Blood transfusion in haematology: A qualitative exploration of patients’ and healthcare professionals’ perceptions

Abstract Objectives Repeated blood transfusions are indicated for the management of patients with cancer or blood disorders. Patients’ perceptions about transfusions may be associated with decision‐making and coping, which has been under‐explored in the haematology context. This study therefore aimed to explore haematology transfusion patients’ and HCPs’ perceptions of blood transfusion, drawing on theory and previously identified themes of transfusion perceptions. Design Semi‐structured interview study with 14 adult blood transfusion patients and 14 HCPs (consultants, registrars, nurses) at two UK haematology units. Methods Patient‐ and HCP‐tailored topic guides were developed based on themes of blood transfusion perceptions identified in a systematic review: ‘Health benefits’, ‘Safety/risk’, ‘Negative emotions’, ‘Alternatives’ ‘Decision making’ and ‘Necessity’. Transcripts were analysed using deductive and thematic analysis. Patient and HCP themes were compared using triangulation methods. Conceptual models (one for patients, one for HCPs) specific to haematology portraying the association between themes were developed. Results Findings for patients and HCPs converged with transfusion reported as beneficial for patients, who were largely involved in the decision‐making. Both groups also reported concerns about transfusion, including iron‐overload, allergic reactions and challenges to deliver transfusions in time‐pressurized services. Themes in the conceptual models included patient ‘Burden’ of receiving repeated transfusions and ‘Supportive relationships’, reflective of patients’ positive interactions with other patients and HCPs in the haematology unit. Conclusion Despite the challenges for patients receiving repeated transfusions, convergent perceptions suggest a shared understanding of patients’ transfusion experiences. Identified challenges could inform ways to improve transfusion services and patients’ experiences.


BACKGROU N D
Patients facing threats to their health (e.g., illness or symptoms) seek to understand these risks by forming illness and treatment perceptions, such as how the threat could be 'controlled' or managed through treatment (Leventhal et al., 1980). Considering an illness as 'controllable' through treatment is linked, for example with improved patient outcomes, such as quality of life (Schoormans et al., 2014;van Wilgen et al., 2008). In parallel with managing emotional responses, such as fear and anxiety, patients are likely to consider the 'Necessity' of a treatment and their 'Concerns' about the treatment (Horne et al., 2013). Patients may hold concerns, for example if they associate a potential treatment with a certain level of risk. Blood transfusion is a very common procedure in hospitals and represents an example of an intervention where patients face potentially significant risks from the transfusion, such as infection transmission or adverse reactions to the blood products (Serious Hazards of Transfusion (SHOT), 2021).
In the United Kingdom (UK), around two million blood components were issued from the UK Blood Services in 2020 with the main indications for treatment being severe bleeding or to replace blood loss due to conditions affecting blood cell production (e.g., bone marrow failure, anaemia) (Narayan et al., 2021). Despite transfusion being used for patients worldwide (around 3.5 million patients across reported concerns about transfusion, including iron-overload, allergic reactions and challenges to deliver transfusions in timepressurized services. Themes in the conceptual models in- Identified challenges could inform ways to improve transfusion services and patients' experiences.

K E Y W O R D S
Blood transfusion, cancer, haematology, outpatients, treatment decisionmaking, treatment perceptions

What is already known on this subject?
• Transfusion perceptions (e.g. 'concerns' about risks; 'alternatives') may influence patient and HCP decision-making. • Haematology perceptions have been under investigated despite repeated transfusion use.

What this study adds?
• New haematology themes presented in the conceptual models inter-link and influence 'decisionmaking' perceptions. • Patients report the 'burden' of repeated transfusions, noticeable 'health benefits' and the importance of 'social connection'. • HCPs report unique challenges with providing a high volume of transfusions, indicating the need for service reforms.
Few studies have used qualitative methods to explore transfusion perceptions (Abdul-Aziz et al., 2018) meaning that we lack in-depth understanding of participants' perceptions and experiences.

