2.1 Theoretical Framework and Methodological Underpinning

Qualitative studies allow for a detailed understanding of an issue, including how the context impacts participants' experiences and perceptions [20]. Lambert and McKevitt advocate the value of anthropology in healthcare research as such an approach regards people's lay perceptions and actions as socially and culturally specific and valid within that frame [21]. This is important because narrowly framing an investigation in biomedical terms limits the ability to design interventions that are responsive to community needs [21]. A systematic review shows that the lay perspectives of health beliefs and experience in affecting donation-decision processes for nondonors still remain understudied [22]. Studies in various cultural contexts have shown that blood donation in nonpandemic times can be influenced by factors such as religious beliefs [23-26], social expectations [26, 27], family ties [28] and subjective trust in a healthcare system [29]. An ethnographic study in India indicates that the context of religious gift-giving, sacrifice, caste, kinship, and nationalism can affect blood donation practice [30]. Whereas during pandemic times, personal perceptual factors, government institutional factors and social factors that developed from the new social norms are noted as factors in donating blood for active donors in a Chinese community [31]. Although the pandemic is shown to have less impact on active donors' motivation in Germany [16], it has a stronger impact on nondonors in Australia and Germany during pandemic times, with subjective norms and self-efficacy affecting their intention to donate [16, 32]. Hence, across different social contexts, multilevel social and cultural factors were identified as relevant influences on blood donation during pandemic and nonpandemic times, which suggests the utility of taking a holistic approach to the study of this phenomenon. Anthropology, and specifically CMA, are therefore appropriate tools for understanding complex decision-making processes around blood donation during the COVID-19 pandemic. CMA investigates the interrelationship between people's micro-level experience, the intermediate-level of social organisation action and the macro-level of social structure (Figure 1). Furthermore, it explicitly recognises connections among social groups to larger national and global societies [33]. This framework emphasises how power structures can affect people's health beliefs and behaviours, particularly in and through social relationships [34].

According to CMA, global, capitalist, political and economic forces can affect human relationships, social behaviours and collective experiences [19]. Further, factors at the individual, micro-social, intermediate-social and macro-social levels are interlocking and can help to explain health behaviours [19]. At the individual level, personal factors and social networks can influence a person's individual donation behaviour. At the micro-social level, the interaction between people and healthcare personnel can influence decisions to donate. At the intermediate-social level, institutional policies of governments can affect people's decisions. Finally, at the macro-social level, ethnocultural beliefs of a society and the political economy embedded in the global system can affect a person's decision.

Anthropological studies are important for placing objects of study in a social context; this frequently involves descriptive analysis [34]. A qualitative descriptive approach was used in this study to describe the perceptions and experiences of participants [18]. Qualitative description recognises the subjective nature of human experience, and the findings are presented in a way that directly reflects or closely resembles the terminology used by research participants [35]. It illuminates how people feel and experience things and the factors they think facilitate or hinder their behaviours [36]. Qualitative description is characterised by a lower level of interpretation and inference from researchers, so the researchers can stay closer to the data and the words used by informants [37]. The CMA framework adopted in this study is well-suited to contextualise perceptions and experiences in the social and cultural context specific to a Chinese population.

2.2 Ethical Considerations

Ethics approval was obtained from the Human Subjects Ethics Subcommittee of The Hong Kong Polytechnic University before the start of the study (HSEARS20171013002). Written informed consent (for face-to-face interviews) or audio-recorded informed consent (for online interviews) was obtained from the participants before data collection.

2.3 Data Collection

Study participants were selected by purposive and snowball sampling. The following criteria were used for the selection: age 18 to 65 years, never having donated blood (nondonors) or not having donated blood for 3 years after a prior donation made before 1 January 2018 in Hong Kong (lapsed donors), being born and educated in Hong Kong, residence in Hong Kong at the time of the study. Blood donors were considered as lapsed if they had not donated for 36 consecutive months at the time of recruitment following literature [38].

These sampling criteria ensured that the participants had been exposed to the social environment of Hong Kong over the long term. The eligible age for blood donation ranges from 16 to 75 years in Hong Kong [39]. However, those aged 16–17 years can only donate blood with parental consent, and those aged over 66 years are required to undergo annual health assessments and receive approval from the BTS [39]. Therefore, only those who were between 18 and 65 years of age during the study period were sampled to minimise the potential impact of parents and health conditions.

The participants were recruited in three ways. Some participants (N = 4) were undergraduate students who had responded to a survey conducted at a university about the perception of blood donation [40]. Others were recruited via materials posted in public areas at the same university (N = 6 nondonors and 2 lapsed donors). The remaining participants (30 non- and 38 lapsed donors) were recruited from the broader community, including from social service agencies and residents' associations in different districts, a doctor's fan page on Facebook (with 25,000+ followers during the study period), and referrals from prior participants. Informational recruitment posters indicating the eligibility criteria were provided to these organisations and the owner of the Facebook page after obtaining the consent from these agencies, associations and the doctor's Facebook for participant recruitment.

