FormalPara Key Summary Points

Every day matters to enable patients with acute myeloid leukemia (AML) to spend more time with family/friends and do hobbies, but not at the sacrifice of good health

Incorrect perceptions and limited understanding about treatment options may limit patients’ treatment acceptance resulting in worse outcomes

Patients consider physicians to have the most influence in treatment decision-making; however, priorities and treatment goals of patient and physicians do not always align


Acute myeloid leukemia (AML) is one of the more common blood cancers, characterized by peripheral blood and bone marrow myeloblast proliferation, leading to bone marrow failure and rapid death without treatment. AML primarily affects older adults, with a median age of almost 70 years [1, 2]. Previous research indicates that older patients with AML have substantial misperceptions about their prognosis and often feel distressed, overwhelmed, and uninformed about available treatments, yet are faced with the urgent need to make these potentially life defining decisions [3,4,5,6,7,8].

With the recent introduction of several new drugs [9], treatment decisions for patients with AML have become even more challenging, with difficult risk-to-benefit assessments to be made when deciding among multiagent intensive chemotherapy (IC), non-intensive chemotherapy (NIC) [3, 10] or best supportive care (BSC; e.g., blood product transfusions) alone [11,12,13]. Consideration must be given, for example, to likelihood of achieving complete remission (CR) and disease-free survival versus experiencing substantial toxicities or even early death. In recent years, it has been acknowledged that traditional survival endpoints in oncology clinical trials overlook disease- or treatment-related events that impact patients’ wellbeing. For this reason, health-related quality of life (HRQoL) has been recognized as an important endpoint and predictor of outcomes in AML [14], with worsening HRQoL potentially offsetting the value of prolonging survival. For example, one study found that, while patients with AML highly valued the chance at CR, they were willing to forgo small increases in the probability of remission to improve other outcomes, notably long-term side effects [15]. Other treatment factors, including treatment delivery and convenience, may also play a significant role.

However, for older patients with AML, there are other factors that contribute to a poorer prognosis including physical frailty, a higher incidence of comorbidities and adverse cytogenetic abnormalities, and an increased risk for treatment-related morbidity and mortality [16,17,18]. As such, for older patients with AML (particularly those who are over the age of 75 years and have significant comorbidities), IC is not considered a suitable treatment option [19]. Thus, for this population, treatment options are narrowed, adding further complexity to treatment decision-making. Despite this, there is very little published in the literature on the treatment decision-making process in older patients with AML who are unsuitable for IC, including the factors that influence decision-making, the trade-offs these patients are willing to make and patients’ expectations about treatment outcomes—particularly from a multi-stakeholder perspective. The current research aimed to substantiate and build on the current literature by exploring the value of extending survival in older patients with AML who are unwilling or unsuitable to receive IC and the associated complex, multifactorial decision-making process and inter-stakeholder dynamics from the perspective of patients, their relatives and physicians.


Study Design

This was a non-interventional, cross-sectional, qualitative interview study involving patients with AML as well as their relatives and independent physicians recruited in the United States (US), United Kingdom (UK) and Canada.

To support a patient-centered approach and ensure the perspectives of different stakeholders were considered, four physicians with expertise in AML and four patient advocates formed a steering committee (SC) to provide input and consultation at key stages throughout the research. Figure 1 presents an overview of the study design.

Fig. 1
figure 1

Overview of study design


It was planned that a total of 10 physicians (not part of the SC) and 25 patients (and their relatives) would be recruited to take part in this study.

Physicians specialized in treating adults with AML (> 10 cases per year, with ≥ 3 cases where the patient was unsuitable for IC) based in the US, Canada and UK were recruited via research recruitment agencies and were not physicians of the participating patients.