Setting and participants
Haematology patients and HCPs from NHS haematology day units were jointly included in the study to understand both groups' perceptions in parallel, with few studies having previously explored both groups' perceptions concurrently (Abdul-Aziz et al., 2018). HCPs are a key part of a transfusion, often responsible for transfusion decision-making on behalf of the patient and their perceptions may not align with those of patients (Morita et al., 1999). This may prevent treatment goals from being readily established and perceptions being discussed openly, leading to misunderstandings (Barry et al., 2000;Mead & Bower, 2000).
The specific patient and HCP inclusion criteria for this study were: -Adult patients (aged 18+) with a non-acute haematological disorder attending UK NHS haematology day units for a blood transfusion were eligible for inclusion. Patients were excluded if they had limited English or cognitive impairment. -HCPs, including consultants, physicians and haematology nurses, working in the same haematology unit as the recruited patients, who discussed or delivered transfusions to their patients, were eligible for inclusion.
Ethical approval was granted from South Central -Hampshire B Research Ethics Committee (15/ SC/0757).

Recruitment
An on-site member of staff facilitated recruitment and informed each eligible patient attending for a transfusion about the study providing a Participant Information Sheet (PIS). The study was also verbally promoted to HCPs in the haematology units by the site Investigator and a HCP-specific PIS was provided. A minimum of 13 patients and 13 HCPs was the target sample size following data saturation sample size guidance (Francis et al., 2010). Interviews after the tenth interview per group (e.g., 11th, 12th…) were assessed during data coding to determine whether additional themes could be created or whether the themes were completed and sufficiently populated.

Topic guide
Patient-and HCP-specific topic guide questions were structured around the themes of blood transfusion perceptions identified in the review: 'Safety/risk', 'Alternatives', 'Health benefits', 'Negative emotions' and 'Decision making' (Abdul-Aziz et al., 2018), in addition to the construct of 'Necessity' (Appendix 1). This construct was added as it is central to the understanding of treatment perceptions, defined as 'the perceived role of [the treatment] in protecting against deterioration of the present and future health status of the patient' (Horne et al., 1999b). The topic guides were developed in collaboration between psychologists, a patient representative and Consultant Haematologist, and piloted prior to data collection.

Procedure
Participants were interviewed individually, face-to-face, in the haematology unit or by telephone if preferred. For face-to-face interviews, patients were offered to be interviewed before, during or after their transfusions in the haematology unit. A demographic questionnaire was used to collect participant demographic data. All participants provided informed consent before interview. Interviews lasted between 30 and 45 min and were audio-recorded, transcribed verbatim and fully anonymised. Informed consent was not collected from participants to share anonymised transcripts, therefore transcripts were unable to be uploaded to open access repositories.

Analysis
Transcripts were first analysed using a deductive directed content analysis approach (Hsieh & Shannon, 2005), with participant responses coded into the blood transfusion themes that they were judged to best represent (Abdul-Aziz et al., 2018). For example, a quote 'patients benefit from transfusions to restore their energ y' was coded to the theme 'Health benefits' or 'I do worry less about how risky the blood is nowadays', coded to 'Negative emotions' and 'Safety/risk'. Perceptions that could not be coded into existing themes were coded into a temporary 'other' category. Two researchers independently coded one patient and one HCP transcript (7% of total), and inter-rater reliability assessed using percentage agreement with >75% considered an acceptable level of agreement (Stemler, 2004).
Inductive thematic analysis was subsequently performed to generate subthemes for data deductively coded to each transfusion theme. Content coded to the 'other' category were also developed into themes, identified as patterned responses (Braun & Clarke, 2006) to describe the blood transfusion perceptions reported by patients and HCPs. This involved generating initial codes for interesting features of the data and collating codes into potential themes, which were then reviewed and named (Braun & Clarke, 2006). Data saturation was assessed by BV during coding to determine that the final three interviews from each group produced no new themes (Francis et al., 2010).
Themes between patients and HCPs were compared using a triangulation approach (Farmer et al., 2006;Hopf et al., 2016). For this, a convergence matrix was generated and subthemes were tabulated as either: 'agreement', a comparable subtheme present for both groups; 'partial agreement', partially comparable subthemes; 'disagreement', a contradictory finding between subthemes or 'silent', subtheme recognised by only one group.
Final themes were organised into patient-and HCP-specific conceptual models of blood transfusion perceptions, specific to the haematology context.

