapy in the past 6 months; (d) organ transplant receiver or blood transfusion within the last year; (e) Laboratory re-identification of infectious disease infections; (f) medical, psychological, social, or other conditions inconsistent with the protocol or affecting the ability to sign informed consent as assessed by clinician.

All potential blood donors were first screened by HIV, HBV, HCV rapid test and ABOs test at blood donation mobile health vehicles. After excluding invalid donors, eligible donors had a brief physical examination by nurses (including blood pressure, pulse rate, and body temperature measurement) and then started blood donation. According to the legal and regulatory standards in China, donors can only choose to donate whole blood in volumes of 400, 300, or 200 ml each time. For apheresis platelets, the donation amount is 1–2 therapeutic units, with each therapeutic unit not exceeding 200 ml. Prior to recruitment, donors were clearly informed of the study purpose, procedures, and potential risks/benefits. They were explicitly told that refusal to participate would not affect their blood donation services or subsequent benefits, and participation was entirely voluntary. Only those who signed the written informed consent form proceeded to the survey. Following the recruitment process, participants completed anonymous questionnaires via WeChat QR code, providing only a unique blood donation code (for data verification) instead of personal identifiers (e.g., name). This code was not linked to any personal information in the final dataset. After donation, participants stayed for 15-min safety observation and received incentives.

In the pilot study, feedback regarding the clarity, length, and item relevance of the questionnaire was gathered. Based on this feedback, the questionnaire was revised and subsequently validated. The validation results indicated a high level of internal consistency (Cronbach's α = 0.85) and outstanding construct validity (Kaiser Meyer Olkin, KMO = 0.91). Participants in the pilot study were excluded from the subsequent investigation. This formal questionnaire had several sections: socio-demographic characteristics, health status, donation willingness and attitude, and incentive attribute preference To ensure validity, an expert panel (including an epidemiologist, blood collection nurses, blood center laboratory director, and chief transfusion technician) formulated and evaluated all items. The socio-demographic section included variables like gender, age, and educational attainment. Basic health status information covered infectious diseases, chronic conditions, and medication. Also, the survey included blood donation willingness, attitudes, and incentive preferences. The details of the questionnaire were presented in Supplementary File 1. A password-protected master identification list was stored exclusively on investigator device with limited access. Data was encrypted and regularly backed up to maintain the confidentiality and reduce data loss risks.

This study aimed to investigate China's blood re-donation rate and factors influencing re-donation willingness. In China, the blood donation rates exhibit variations across different regions. Based on the donation history data from the Jiangsu Provincial Blood Center, the estimated blood donation rate of the general population ranges from 10 to 35%, and the re-donation rate stands at 60% among donors. Hence, we assumed first donors would be 35% and repeat donors would be 50%. A 95% confidence level (α = 0.05) and 90% statistical power (β = 0.10) were set. Considering a 10% dropout rate, a minimum sample size of 510 participants was needed. The sample size was calculated using Power Analysis and Sample Size software (Version 15.0.5).

Descriptive statistical methods were used to present background characteristics and willingness profiles. The Kolmogorov-Smirnov test evaluated data normality for continuous variables. Normally distributed data were described by mean ± SD and analyzed with the t-test, while non-normally distributed data were presented as median with IQR and analyzed by the Mann-Whitney tests. The Fisher's exact test or the chi-square test was used for categorical data like AEs. A binary logistic regression model was used to study the factors influencing vaccination willingness. Then, a multivariable logistic regression analysis was conducted by including variables with p ≤ 0.1 from the prior binary logistic regression analysis, while adjusting for age, gender, and education degree. Only variables with a two-tailed p < 0.05 were considered significant. Comparisons were shown by aORs and 95% CI. Cases with missing data were excluded. R software (version 4.3.0) was used for statistical analyses. Figures were made with GraphPad Prism (version 9.0.0), and results were presented according to the STROBE guidelines.