Patients and relatives were recruited as dyads to facilitate comparisons between patient and relatives’ perspectives. In the US, patients were recruited by a recruitment agency via referrals from physicians working in hematology and/or oncology (any site names/locations were blinded to the authors). In the UK and Canada, patients were recruited via physician referrals from The Russell Centre for Clinical Haematology at Nottingham University Hospitals NHS Trust (Nottingham, UK), Bristol Haematology & Oncology Centre at University Hospitals Bristol NHS Foundation Trust (Bristol, UK) and Princess Margaret Cancer Center (Toronto, Canada) facilitated by authors NHR and DM.

Eligible patients were ≥ 65 years old, had a physician-confirmed diagnosis of AML according to the World Health Organization (WHO) 2016 Classification [20] and were unsuitable or unwilling to receive IC. Suitability for IC was determined by the recruiting physician who considered factors such as poor performance status, significant comorbidities and adverse cytogenetics/molecular genetics. [21]

Relatives were referred by the patient and had to be ≥ 18 years old and have frequent contact with the patient.

A quota sampling approach was employed for the patient sample to ensure representation of key demographic and clinical characteristics [22], including age, time since diagnosis and treatment option. However, due to the rarity of the condition and how unwell patients were, these were target quotas only (i.e., flexible).


The study was approved by relevant ethical review agencies in the US, Canada and the UK. Written informed consent was obtained prior to the collection of any data.

Interview Process and Content

Each participant took part in a single, one-on-one, 60-minute interview conducted via telephone or video conference with a trained interviewer with extensive expertise in qualitative techniques (JS and LO). Interviews were conducted between April 2019 and December 2020.

A semi-structured interview guide, developed based on findings from a literature review [10] and with SC input, was used for each participant group to ensure all topics of interest were discussed. The interviews were conducted using a concept elicitation (CE) technique where open-ended, exploratory questions were asked to elicit spontaneous, unbiased responses about the participant’s experience, followed by more focused questions on topics that may not have naturally emerged during the interview [23, 24].

The patient interviews explored the AML journey and the value of living longer, specifically, key milestones that patients had achieved since diagnosis and hoped to achieve in future (e.g., family weddings or travel). The interviews also explored the treatment decision-making process and the role of the various stakeholders (e.g., how information was provided and what treatment options were discussed). The interviews with relatives explored similar topics but from the relatives’ perspective. The physician interviews covered these topics and also the clinical management of patients who were unsuitable for IC.

As part of the interview, all participants completed a quantitative questionnaire where they rated the importance of various factors in AML treatment decision-making from 0 (not important) to 3 (very important).

Data Analysis and Sample Size Considerations

In qualitative research, sample size is typically determined based on the goal to achieve ‘concept saturation’ (a point at which no new concepts are likely to emerge with further interviews). Research suggests that, in a relatively homogeneous population, approximately 85% of all concepts will be elicited after 10 interviews and > 90% of all concepts after 15 interviews; as such, a planned sample size consisting of 10 physicians and 25 patients/relatives was considered sufficient [25, 26].

The interviews were recorded and transcribed verbatim. Transcripts were analyzed via Thematic Analysis methods in ATLAS.ti [27] by authors JS and LO (and overseen by authors LM, CP and AG). Using an agreed code list to ensure consistency, participant quotes pertaining to the research objectives were assigned corresponding concept codes. New codes were also organically added, and previously analyzed transcripts were reanalyzed to ensure new codes were applied across all transcripts [28, 29]. Data were then converted into a filterable Microsoft Excel data extraction spreadsheet using the statistical software R/RStudio to facilitate calculation of discrete frequency counts of the number of participants reporting each concept code [30, 31].

While the analysis did not rely solely on frequency counts, they can be useful supplementary data, alongside descriptions of emergent themes and supportive patient quotes, to understand the relative importance and relevance of concepts. This information is particularly valuable when devising evidence-based recommendations for patient care [32].

Consolidated Criteria for Reporting Qualitative Research (COREQ)

The reporting of the study adhered to criteria in the COREQ qualitative research checklist [33].