Participant characteristics
Fourteen patients (eight from site 1, six from site 2) and 14 HCPs (seven from site 1, seven from site 2) were included in the study out of 59 patients and 43 HCPs who were informed about the study from both sites and invited to participate. Participant demographics are presented in Tables  1 and 2. Red blood cell transfusions were provided to 93% of patients and 50% had previously received platelets. Thirteen patients were interviewed during transfusions and one interview started pre-transfusion.

Reliability of perceptions and data saturation
Inter-rater agreement for coding decisions was 67% for patient transcripts and 60% for HCP transcripts. Full consensus was reached on all disagreements.
Data saturation was reached at the 11th patient and 11th HCP interview with no new themes identified beyond this point. Yet due to the 12th and 13th participants in both groups continuing to provide rich data, populating the themes further, and additional 14th interview for both groups was conducted and included in the analysis being assessed for saturation (Appendix 2).

Patients' perceptions of blood transfusion
Perceptions were coded deductively into all six blood transfusion themes: 'Health benefits', 'Necessity','Negative emotions', 'Alternatives' and the re-titled constructs of 'Awareness of risk/Safety' and 'Involvement in decision making'. Three new themes were developed from data initially coded to 'other': 'Social connection', 'Burden' and 'Distinguishing between blood products' (Tables 3 and 4 provides sample quotations with a full list of themes provided in Appendix 3). Patients recounted negative consequences that they experienced during and after transfusion, mainly from painful cannulisation and beliefs about adverse effects; high iron levels risking organ damage. Some patients reported that reactions were only likely at the start of the transfusion or had no negative consequences to report.

Health benefits
Transfusions aided patients to keep going with their daily lives, reducing tiredness, improving blood levels and well-being. Many patients were persuaded of the benefits by HCPs or significant others and thought that other patients must also perceive their transfusions as beneficial. Two patients commented that the benefit was variable and can take time to be felt or that having transfusions left them feeling drained.

Necessity
Patients reported that their transfusions were a supportive treatment for immediate health needs (e.g., to support chemotherapy) or for long-term condition management. Some patients recognised that they needed transfusions through their symptoms (e.g., lethargy; n = 8). Other patients relied on HCPs and clinical indicators to determine the transfusion necessity, being often problematic for patients to establish a transfusion routine. Interaction with patients in the haematology unit Daily (discussion and blood product administration) 2 (14%) Weekly (discussion and blood product administration) 3 (21%) Daily discussion 4 (29%) Weekly discussion 3 (21%) Monthly discussion 2 (14%)

Negative emotions
Some patients reported worries, fears and frustrations with receiving transfusions or found them hard to initially deal with. Some patients disliked being cannulated or the time length of the transfusion. Despite this, a greater number of patients discussed having no concerns with transfusion or that any negative emotions were managed by communicating with HCPs or keeping a hopeful and positive outlook. Some patients liked attending for transfusion, expressing gratitude that blood was available.

Alternatives
Patients discussed alternative treatments (e.g., tablets: unspecified) trialled to reduce their transfusion requirement, or previous unsuccessful or non-feasible treatments (e.g., bone marrow transplant). Some patients reported preferring, yet lacking, an alternative to transfusion. Other patients commented that their body may naturally, with time, restore depleted blood cells.

Involvement in decision-making
Patients reported either being involved in the initial transfusion decision-making, or that they deferred decisions to HCPs, or that HCPs solely made the decisions. Many patients accepted HCPs' decisions, finding transfusion discussions positive or had limited or no choice but to accept the transfusion. For some patients, transfusions were almost an 'automatically' prescribed treatment.

Social connection
Six patients reported interactions with other patients in the unit and were positively involved in their transfusions (e.g., through asking HCPs general questions or to discuss their transfusion regimens). Some patients reflected on wider family and General Practitioner (GP) support or talked of a curiosity or connection to the blood donors. Yet, other patients felt that they lacked or had limited interaction with other patients, especially patients with different haematological conditions or to respect their privacy, feeling that greater involvement was complex.