A total of 1,460 volunteers were recruited and completed the survey. 102 questionnaires were excluded due to inconsistent responses or missing data. Finally, 1,385 questionnaires were included in analysis, with 320 (23.56%) first-time donors and 1,065 (76.44%) repeat donors. Even with the higher-than-expected proportion of repeat donors in the actual sample, this subgroup further enhanced the statistical power for between-group comparisons. The cohort had 757 males and 628 females. Among first-time donors, half were young, while middle-elderly were more common among repeat donors. Blood types A, B, and O each accounted for 29.23%−30.71%, and only 10.68% were AB. The Rh-positive type was predominant (98.53%). Most donors (75.11%) donated 400 ml of whole blood, and the rest preferred ≤200 ml or 300 ml red blood cell components. 843 (62.08%) lacked local household registration, and 1,014 (74.67%) had long-term local residency. Also, over 60% of donors had a higher–education degree. The 40-h workweek was the most reported work duration (37.04%; Table 1).

Over half of donors were familiar with blood-donation knowledge and policies, and only 2.6% were completely unaware. Notably, first-time donors had lower awareness of donation policies than repeat donors (χ² = 170.413, p < 0.001). For stimulating blood donation willingness this time, personal willingness was the main factor (54.0%; Figure 1A). Sub-group analysis showed that on-site promotion might have a greater influence on first-time donors, though not significantly. In contrast, repeat donors cited personal willingness at a significantly higher rate (57.6 vs. 42.2%, p < 0.001). Altruism was the main purpose of blood donation in both groups (Figure 1B). However, first-time donors were more likely to donate out of curiosity and for free blood-type testing (20.6 vs. 4.5% and 15.6 vs. 6.1%, respectively).

Regarding geographic preferences, 51.0% of donors prioritized city centers or commercial districts as blood-donation sites and 63.0% donors preferred 15-min commute time radius. During the blood-donation process, the technical proficiency of nurses was the key concern (461 responses, 33.9%). First-time donors expressed a higher level of anxiety regarding pain (44.7 vs. 35.8%, p = 0.004) and adverse reactions (44.1 vs. 32.4%, p < 0.001) compared to the repeat-donor group. The primary reason for non-donation across all donor groups was temporary physical unfitness (51.2%), with repeat donors reporting this reason more frequently (53.7 vs. 43.1%, p = 0.001). Fear of blood donation was disproportionately higher among first-time donors (26.9 vs. 11.3%, p < 0.001).

In terms of the donation incentives, first-time donors preferred educational initiatives related to blood-donation knowledge (59.1 vs. 51.5%, p = 0.018), while repeat donors prioritized policy incentives such as blood reimbursement (44.4 vs. 37.2%, p = 0.022; Figure 1C). Reward preferences diverged significantly: first-time donors preferred donation certificates (49.4 vs. 41.9%), whereas repeat donors favored priority medical services (41.6 vs. 32.5%, p = 0.004) and medical registration fee waivers (33.8 vs. 25.9%, p = 0.008; Figure 1D). Communication preferences also differed, with first-time donors preferring phone notifications of test results (40.0 vs. 32.6%, p = 0.014) and repeat donors favoring short messaging service alerts (68.6 vs. 55.9%, p < 0.001; Table 2).

Blood donation was highly safe among all donors, with the proportion of adverse reactions being merely 2.7%. The majority of adverse reactions were mild or moderate, and no serious adverse events were observed. In total, 11 (3.4%) out of 320 first-time donors and 26 (2.5%) out of 1,038 repeat donors reported at least one adverse reaction during the blood donation process (Table 3). Fever and sweating, as well as hypotension, fainting, and shock, were the two most prevalent kinds of adverse reactions among all donors (0.8 vs. 0.9%). Although the proportion of total adverse reactions reported by first-time donors was numerically higher than that of repeat donors, no statistically significant difference was detected (3.4 vs. 2.5%, p = 0.370). Each type of case was documented sporadically. First-time donors reported a significantly higher incidence of local pain compared to repeat donors (1.3 vs. 0.1%, p = 0.012). Additionally, first-time donors experienced a higher proportion of fever and sweating reactions than repeat donors (1.9 vs. 0.9%, p = 0.025). Conversely, repeat donors reported more cases of nausea and vomiting, facial pallor, hypotension, fainting, and shock than first-time donors, yet these differences were not statistically significant.

After rectifying biases by incorporating variables with p < 0.1 into multivariable logistic regression analysis, over 10 characteristics were identified as significant factors. Donors who previously donated 400 ml of blood were more willing to re-donate than those who donated <200 ml (aOR = 4.66, 95% CI: 2.21–9.85, p < 0.001). Moreover, as age advanced, 41–60 year-old donors were more likely to repeat donation than younger donors (aOR = 4.78, 95% CI: 2.03–11.23, p < 0.001). Additionally, repeat donors were more likely to have long-term local residency than non-permanent residents (aOR = 2.29,95% CI: 1.55–3.39, p < 0.001). Besides, participants who worked more than 5 days per week were more inclined to re-donate than those who worked 1–4 days per week (aOR = 2.11, 95% CI: 1.25–3.57, p = 0.005; Table 4).