Sample Characteristics

Twenty-eight patients with AML, 25 of their relatives and 10 physicians from the US, UK and Canada participated in final interviews. The demographic and clinical characteristics of the patient sample are summarized in Table 1. Due to recruitment challenges, patients on the various treatment options were not proportionately represented across countries, with most of the US patients having untreated AML while all UK and Canadian patients were currently or had previously been treated with active therapy. Characteristics of participating relatives are summarized in supplemental data Table 1. Characteristics of the physicians who took part in the interviews are described in Table 2. On average, physicians had experience of treating patients with leukemia for 17.3 years and were seeing 22 patients with AML per month, with on average 10 of these patients considered not to be suitable to receive IC, as determined by the physician.

Table 1  Demographic and clinical characteristics of patient sample, broken down by country (n = 28)
Table 2 Physician characteristics (n = 10)

Interview Findings

To align with the two key topics explored during the interviews, the results have been presented in two key sections: the value of extended life and the treatment decision-making process. Within each section, the keytheme(s) have been pulled out and presented as sub-sections.

Value of Extended Life

Every Day Matters to Enable Patients to Spend More Time with Family/Friends and Do Hobbies

During the interviews, all patients and relatives were asked about life milestones (goals, events or activities) they/their family member had achieved since diagnosis or hoped to achieve in the future. The life milestones discussed pertained to five high-level domains of HRQoL, which broadly align with categories of life goals described by Pinquart et al. [34] most notably: social goals (focusing on maintaining or enhancing social relationships), leisure goals (focusing on meaningful and self-rewarding activities) and health-related goals (focusing on maintaining or improving physical health). Other domains included activities of daily living (ADL) and important dates.

As illustrated in Fig. 2, almost all patients and relatives mentioned social goals when discussing life milestones that they/their family member hoped to achieve (patients n = 27/28, 96.4%; relatives n = 23/25, 92.0%), indicating their importance. Patients and relatives described wanting to spend time with family/friends, while others discussed wanting to do particular activities with family, such as meals out or at home, or travel/vacations. Example quotes in support of the key findings are summarized in Table 3.

Fig. 2
figure 2

Overview of future life milestones

Table 3 Key findings and supporting quotes

Patients and relatives also discussed leisure goals they/their family member hoped to achieve in the future (patients n = 24/28, 85.7%; relatives n = 21/25, 84.0%). This included going on walks, watching TV, reading, games and crafts, going to the cinema, gardening, doing some sport/exercise, travel/vacations and listening to music.

To a lesser extent, patients and relatives also described hoping to achieve health-related goals (e.g., better health, longer survival), maintain activities of daily living (e.g., daily chores, housework) and reach important dates (e.g., family occasions). Patients described being limited in some activities because of poor health and/or coronavirus disease 2019 (COVID-19) pandemic restrictions.

These responses resonated with the physicians, who described patients wanting to achieve important dates (n = 7/10, 70%), such as family occasions and public holidays, mostly to spend more time with family (n = 5/10, 50%).

Patients’ Outlook and Expectations About the Future May Be Impacted by Treatment Experience

During the interviews, patients and their relatives had mixed expectations for the future. Patients and relatives had negative expectations and worries about their/their relative with AML’s physical health including deteriorating health, treatment not working or even imminent death, and this remained a key concern over time. However, there were positive expectations related to time being spent with family, with patients and relatives both describing upcoming family events and visiting family members who may have lived further away.

Among patients who had no prior exposure to AML treatment, outlook for the future became more negative over time. Few patients (n = 4/13, 30.8%) recalled experiencing negative expectations about the future at diagnosis, while most (n = 10/13, 76.9%) did express negative outlooks at the time of the interview. Conversely, patients with AML treatment experience appeared to have an increasingly positive outlook. Most of these patients (n = 11/14, 78.6%) recalled feeling negative about the future at diagnosis, whereas at the time of the interview, most (n = 9/14, 64.3%) felt hopeful about future events.

