Perceived risk of death among patients with advanced cancer: a qualitative directed content analysis

Abstract

Background

Risk perception with respect to death is a prerequisite for patients with advanced cancer when the time comes to make medical decisions. However, the nature of death risk perception remains unclear.

Method

In-depth interviews were conducted with 28 patients with advanced cancer who were recruited from two hospitals and one home-based hospice in Fujian, China. Interviews were transcribed and directed content analysis applied. The Tripartite Model of Risk Perception was used as a theoretical framework.

Results

Patients with advanced cancer perceived their risk of death in different ways. Professional communication about death risk and data-driven risk perception were common in clinical settings. Affective influences, inherent cognition, and comparisons to others or oneself also contributed to the subjects’ self-perceived death risk.

Conclusion

This theory-informed qualitative study clarifies the nature of the perceived risk of death among patients with advanced cancer. The study findings offer healthcare providers a more nuanced understanding of the perceived risk of death among patients with advanced cancer.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12904-024-01584-3.

Background

Risk perception is a subjective judgement about the likelihood of a negative occurrence. It has become a central construct in numerous frameworks designed to predict engagement in health-related behavior [1, 2]. It has also been used to assess the probability of health-related occurrences such as injury, disease, and death [3]. Previous studies have revealed that death risk perception is a prerequisite for patients with advanced cancer who need to make informed medical decisions and prepare for death [4, 5]. Therefore, it is essential for healthcare providers to understand patients’ self-perceived death risk.

Bias in death risk perception is common among advanced cancer patients. Based on an integrative literature review, the accuracy of the perceived risk of death among such patients ranges from 5–71% [6]. Underestimation often leads to unrealistic expectations and hope in these patients, causing them to opt for aggressive treatment. Such decisions can be financially straining and waste limited medical resources and prolong the painful death process [6, 7]. In contrast, overestimation of death risk may have adverse mental impacts such as death anxiety and depression and result in a failure to reflect on death and establish reasonable coping behaviors, consequently lowering patients’ quality of life [8]. However, despite its importance, the death risk perception of patients with advanced cancer remains poorly documented.

Previous studies on perceptions of the risk of death have emphasized objective disease information such as the estimation of recovery chance, therapy goals, and the course of the disease among patients with advanced cancer [9–11]. Several studies have argued that it may not be sufficient to understand the death risk perception of patients with advanced cancer, if doing so is limited to objective medical information [8, 12]. Derry et al. [13] further found that negative psychological factors such as stress and extreme emotions play an important role in the perceived risk of death among this category of patients. Thus, it is important to explore death risk perception as expressed by the patients themselves.

According to risk perception theory, human beings comprehend risk in two fundamental ways: analytic and experiential systems [14]. Ferrer’s tripartite model of risk perception [15] (TRIRISK), which was developed and validated for a US population, proposed that risk perception is composed of three dimensions: deliberative, affective, and experiential components. Deliberative risk perception is systematic, reason-based, and rational. Affective risk perception refers to the valence and associated arousal of affective responses to risk events. Experiential risk perception is conceptualized as heuristic risk judgements or gut-level feelings of vulnerability to a threat resulting from previously acquired or learned associations. The tripartite model has been used in quantitative research to predict disease-related protective motivation or actions resulting from cancer, heart disease, and diabetes [15, 16]. For example, Riedinger et al. [1] employed an 18-item TRIRISK instrument of risk perception for a UK sample to examine the predictive utility of cancer risk perception with regards to intention. Although the TRIRISK model has not been utilized in qualitative research, it likely offers a suitable avenue to explore how patients with advanced cancer perceive their risk of death. Therefore, we employed the TRIRISK model to explore the experiences of patients with advanced cancer in perceiving the risk of death, to aid healthcare providers in gaining a more accurate understanding of how patients acquire their perception of their death risk.

Methods

Design

This study was qualitative and based on a theoretically informed survey. The TRIRISK model was used as the study theory, which guided data collection and analysis. Data on the experiences of death risk perception among patients with advanced cancer were analyzed using directed content analysis, a method employed to validate or extend a theory or framework [17].