Burden
Half of the patients interviewed reported that transfusions were a part of their routine life, that attendance was not a great burden and that the experience was consistent over time and easy (switching between red cell and platelet transfusions). Around an equal number of patients (n = 8) reported, however, that transfusions were inconvenient, involving lengthy appointments, frequent hospital trips and restriction to activities such as travelling. Alternatives: 'most of these patients that are on such regular transfusion programmes are on EPO (erythropoietin) are on iron, are on all the other kind of alternatives to blood that they can be on and despite that are still requiring a blood transfusion' (HCP 10, Senior House Office, Site 2) Burden: 'because these patients if they don't have their transfusions, they aren't able to get up and be, have that level of activity that they would have without that transfusion, so despite having to come in and have a cross-match one day, blood the following day and then only really have one day out of hospital a week, they still manage to maintain some quality of life' (HCP 14, Specialist Registrar, Site 2).
Organizational factors: In relation to the complexity of managing the slots: 'we'll sit down and we'll look and "well if I have to cancel this person this week, if they needed a transfusion in three weeks" time, are we going to be able to have a slot?' you know, that sort of thought behind it, so it's quite complicated … we try and stretch it because of the lack of availability of slots, we try and stretch it to the absolute maximum, so we/ and we try not to prescribe if we don't have to prescribe' (HCP 3, Specialty Doctor, Site 1).
Stability and variability of transfusion perceptions: 'before blood transfusions were a one-off sort of thing … since being here, it's actual haematology conditions that require regular blood transfusions, so I see a different side now, and I see, erm how reliant people are on them' (HCP 5, Consultant Haematologist, Site 1) Supportive relationships: 'Erm, some of them might get some benefit out of that [lengthy appointments] cause it gives them an opportunity to have a chat with patients who are going through a similar experience and see that they're not alone' (HCP 6, Consultant Haematologist, Site 1).

Distinguishing between blood products
Four patients shared perceptions about specific blood products, mostly platelets. These patients perceived platelets fairly positively (e.g., quicker to infuse) and had less curiosity over their donor origin. However, there were some knowledge gaps about platelets; what platelets did and where they were harvested from, and one patient specifically reported that she was receiving irradiated blood, with no greater elaboration on the distinction of this.

Conceptual model of patients' haematology blood transfusion perceptions
In the patient's model (Figure 1), 'Burden' and 'Safety/risk' are proposed to be linked with 'Negative emotions' due to some patients reporting their transfusions to be time-consuming or involving risk. 'Health benefits' and 'Social connection' are associated with perceptions in the 'Safety/risk' theme, due to patients often deliberating risk vs. benefit and efforts made to have positive social interactions (e.g., with HCPs) reportedly eased some patients concerns or information needs. The 'Clinical factors' box displays three themes that impact on the behavioural, cognitive and emotional perceptions, such as the availability of alternatives, differences in perceptions depending potentially on the blood product being transfused and the transfusion 'Necessity' determined by HCPs or clinical indicators. In total, all themes are influences on 'Decision-making', such as transfusion agreement (patient consent) or HCP transfusion prescription.

Health care professionals' perceptions of blood transfusion
HCPs' perceptions were coded deductively into all six existing blood transfusion themes and four themes were additionally identified: 'Burden', 'Organisational factors', 'Stability/Variability of transfusion perceptions' and 'Supportive relationships'. A full list of themes is provided in Appendix 3.

Awareness of risk/safety
Half of the HCPs interviewed reported iron overload as the key risk for this patient group and HCPs widely discussed risk management through safe transfusion and blood testing practices. Patients were informed of risks such as iron overload, antibody production, infections or allergic reactions, as well as risks associated with non-transfusion verbally by HCPs. A few HCPs identified other medical and psychological risks with transfusions, such as bone marrow damage, venesections, nausea or patient dependency on transfusions.

Health benefits
HCPs reported that transfusions improved patients' symptoms and helped them to maintain their quality of life and daily activities. HCPs discussed transfusion benefit with patients to continue their transfusion yet were aware that the experienced benefit for patients will end at some point. A number of HCPs felt that their colleagues shared their view of transfusion being beneficial, yet some risk-benefit considerations in particular patient cases were questionable.

Necessity
HCPs reported that transfusions aided patient survival and protected their health, (e.g., when receiving chemotherapy or suffering from anaemia). HCPs reported prescribing transfusions based on clinical factors, such as haemoglobin levels and patient functioning. HCPs reported that some patients who depend on their transfusions experienced anguish when transfusions are deferred or delayed.

Negative emotions
HCPs expressed concern about the burden and physical impact of transfusion for their patients, with safety procedures reducing some of this concern and HCPs experienced capacity pressures. Despite this, HCPs offered patients reassurance to help reduce their anxieties, feeling that patients' fears may inhibit their involvement in transfusion discussions. HCPs found that some patients expressed positive emotions when transfusions were not required or viewed transfusions as a 'lifeline'.