Furthermore, among those with treatment experience, positive expectations were more frequently reported among those diagnosed ≥ 6 months prior to interview (n = 6/7, 85.7%) compared to those diagnosed < 6 months prior to interview (n = 3/7, 42.8%). This was also supported by findings from physician interviews, where seven (n = 7/9 asked, 77.8%) physicians noted that patients’ feelings changed over time depending on treatment success.

Treatment Decision-Making Process

The Shock of Diagnosis May Impact Patients’ Understanding of the Various Treatment Options

Most patients discussed their emotional reactions when first diagnosed, which most commonly included feeling shocked (n = 19), concerned/worried (n = 13) and overwhelmed (n = 10). All physicians (n = 10) reported discussing NIC treatments with patients unsuitable for IC. While patients confirmed that they discussed treatment with their physician, most patients (n = 15/28, 53.6%) did not articulate a difference between IC, NIC and BSC during the interview, but were not directly asked if they understood the distinctions. The remaining patients (n = 13/28, 46.4%) were directly asked whether they had been made aware of the difference between various treatment options (e.g., IC vs NIC, NIC vs BSC). Most patients (n = 9/13, 69.2%) explained that they had been provided with some materials, such as leaflets, which provided some treatment-related information. However, despite this, understanding of treatment options was limited. The remaining four patients (n = 4/13, 30.8%) reported personally seeking out treatment information.

Extending Survival in Good Health is More Important Than Extending Survival Alone

Figure 3 presents the proportion (%) of patients, relatives and physicians who rated each factor as ‘very important’ in the quantitative questionnaire. Overall, improved HRQoL (patients: 92.6%, relatives: 96.0%, physicians: 70.0%), extended survival (patients: 89.3%, relatives: 88.0%, physicians: 80.0%) and relief of symptoms (patients: 88.9%, relatives: 88.0%, physicians: 60.0%) were most commonly rated as ‘very important’ in AML treatment decision-making. However, patients and their relatives further described certain trade-offs they made or would be willing to make when deciding whether to pursue treatment. Overall, patients across all treatment options (n = 19/28, 67.9%) and their relatives (n = 15/25, 60.0%) most frequently reported that QoL was more important than extended survival.

Fig. 3
figure 3

Treatment factors rated ‘very important’ by patients, relatives and physicians when making AML treatment decisions

Conversely, a smaller proportion of patients reported that they would be willing to tolerate side effects if survival could be extended (n = 7/28, 25.0%), though most of these patients were already on NIC (n = 5/7, 71.4%). The perceptions and reality of side effects are discussed below.

Incorrect Perceptions of Treatment Side Effects May Limit Treatment Acceptance and Result in Worse Outcomes

As detailed in Fig. 3, side effects were rated as ‘very important’ by a greater proportion of patients (n = 15/25 who answered, 60.0%) and their relatives (n = 16/25, 64.0%) compared to physicians (n = 4/10, 40.0%). Furthermore, side effects were rated as ‘very important’ by a greater proportion of patients not on treatment (n = 9/13, 69.2%) and their relatives (n = 12/13, 92.3%) compared to those with experience of NIC (n = 5/11 who answered, 45.5%) and their relatives (n = 3/11, 27.3%).

Patients with no treatment experience (n = 9/13, 69.2%) and their relatives (n = 7/13, 53.8%) reported a fear of side effects being the primary reason for opting not to pursue AML treatment. Specifically, many of these patients (n = 6/13, 46.2%) were worried that side effects would be worse than their AML symptoms. Although it was not clear whether patients were distinguishing between the side effects of IC and NIC, patients generally perceived side effects as severe (n = 9/13, 69.2%) and mentioned nausea/vomiting, pain, hair loss, multiple organ failure and death. Four patients further described being influenced by seeing or hearing about other cancer patients who suffered from side effects.

Although two patients’ experiences of side effects resulted in them discontinuing NIC (n = 2/14, 14.3%), most patients with experience of NIC (n = 9/14, 64.3%) reported a small number of side effects that had little impact on their life, including fatigue, reduced appetite, generally feeling unwell, nausea and injection site irritation. Nausea (n = 5) and risk of infection (n = 3) were the treatment side effects most frequently reported by physicians.