An open-ended interview guide (Supporting Information: Appendix 1) was developed after referencing literature on blood-donation perceptions and barriers [1, 25, 41-45]. Although the questions were designed to correspond to the four levels of the CMA framework to examine the factors, probing questions that are not in the interview guide were asked following the responses of each participant to dig into the participants' experiences. The participants were interviewed individually by the same interviewer, who had undergraduate and master's-level education in sociology. The interviewer received intensive training and supervision from the first and second authors throughout the data-collection process.

All interviews were conducted in Cantonese (the native language of the participants and the interviewer). Because the interviews were conducted during the COVID-19 pandemic, the participants had the option of an online interview. Ultimately, four interviews were conducted face-to-face, whereas the remaining 76 interviews were conducted online. Each interview lasted 1–1.5 h, and the audio was recorded with the participant's consent. To compensate the participants for their time, each one was given a supermarket cash coupon worth HK$200 (approximately 25.64 USD) upon completion of the interview.

2.4 Data Analysis

After each interview, the interviewer documented the key points and impressions on each interview in an interview diary. The first and second authors also noted the key points and impressions when listening to the audio recordings of all interviews. The notes by the first three authors were compared and discussed throughout the data-collection process. Interviews were transcribed verbatim, and the transcripts were examined line by line in the coding process. The first, second and third authors separately coded the interview transcripts using thematic analysis, according to Braun and Clarke [46]. The codes, categories and themes identified along with supporting interview quotes were documented in a coding table [47].

As this study follows the CMA framework, identified codes, categories and themes were consolidated into four social levels for analysis. Other data not following the CMA framework was also coded inductively and documented in the coding table. Data saturation is defined as data redundancy with no new themes or codes emerge from further interviews [48], which was achieved at the 31st interview for nondonors and at the 34th interview for lapsed donors. Additional interviews were conducted for both nondonors and lapsed donors to validate and expand the already identified patterns, concepts and themes [49]. The coded data were compared and discussed among the first three authors in routine research meetings and were discussed with and agreed upon by the fourth author, who was a department operations manager and a registered nurse of the BTS during the study period. Selected interview quotes were translated from Chinese to English, and back-translation from English to Chinese was then conducted to confirm the accuracy of the translations.

2.5 Study Rigour and Data Trustworthiness

To ensure the rigour of the study design and methods, the criteria developed by Lincoln and Guba [50] were considered. Data collection and analysis were conducted in accordance with the Consolidated Criteria for Reporting Qualitative Research (Supporting Information: Appendix 2) [51]. The interviewed participants were asked to read their transcripts to ensure that their meanings were accurate, and all the participants confirmed the content of their transcripts. Coding meetings were held monthly among the first three authors to ensure agreement. The fourth author, as a department operations manager of the BTS, regularly provided opinions on the coding procedure in these meetings. Consensus was achieved among all four authors.

3.1 Theme 1: Barriers to Blood Donation During the COVID-19 Pandemic

The barriers discussed in this theme are pandemic-related and are started at the CMA's intermediate-social level, which the participants mentioned as having the most proximate and fundamental effects on their perceptions and behaviours at other levels.

3.2 Theme 2: Underlying Barriers to Blood Donation Preceded the Pandemic

As mentioned, the reasons participants did not donate blood during the COVID-19 pandemic were not merely due to the pandemic itself, but also embedded in their perceptions and past experiences of blood donation before the pandemic. Prepandemic deterrents to donation spanned the levels of the CMA framework.

4.1 Participants' Nondonation as a Product of Interaction of the Four Social Levels of the CMA Framework

The adopted CMA framework has shown that the participants' experiences are a product of interaction among people's actions at both individual and microsocial levels, policies at the intermediate-social level and political and economic forces at the macro-social level. The decision not to donate during the pandemic cannot be explained by pandemic factors alone. Although the participants' sense of being a ‘good citizen’ arising from the new social norms developed in the pandemic at the intermediate level (quarantine policy) and the macro-level social structure (collective responsibility) had affected their micro-level perceptions (blood donation as unnecessary and risky and healthcare personnel as dangerous), their experiences at different social levels preceded the pandemic had played an important embedding role in reinforcing their nondonation during the pandemic (Figure 2) [19]. Simultaneously, the microspheres reinforced the macrosocial structures.

The government's quarantine and infection control policies at the intermediate-social level are the most proximate factors for the participants not donating. Interviewees interpreted these policies as indicating that donating blood was an unsafe, unnecessary and nonurgent activity that involved risky crowd gathering. In particular, the environment of donor centres was demotivating, which is consistent with previous studies showing the concerns about close contact with other blood donors dissuaded people from donating blood during the pandemic [52]. Blood donation conflicted with the participants' understandings of institutional appeals for infection control. As noted in past studies, those with a stronger sense of adherence to the COVID-19 infection-control guidelines are less likely to donate blood [53]. Indeed, all the participants perceived adherence to the infection-control guidelines as a collective responsibility, which significantly prevented them from donating blood in the COVID-19 pandemic. Such a sense of collective responsibility made the participants believe that blood donation during the pandemic would not be supported by their social network, and peer pressure and the acceptance of others have been crucial for people donating blood during the COVID-19 pandemic [54]. Quarantine policies at the intermediate-social level affected macro-social ideology and then the participants' perceptions and behaviours at the individual level. Under the established policies, the prevention of infection was socially constructed as a collective responsibility, which had become a social and cultural norm at the macrosocial level during the pandemic. A study in Germany notes that personal moral norms of donation had become even more ineffective for nondonors during the pandemic when the personal moral norms had become focused on infection reduction [16]. Without the government's appeal about the importance of blood donation during the pandemic, social and cultural macrosocial norms made infection prevention become the most prioritised. This overrode the importance of blood donation and demotivated the participants from donating blood as a result.