Involvement in decision-making
The majority of HCPs reported advocating for and involving patients in transfusion decision-making despite some decisions being made without patient involvement. Many HCPs made transfusion decisions with other HCPs, laboratories or used guidelines and transfusions were often part of a broader treatment plan. A slight tendency was reported towards providing transfusion if symptoms indicated so, but patients did often question decisions or resisted transfusions. F I G U R E 1 Conceptual model of patients' haematology blood transfusion perceptions. Note: Single headed arrows indicate a direct one-way relationship, double headed dashed arrows indicate a bi-directional relationship between themes. Influences on 'Decision making' are shown using the central arrow Alternatives HCPs supported greater consideration into the use of alternatives and consulted patients who may request alternatives. Many alternatives, however, were either being tried to some benefit (Erythropoietin injections) or alternatives generally were not an option for this patient group. Some HCPs understood that patients may prefer alternatives, but opted for transfusion due to symptoms, clinical indicators or the evidence base.

Burden
The majority of HCPs anticipated that frequent and lengthy appointments must be burdensome for patients, and seven HCPs commented that transfusions have become a part of patients' routine lives. One HCP commented that patients could maintain some quality of life despite frequent hospital visits.

Organisational factors
Many HCPs from both sites acknowledged constraints to fully discuss patients' transfusion perceptions, such as busy haematology units. HCPs clarified the complexity of managing limited transfusion slots and high and costly blood use, feeling that solutions were needed to improve processes and ease capacity strain. Patient-or HCP-education was signalled as potentially required to enhance communication.

Stability and variability of transfusion perceptions
Many HCPs reported that their views about transfusion were stable over time and similar to their colleagues. Their views did, however, develop with haematology exposure with HCPs recognising their increased tolerance of lower haemoglobin levels and broadening knowledge about the varied and longterm use of transfusions.

Supportive relationships
HCPs discussed how they remained approachable to patients and established bonds by listening to patients and providing information like blood counts, enhancing patients' feelings of control. HCPs aimed to increase patients' level of comfort during the transfusions, or patient support was provided from other patients or patients' acquaintances whom they attended the transfusion with.

Conceptual model of health care professionals' haematology blood transfusion perceptions
In the HCP haematology conceptual model (Figure 2), three themes are proposed to link directly with 'Negative emotions': transfusion 'Burden', 'Safety/risk' concerns and 'Supportive relationships', which could help ease patients' concerns or worries. 'Supportive relationships' such as positive contact with HCPs and other supportive contacts, whilst patients receive their transfusions may also ease perceptions of transfusion 'Burden'. 'Safety/risk' vs. 'Health benefits' decisions are associated, being routinely considered and discussed with patients and the model also displays 'Clinical factors' which influence 'Decision making'. This is transfusion 'Necessity', the suitability of 'Alternatives' and 'Organisational' appointment availability etc. Prior to decision-making, the 'Stability or variability of HCPs' transfusion perceptions may influence tolerance of (lower) haemoglobin levels, practice consistency, transfusion acceptability and team decision-making agreement.

Triangulation of themes from patients' and health care professionals' perceptions
There was moderate convergence of the 95 patient and HCP subthemes (42%). Full agreement was evident on 24 occasions (35%), partial agreement on seven occasions (8%), disagreement on four occasions (5%) and silence on 39 occasions (52%). Appendix 4 lists all themes with their agreement ratings.
Agreement was greatest for the 'Health benefits', 'Alternatives' and 'Involvement in decision-making' themes, for example HCPs reported similar benefits of blood transfusion as patients. Partial ratings occurred through HCPs reporting patient-involvement in their transfusion decisions, yet some patients commented that although involvement was generally positive, transfusions were often not discussed. Disagreement in themes occurred when patients expressed having life restrictions ('Burden') in respect to travel, whereas some HCPs reported that transfusions can become a part of a patient's routine life.
Silence (i.e., theme was evident for one sample but not the other) occurred for themes, in which patients or HCPs expressed their particular patient/HCP role experiences (e.g., no negative consequences from transfusion (patients) or that patients are reviewed and consulted with when transfusions may need to end (HCPs)).