Patient and Physician Priorities Are Not Always the Same, Potentially Resulting in a Misalignment of Treatment Goals

There was evidence of other differences in priorities between physicians and patients and their relatives such as the possibility of being hospitalized, risk of infection and location of care (Fig. 3).

The possibility of being hospitalized was rated as ‘very important’ by a greater proportion of patients (n = 14/27 who answered, 51.9%) and relatives (n = 16/25, 64.0%)%) than physicians (n = 1/10, 10.0%). Most patients and relatives described how hospitalization was to be avoided because it meant being away from home or because it indicated being seriously ill. Others considered it beneficial because it afforded access to medical care. In contrast, physicians tended to view hospitalization as a standard and necessary part of treatment.

The risk of infection was also rated as ‘very important’ by a greater proportion of patients (n = 14/26 who answered, 53.8%) and relatives (n = 19/24, 79.2%), compared to physicians (n = 3/10, 30.0%). Patients and relatives commented on susceptibility to infection further impacting quality of life. In contrast, most physicians explained that infections are usually caused by the disease, rather than treatment, and can be managed.

Finally, location of care was rated as ‘very important’ by a greater proportion of patients (n = 13/26 who answered, 50.0%) and their relatives (n = 13/24, 54.2%) compared to physicians (n = 2/10, 20.0%) because of the desirability of having easily accessible treatment closer to home or at home.

Patients Consider Physicians to Have the Most Influence in Treatment Decision-Making

During the interview, all participants (patients, their relatives and physicians) discussed a shared treatment decision-making process involving the physician(s)/hematologist(s), the patient, family members (including spouses, children or other family members) as well as other healthcare professionals (nurses, pharmacists and support workers). Participants also discussed who had the most influence on treatment decision-making (Fig. 4). All physicians asked (n = 6/6, 100.0%) and most patients asked (n = 20/23, 74.1%) stated that it was the physician who had the most influence due to their knowledge and experience, while most of the relatives asked (n = 15/23, 65.2%) thought it was the patient. A small proportion of patients (n = 2/27, 7.4%) said that the decision was joint between them and a family member.

Fig. 4
figure 4

Who had the most influence in treatment decision-making?


The overall objective of this study was to characterize the experiences of patients with AML who were unwilling or unsuitable for IC and to better understand the treatment decision-making process from the perspective of patients, their relatives and physicians and how this links to the perceived value of extended survival. The key findings from this study were that:

  • Every day mattered to enable patients with AML to spend more time with family/friends and do hobbies, but not at the sacrifice of good health;

  • Patients’ incorrect perceptions and limited understanding about treatment options may limit treatment acceptance and result in worse outcomes;

  • Patients consider physicians to have the most influence in treatment decision-making; however, patient and physicians’ priorities and treatment goals do not always align.

First, the findings of this study showed that extending survival in good health was valuable for both patients and their family, particularly in terms of the opportunity to spend more time together, ability to do daily activities and attain meaningful goals and milestones. These findings are supported by previous research where cancer patients discussed the importance of spending time with family [34, 35] and maintaining a ‘normal life,’ which included engaging in hobbies and remaining active [36]. As a novel finding, this study indicated that patients and physicians’ priorities did not always align when making treatment decisions, for example, regarding the perceived importance of the possibility of being hospitalized and the risk of infection. These findings have an important implication for the AML treatment decision-making process, highlighting the need for structured patient-physician discussions to allow the physician to align recommended treatments and plans with the patient/family’s goals and expectations. Tools to facilitate this could include discussion guides for both the physician and the patient, which may include topics and questions to cover (e.g., goals and important considerations) and techniques to enhance engagement of the patient.