Study setting and participants

Participants were recruited through clinical care teams in an oncology department of a tertiary general hospital, cancer hospital, and home-based hospice in Fujian, in southeast China. A purposive sampling strategy was used to recruit patients of a variety of ages, genders, levels of education, marital statuses, and disease courses. The inclusion criteria were as follows: (a) diagnosed with advanced cancer at Stages III or IV, (b) aged ≥ 18 years, and (c) aware of their diagnosis, disease condition, and treatment. The exclusion criteria were as follows: (a) severe disability or critical illness (Karnofsky performance status < 20%) and (b) verbal communication or cognitive impairment (score on Short Portable Mental State Questionnaire < 2).

Data collection

Semi-structured interviews were conducted and audio-recorded by the interviewer (CGJ) from March through June of 2022. The interviewer spent three months in the study setting as an intern in order to become familiar with the study population. The interview questions were constructed based on the TRIRISK model. Due to the sensitivity of the interview topic, the interviews began with an open-ended question about the experience of being ill in order to build a trust relationship and provide an opportunity to further discuss experiences related to death risk perception. Then, the interviewer explored participants’ specific experiences related to deliberative, affective, and experiential death risk perception, based on the TRIRISK (Supplementary File 1). During the interviews, probing questions were used to further explore their connotations of death risk perception, such as “Could you explain more about that?” and “When you say …, what do you mean?”. According to Francis’s recommendation, when three consecutive interviews do not yield additional information, data saturation have been achieved [18]. Each interview lasted 30 to 90 min. Interviews were conducted at participants’ homes if they were receiving home-based hospice care, or in a private room in the hospital if they were hospitalized.

Data analysis

All interviews were transcribed verbatim within 24 h of completion. Transcripts were checked against the audio files for accuracy and imported into qualitative data management software (NVivo 12.0). A directed content analysis was performed to deductively analyze the transcripts for dimensions of the theoretical model. Then, through iterative discussion, emergent categories were inductively explored [17]. Two authors (CGJ and CZX) were well-trained in qualitative research methods prior to this study, and independently coded each transcript. After coding the first transcript, they developed a codebook through iterative discussion and consensus-building to ensure consistency of terms. Data analysis and subsequent interviews were conducted concurrently to enable the research team to explore emergent categories. Discrepancies were resolved by consulting another author (XHM), a senior qualitative researcher in oncology nursing. The original language of the transcripts was Chinese. Excerpts presented here were translated into English by CGJ and revised by XHM.

Ethical considerations

The present study was approved by the Ethics Committee of Fujian Medical University of Biomedical Research (No. FMU202272). All interviews began after informed consent was obtained from the participants. Participants were free to withdraw from the interviews at any time. If they demonstrated negative emotional reactions such as crying or sobbing, the researcher paused the interview and provided emotional support. In addition, a protocol was developed to address participants’ strong adverse psychological reactions, if necessary. All information related to the participants’ privacy was anonymized and kept confidential.

Results

Thirty patients participated in the study. One withdrew from the study because he avoided discussing death-related topics, and another stated that he had no risk of death. Ultimately, 28 participants were included in the analysis. Among them, 13 were males and 15 were females, 21 were under 60 years old, 15 had a high school education or lower, and seven held a bachelor’s degree. The details about the demographic characteristics of these participants are presented in Table 1. This study revealed five categories: professional communication about death risk, data-driven risk perception, affective influences, inherent cognition, and comparisons to others or oneself.

Table 1.

Demographic characteristics of participants (n = 28)