Notably, donors with comprehensive blood donation knowledge were more willing to repeat donation (aOR = 8.76, 95% CI: 3.24–23.66, p < 0.001). However, introduction by medical staffs (aOR = 0.52, 95% CI: 0.31–0.89, p = 0.016) and impulses or curiosity (aOR = 0.23, 95% CI: 0.12–0.42, p < 0.001) may decrease the re-donation rate. Furthermore, first-time donors were more likely to choose free blood type testing than repeat donors (aOR = 0.41, 95% CI: 0.24–0.73, p = 0.002). Nonetheless, repeat donors considered personal willingness as the key factor in enhancing willingness (aOR = 2.00, 95% CI: 1.40–2.86, p < 0.001). Rewards were another significant factor as increasing the re-donation willingness among existing donors (aOR = 3.33, 95% CI: 1.32–8.41, p = 0.011).

In the blood donation process, the environmental experience at the donation site (including air quality, lighting, and comfort) was more critical for enhancing re-donation willingness than nursing staff skills or attitudes (aOR = 2.51, 95% CI: 1.19–5.53, p = 0.016). Donors who had adverse events during donation, especially those with local pain or fear, were less likely to donate again (aOR = 0.49, 95% CI: 0.34–0.71, p < 0.001; aOR = 0.01, 95% CI: 0.00–0.24, p = 0.005). First-time donors with fear also refused to re-donate (aOR = 0.37, 95% CI: 0.23–0.59, p < 0.001). Issuing certificates post-donation had no effect on re-donation willingness (aOR = 0.60, 95% CI: 0.42–0.84, p = 0.003). However, increasing donation award value (aOR = 1.70, 95% CI: 1.06–2.73, p = 0.029) and providing a one-time flu vaccination (aOR = 1.66, 95% CI: 1.06–2.61, p = 0.028) attracted interest in re-donation. Variables with p > 0.05 were put in Supplementary Table 1.

Nevertheless, this cross-sectional cohort study has several limitations. Gender and age disparities between the two groups may bias result interpretations, so we controlled confounding variables in the multivariate logistic model. The single-time-point study design makes it difficult to establish causal relationships between investigated factors and repeat donation behavior due to unmeasured variables or temporal variations. Also, all participants being donors may introduce selection bias as those with positive experiences are more likely to participate, distorting results and limiting generalizability. Moreover, the fixed order of options in multiple-choice questions and excessive questions may make blood donors impatient and averse, increasing the likelihood of choosing early options. The data rely on self-reported information from donors, which may have biases like recall or social desirability bias, affecting data accuracy on motivators and concerns.

Aside from the aforementioned limitations, this study also has several drawbacks that warrant further addressing. Given that our data collection was restricted to the period from December to June, the observation window was comparatively limited. This resulted in insufficient sample sizes across potential seasonal subgroups, precluding robust statistical inference. Future studies should therefore extend data collection to a full calendar year to enable rigorous exploration of seasonal factors influencing donation behavior. Additionally, the omission of crowd pressure leads to an incomplete representation of the determinants of donation behavior. Future research should incorporate a more comprehensive scale to quantitatively analyze its association with re-donation willingness. Furthermore, since many religious teachings emphasize altruism, which may correlate with blood donation willingness, future studies should include religion as an additional potential factor to examine its association with re-donation willingness in the Chinese context.

The findings of our research indicated that the factors associated with the willingness to donate blood vary between first-time donation and repeat donation. The individual willingness, local residence and donation knowledge popularization were significantly correlated with the repeated donation behavior. Additionally, providing rewards, enhancing the value of the gifts and offering free one-time vaccine immunization also had a great impact. This suggests that repeated donation is highly dependent on the cohort formed by feedback. The optimization of national blood transfusion services and the implementation of evidence-based targeted strategies by increased government financial investment to elevate the donation rates, including augmenting knowledge dissemination among first-time donors and reinforcing policy incentives for repeat donors, are crucial for ensuring blood availability.