Second, this study illustrated that older AML patients had limited understanding of the different treatment options. No standard treatment approach exists for older patients with AML; those who are unsuitable for IC are typically offered NIC or BSC [37, 38]. Despite physicians reporting that they discussed various treatment options with their patients, most patients in this study did not articulate or had limited understanding about the difference between IC, NIC and BSC. Patients noted the overwhelming shock upon diagnosis [39,40,41], which can make it difficult to process complex information about treatment options [5, 41] and to feel involved in treatment decision-making [42]. Several patients also described inflated concerns regarding the treatment side effects based on their experiences with family/friends on other chemotherapies. Since NIC in AML is less intensive than many typical solid tumor chemotherapy regimens, it is important to recognize that patients with AML may have pre-conceived perceptions about the side effects of NIC that could limit treatment acceptance[10] and lead to worse outcomes. Physicians should carefully consider how information is delivered, explicitly assessing patient and family understanding and expectations with attention to correct false assumptions. The highlighted lack of patient understanding also calls for the need for more accessible, patient-friendly resources, such as videos, blogs/vlogs and brochures. For example, education videos and vlogs can be effective and powerful tools for patients by facilitating knowledge and providing accounts of the “real lived” experience that can mitigate misconceptions (e.g about different treatment pathways) and provide useful tips to enhance coping skills and self-care. Ensuring the resources are patient-centered and involve input from patients is critical. Given the influence of other patients' experiences, it might also be useful to provide opportunities for peer support and education between patients who have had a similar diagnosis or are on the same treatment path.

Finally, this study illustrated that involvement during treatment decision-making is further complicated by who the different stakeholders perceived to have the most influence in the decision. There was consensus among physicians and patients that the physician had the most influence, demonstrating the importance of the physician acting as a facilitator to engage the patient. This is especially vital in the current landscape of increasing numbers of available therapies and the patients’ limited understanding of the treatment options available to them. By actively creating space for the patient to voice their own concerns and questions, the physician can ensure that the process achieves the oncology gold standard of shared decision-making.

While the qualitative study design of this research provided considerable depth of insight regarding varied stakeholder perspectives of AML and the treatment decision-making process, it is necessary to acknowledge some limitations. First, while the overall sample size was considered sufficient for qualitative research, it is important to acknowledge that the data are based on a small number of participants, therefore limiting the generalizability of the findings. Additionally, while the overall sample was demographically and clinically diverse, samples within each country were less so. There was no representation of black African or multi-racial ethnicities in the UK or Canadian samples, education levels were higher in the UK and Canadian samples, and all Canadian patients were male. In terms of treatment paths, there was just one patient on BSC and patients with untreated AML were represented in the US sample only. Of note, the low treatment uptake in the US sample is supported by the Surveillance Epidemiology and End Results (SEER)-Medicare database, which reported that approximately 50% of older US AML patients remain untreated for reasons that are not well understood [43]. It is also important to acknowledge that the access and funding of healthcare are different in the US compared to the UK and Canada, and this may have influenced patients’ treatment decision-making. Given some of the other differences in the US, UK and Canadian samples (including representation of patients on different treatment options), no comparisons were made at the country level. Finally, no information about patients’ remission status was collected, and recruitment for the study was also prolonged by the global COVID-19 pandemic, although, the data collected from interviews was rich and descriptive and allowed for the exploration of patients on different treatment paths. Patients recruited and interviewed during the COVID-19 pandemic were also able to effectively differentiate between any limitations to their HRQoL as and when this was due to their AML experience or to COVID-19.


In conclusion, this qualitative study illustrates the complexities associated with treatment decision-making in older patients with AML who are unsuitable for IC and how patient priorities (such as every day HRQoL) do not always align with physicians’ priorities. The misalignment in stakeholder priorities and the perceived weighty influence of physicians must encourage physicians to strive for structured discussions about diagnosis and treatment with their patients, allowing them the space and opportunity to be informed and to discuss their goals and concerns surrounding receiving treatment at an emotionally challenging time. Additionally, future research efforts should focus on the development of patient-centric resources to ensure patients and their families are supported to make fully informed decisions.