DISCUS SION
This qualitative interview study found that haematology patients and HCPs considered transfusions beneficial for managing patients' haematological disorders, with HCPs and patients working together to discuss transfusion decisions, risks and alternatives. However, both patients and HCPs experienced negative emotions, linked to the potential harms and 'Burden' of transfusion for patients; HCPs also reported concerns regarding constraints with transfusion services. Strategies to manage negative emotions were reported as remaining positive and hopeful (patients) or offering reassurance to patients or trusting clinical safety measures (HCPs).
There was moderate convergence between patients' and HCPs' perceptions, reflective of the groups having different roles (patients vs. providers). For example, HCPs' perceptions often exemplified their compassionate patient care: ('Supportive relationships'), within a busy and pressurised environment: ('Organisational factors') whilst patients were able to reflect on their interaction with other patients and HCPs ('Social connection').
The findings from this study extend previously identified themes related to blood transfusion treatment perceptions (Abdul-Aziz et al., 2018), with 'Decision making', for example, being a key theme. In contrast to other findings, many patients in this study did not find their transfusions to be overly anxiety provoking or hard to come to terms with (Randall & Wearn, 2005). Instead patients remained hopeful, keeping a positive outlook, similar to other haematology patients who reported acceptance of their conditions, changing life's priorities, and increasing engagement with HCPs and significant others as a way to cope with their conditions (Bulsara et al., 2004) (Prip et al., 2017). Patients may have used appraisal-focussed coping to accept the reality of their situation, redefining it as acceptable (Moos & Schaefer, 1984), similar to dialysis patients who perceive dialysis as a life-sustaining gift (Reid et al., 2016).
Some patients in this study reported limited interaction whilst in the transfusion units with both other transfusion patients and HCPs. With HCPs there was often a lack of involvement in the transfusion decision-making, as also cited in the broader cancer literature (Bruera et al., 2001). However, other patients enjoyed attending for their transfusions, with some not requiring extensive interaction with other patients. Such patients are likely to have developed ways that they cope with their transfusion, such as occupying themselves during their transfusions and engaging with patients and HCPs as and when available. As described in the Common Sense Self-regulation Model, the positive outcomes of such strategies inform the development of new perceptions (Leventhal et al., 1980(Leventhal et al., , 1984, which is evidenced with patients discussing positive emotions, such as gratitude. This research adds perceptions of blood transfusion as a life-sustaining intervention to the treatment perceptions evidence base, similar to perceptions of other repeated treatments, such as asthma medications (Østrem & Horne, 2015) or dialysis therapy (Karamanidou et al., 2014). How 'burdensome' a treatment is informs peoples' needs and concerns about the treatment (Horne, 2003). 'Treatment burden' for patients with long-term conditions is identified as the loss of freedom and independence for some patients (Demain et al., 2015). Relational disruptions can also occur due to treatment burden, including strained family and social relationships and feeling isolated (Demain et al., 2015). Some of these aspects were reported by the patients in this study, in terms of the time commitment required to attend for transfusions or necessary life-alterations (e.g., travel). Thus, it is important to continue to explore the construct of 'Burden' in more detail with transfusion patients as treatment burden may impact a haematology patient's life to a greater extent than shared by patients in the current interviews.
Strengths of this study include the concurrent investigation of both patients' and HCPs' transfusion perceptions from an under-researched transfusion context, where participants across two NHS sites reported similar perceptions. Patients were interviewed during or just prior to their transfusion, minimising recall bias and potentially heightening the vivid and actual account of their beliefs and experiences. Strengths also include organizing the themes of perceptions into patient and HCP-specific transfusion conceptual models. This helps to portray relationships between constructs. Limitations, however, include the lower than anticipated inter-rater reliability scores for the deductive coding and lack of secondary ratings for the triangulation analysis, increasing risk of subjective bias. However, multiple raters and co-researchers observed and contributed to the final thematic coding and exemplary quotes have been provided as supporting evidence (Nowell et al., 2017).
Future research could extend this study by utilising the applicable patient or HCP conceptual model in a larger quantitative study, translating the themes into questionnaire domains to investigate prevalence of the perceptions on a larger scale and relationships between the themes. This could impact the content of the themes in the existing model as well as formulating themes specific to this health and illness context. The models could also be used as future qualitative frameworks to investigate whether other patient and HCP groups hold similar beliefs about transfusion, such as sickle-cell patients who often face barriers to attend hospital or disclosing information to HCPs (Maxwell et al., 1999). Principles of open science could be more strongly adhered to in future qualitative studies through gaining consent to publish anonymised participant transcripts or larger extracts from interviews in online journal articles or patient targeted websites (Riley et al., 2019).
Through understanding patients' perceptions, greater support could be offered to patients through service enhancements. This study indicated that patients may benefit from being more informed about their transfusion, such as the risks of blood products used and being more involved in the decisionmaking process. If HCPs are more aware of patients' perceptions, they may seek to investigate whether patients hold common misperceptions and may, in some cases, challenge these. This may enable patients to cope better and support them with the decision-making process (e.g., risk vs. benefit evaluations), which may enhance treatment adherence (Horne & Weinman, 1999).
In conclusion, this study has identified a broad range of patients' and HCPs' transfusion perceptions revealing challenges for this context such as the burden of receiving repeated transfusion and transfusion-associated risks. In spite of this, the necessity of transfusion and the benefit of strong links between patients in haematology units and their HCPs were highlighted. HCPs also appeared motivated to improve their transfusion practice, which was reportedly stretched in terms of capacity (e.g., full transfusion appointment slots) with HCPs reporting a high level of goodwill to ensure patients' wellbeing. Service improvements, such re-designed patient consultations to enhance patient involvement in shared decision-making or patient education initiatives to are likely to enhance patients' experiences further.