No.	Gender	Age	Education	Marital status	Occupational status	Economic status	Cancer diagnosis	Course of disease (month)	Type of treatment (now)
P1	Female	41	bachelor’s degree	married	bank clerk	good	lung cancer (IV)	87	palliative treatment
P2	Male	44	middle school	married	worker	good	liver cancer (III)	18	interventional therapy
P3	Female	66	associate’s degree	married	retired	good	breast cancer (IV)	216	chemotherapy
P4	Male	49	primary school	married	farmer	average	liver cancer (III)	6	interventional therapy
P5	Female	57	associate’s degree	married	retired	good	colon cancer (IV)	13	chemotherapy
P6	Female	36	high school	married	housewife	good	breast cancer (III)	14	chemotherapy
P7	Male	70	high school	divorced	retired	average	prostate cancer (III)	30	palliative treatment
P8	Female	79	bachelor’s degree	married	teacher	good	colon cancer (III)	11	palliative treatment
P9	Female	30	bachelor’s degree	married	teacher	average	pelvic mesenchymal malignancy (III)	14 days	awaiting interventional therapy
P10	Female	35	associate’s degree	married	company staff	average	liver cancer (IV)	12	interventional therapy
P11	Female	60	associate’s degree	widowed	employee of public institution	average	breast cancer (IV)	59	palliative treatment
P12	Female	57	middle school	married	worker	poor	breast cancer (III)	10	chemotherapy
P13	Male	87	bachelor’s degree	married	retired	good	bladder cancer (III)	8	palliative treatment
P14	Male	58	middle school	married	worker	poor	liver cancer (III)	21	immunotherapy & interventional therapy
P15	Male	40	middle school	married	sole trader	average	liver cancer (III)	25	immunotherapy
P16	Female	53	bachelor’s degree	married	employee of public institution	average	breast cancer (III)	10	chemotherapy
P17	Male	44	bachelor’s degree	married	teacher	average	lymphoma (IV)	54	chemotherapy
P18	Male	69	bachelor’s degree	married	civil servant	good	rectal cancer (III)	36	palliative treatment
P19	Male	54	primary school	married	sole trader	average	liver cancer (III)	6	interventional therapy
P20	Female	73	middle school	married	worker	poor	pancreatic cancer (IV)	2	palliative treatment
P21	Male	55	middle school	married	fisherman	average	liver cancer (IV)	23	immunotherapy & targeted therapy
P22	Male	59	high school	married	salesperson	average	esophageal cancer (IV)	25	palliative treatment
P23	Male	58	high school	married	worker	average	rectal cancer (IV)	13	interventional therapy
P24	Female	37	associate’s degree	married	company staff	average	breast cancer (III)	2	chemotherapy
P25	Female	51	associate’s degree	married	company manager	good	rectal cancer (IV)	37	interventional therapy
P26	Female	52	primary school	married	housewife	average	breast cancer (IV)	5 days	awaiting chemotherapy
P27	Female	56	primary school	married	farmer	poor	ovarian cancer (IV)	66	chemotherapy
P28	Male	64	high school	married	retired	average	liver cancer (IV)	107	targeted & interventional therapy

Categories

Professional communication about death risk

Professional communication was an essential domain of deliberative risk perception regarding death. Most patients stated that they perceived their own death risk based on information obtained from medical staff, especially doctors. “The doctor said that my cancer type was the least bad of all the kinds of breast cancers. He comforted me and told me not to worry, even though my cancer was in the middle-late stage. So, I felt that death was not so near for me” (Participant 12). “The doctor told me that I have a three-to-five-year life expectancy based on my current situation. So, I don’t need to worry about death at present” (Participant 7).

Patients with advanced cancer confessed that they would speculate about the condition of their disease, according to the treatment purpose stated by the doctor. “The doctor told me that this treatment [interventional embolization] could control the growth of the tumor. If the disease was under control, it would be like a chronic disease! Cancer can still be controlled, albeit gradually” (Participant 14).

In the interviews, almost every patient mentioned that the effect of the treatment prescribed by their doctors was their primary concern. Patients believed that the effect of treatment was directly related to the risk of death, and thus used it as an important clue to assess their risk of death. “After the third round of therapy, I consulted my doctor about my disease. He said the tumor had shrunk and the treatment was very effective. And I really feel relief” (Participant 19). “The treatment period is long, but the doctor told me that the treatment’s effect is not valid. I feel like death may be drawing near to me!” [crying] (Participant 10).

Data-driven risk perception

Data-driven risk perception was another major aspect of deliberative risk perception that was expressed frequently during the interviews. Certain patients admitted that they based their perceived risk of death on data in objective medical examination reports from authoritative hospitals. They claimed that the data were objective and scientific in term of words, numbers, and medical images. “The previous report showed that my cancer was only 1 cm by 1 cm, but now it has grown to 6 cm. Alas! [sigh] It is growing very fast! It does scare me to death!” (Participant 4). Patients noted that imaging reports such as CT and MRI scans relevant to their cancer allowed them to visually understand their disease and risk of death. According to Participant 21, “The CT and MRI scans showed that a dozen points [tumors] were almost gone. It is good news for me because death has once again gone far away from me!”