The stigma about healthcare personnel and facilities as being contaminated served as a notable barrier for the participants at the micro-social level. The participants stereotyped BTS personnel as a risky and dangerous group during the pandemic. To the participants, blood donation and donor centres were associated with hospitals, viruses, bacteria and COVID-19. Such preconceptions have also been documented in other parts of Asia as a demotivating factor for people who might ordinarily donate blood [55]. Besides, blood was perceived as dangerous by participants at the individual level, as many of them extrapolated the blood-borne transmissibility of other infectious diseases to COVID-19. This knowledge discouraged them from donating during the pandemic, which is similar to the situation in other Chinese communities [3]. This perception at the individual level also reinforced the stigma about BTS personnel in donor centres at the micro-social level because they are required to be involved with blood in the donation procedure. People can be culturally regarded as dirty or dangerous when they are in contact with ‘polluted’ and ‘contaminated’ things, which can result in their seclusion from society [56]. These notions strongly discouraged the participants from donating blood during the pandemic.

The nondonation decision of the participants during the pandemic cannot be explained by pandemic factors at four social levels alone. Their experiences at different social levels preceded the pandemic also embedded in reinforcing their nondonation decision during the pandemic. Some studies show that people protest against the new social norms (‘the new normal’) initiated by governments during the COVID-19 pandemic [57], but the present participants complied with the government's pandemic policy. Nevertheless, their nondonation showed hints of resistance against the structural issues in blood-donation practice in Hong Kong that had been embedded in the time before the pandemic. These pandemic-preceded factors are discussed further below.

4.2 Participants' Nondonation as an Expression of Dissatisfaction Embedded in the Time Before the Pandemic

Not engaging in desirable health behaviour has been suggested as a form of advocacy calling for change in health structures, delivery of healthcare, and access to social programs and medical technology [34]. Echoing the recent CMA appeal by Dutta and Basu [58], the decision not to donate blood could show a hint about participants' expression of dissatisfaction toward the blood donation institution and policy, and such dissatisfaction has often been embedded in the participants' minds before the pandemic. Nondonation during the pandemic, we argue, should not be viewed as isolated incidents that only happened in the pandemic, but nondonation has a processual nature in which these demotivating factors have been occurring over time before the pandemic and they influence each other.

Many participants, especially those who had lapsed from blood donation, expressed dissatisfaction toward deferral, BTS personnel and donor screening policy. Participants commonly expressed frustration with deferral and BTS personnel. This should be noted as an important embedding factor as past literature has also demonstrated that deferral [59] and unsupportive donor centre staff [60] can be significant barriers for donors to return for blood donation. Some participants were also dissatisfied with the current donor screening policy, which excludes people with nonmainstream sexual orientation.

In the context of the difficult political relationship between Hong Kong and mainland China at the macrosocial level, our study shows that the political economic situation of blood donation had also demotivated the participants from donating blood before the pandemic. The potential monetary value of blood in mainland China was widely perceived by the participants, which undermined their trust in the BTS and served as a prevalent factor in their nondonation even before the pandemic. The participants' decision not to donate is consistent with past literature [29, 60], which showed that trust in a healthcare system is a significant factor that can affect people's willingness to donate blood. The participants' nondonation can show hints of their dissatisfaction with the political and economic structure of Hong Kong and of their hope for more transparent blood donation policy.

4.3 Comparing Barriers to Donation Reported by Non- and Lapsed Donors

To both the non- and lapsed donor participants, their immediate barriers to donating were pandemic-related. Quarantine policy made both groups of participants perceive the behaviour of blood donation as unnecessary and risky that could impose a negative impact on others. However, lapsed donors tended to have a higher sense of blood donation as important during the pandemic than the nondonors. The pandemic also affected both groups of participants to view healthcare personnel negatively. As BTS personnel were often perceived as healthcare personnel who provided patient care by the participants, this had demotivated them to donate blood during the pandemic. Another alarming finding noticed by both the lapsed- and non-donors is their common perception of blood as the transmission route of COVID-19, which has implications for the blood donation appeal in future pandemics.

Prepandemic barriers affecting the two groups of participants not donating were noted. Lapsed donor participants tended to report more about their unpleasant experiences with the past donation as part of their reasons for lapsing. These unpleasant experiences could be related to deferral, staff of donor centres and donor screening policy. The uncomfortable physical feeling after donation also made lapsed donors to have more hesitation in donating during the pandemic. Nondonor participants also reported barriers prepandemic, but these prepandemic barriers were more due to their suspicion of the blood donation system, which was originated from the macro-level political hardship between Hong Kong and mainland China.