AC K NOW L E DGE M E N T S
The authors thank participating sites, patients and healthcare professionals for their contribution to this research. Funding awarded from City, University of London Doctoral Studentship.

C ON F L IC T S OF I N T ER E S T
The authors declare no conflict of interest.

DATA AVA I L A BI L I T Y S TAT E M E N T
Transcripts are unable to be uploaded to a data sharing repository as informed consent for this purpose was not obtained from participants.

I N T ERV I E W TOPIC GU I DE S Patients
Thank you for agreeing to take part in the study. As described in the information sheet, I am interested to hear your views about blood transfusion, so please be as open with me as you wish to. Are you ok with me to start recording the conversation? If telephone participant: As this is a telephone interview, could you please confirm for the recording that you agree to take part in the study?
Ok, thank you… Are you ready to begin? This is the end of the interview, thank you for your time. Would you like to be sent a copy of the study results?
If yes: enquire how best to provide the patient with the results, collect details below.

Health professionals
Thank you for agreeing to take part in the study about your perceptions of blood transfusion. Some questions may be unfamiliar, for example, rather than looking to understand your content expertise, I am interested to more so understand your beliefs or thoughts, as a healthcare professional, about blood transfusion. What you share with me in the interview is confidential, so please be as open with me as you wish to be. Are you ok with me to start recording the conversation? If telephone participant: As this is a telephone interview, could you please confirm for the recording that you agree to take part in the study?
Ok, thank you… Are you ready to begin? Relief of symptoms such as tiredness 7 'when I received blood transfusion after that, I rest a bit and then I feel strong but before I feel weak' (Patient 10, other Anaemia, Site 2) Anticipated benefits 6 'I just feel as though that if I got a transfusion with fresh blood, surely I've got to feel better from it, I think so' (Patient 1, Myeloma, Site 1) (Continues) A P P E N D I X 2 (Continued)

Frequency (n patients) Example quotations
Patient convinced of benefit by significant others 3 'everyone was telling me I looked better and that in itself made me feel better as well I suppose' (Patient 1, Myeloma, Site 1) Can take time to feel benefit of transfusion 2 'this is a benefit, the blood, but sometimes it takes a while to kick in, it's like if you're taking medication from the doctor, sometimes it takes, erm, a while to kick in' (Patient 11, Lymphoma (CLL) and a second haematological condition, Site 2)

Necessity
Transfusion prolonging life and aiding survival 11 'oh yes, it just seems to me as being essential, erm, as much as say needing oxygen in the air is an essential, without the transfusions I wouldn't be here' (Patient 4, Aplastic Anaemia, Site 1) Transfusions required as a current and long-term supportive treatment 10 'they gave me two courses of ATG, it's an immunogloubulin … the second time, I didn't have any response.