Affective influences

On the TRIRISK, affective influences are the second element of perceiving death risk. In the present study, patients expressed three types of emotional responses to death/cancer, including optimism, neutrality, and negativity. Some patients believed that with an optimistic outlook and unwavering commitment to their treatment regimen, they could overcome the disease and live as long as their healthy peers. Such patients often expressed less perceived risk of death. “To be strong, do not be pessimistic! I always tell myself I can overcome it. No worry about the disease. It will be under control. Death is nothing. I don’t need to think about it” (Participant 25).

Patients who believed that life, aging, illness, and death were natural laws explained that they did not fear death. “We comply with our normal fate; we live the way life is. I will go to the crematorium when my time comes. I think it is okay to let nature take its course” (Participant 28). These patients spoke relatively calmly about their own death.

In contrast, those patients with negative affective reactions such as fear or avoidance confided that the fear of death had followed them like a shadow since the cancer was diagnosed. “Since I was diagnosed with breast cancer, I have been scared and terrified. I keep thinking that death is going to knock at the door” (Participant 24). Some people stated that they avoided thinking about death and discussing it with others, and instead focused on their cancer treatment. However, anything related to disease or death would cause them to perceive their death risk. “I try not to think about death. I comply with the doctors’ orders, and actively treat it. I always ponder how to cure the disease. But, you know, it does not work sometimes” (Participant 1).

Inherent cognition

The term “inherent” highlights the basic or permanent attributes of individuals as shaped by their past experiences. In the interviews, patients had a variety of viewpoints on contemporary medical technology, doctors’ professional skill, and their disease/treatment, all of which impacted their perception of death risk.

Some patients believed that once they cooperated with their doctors, the disease would be controlled and their lives prolonged. “I believe in the development of science and medicine now! Just follow the doctor’s instructions. Cancer will be cured. I am sure my life is guaranteed!” (Participant 3). However, some patients doubted the medical ability of contemporary doctors. They thought that the treatments might not prolong their lives and over-treatment might increase their risk of death. “Seriously, the hospital’s medications are all excessive. Originally, one round of chemo was enough, but they told me, they said they’ll give me two or three rounds. Patients simply can’t bear it! The hospital’s medications will only make me die faster. Taking those drugs, I’ll die more miserably. If he [the doctor] makes that decision, I’ll be gone sooner” (Participant 22).

Patients with different levels of education and types of life experience had varying levels of understanding of the disease and treatment. Certain patients supposed that cancer meant waiting for death. “Cancer! [laughs] It usually means close to death” (Participant 18). Other patients felt that cancer was not too difficult to control. “In fact, I have always felt that I am the same as common sick people. Get sick, get treatment! Nothing to worry about” (Participant 15).

Treatment plans could also affect patients’ perceived risk of death. Certain patients believed that if a cancer operation was accessible, the disease was not severe and the threat to life low. “I cannot undergo surgery now, which means my disease is relatively serious and death may well be imminent” (Participant 26). Conversely, if only traditional Chinese medicine was prescribed for their disease, they felt that the disease was incurable and the threat to life high. “If you have cancer and only take Chinese traditional medicine, it means you have no alternative. In about eight months, you will pass away” (Participant 11).

Comparisons to others or oneself

Comparisons to others or oneself were also important components of the perceived risk of death. Some patients compared their disease condition and treatment to those of other cancer patients or to their previous physical condition. Such comparisons helped form their perception of death risk.

Some patients felt that they were at imminent risk of death when they experienced the same physical symptoms as dying patients. “In our village, if your feet swell, you are done! Look at my feet. They are so swollen now, and I think my time is coming” (Participant 27).

Some patients felt that their current disease condition was worse than that of other patients; therefore, they assumed that their survival time would be shorter than that of others. “I know some patients with cancers that are less severe than mine. They tried chemotherapy and hoped to live longer, but they still passed away soon afterward. My disease is more complicated. I need not be in the hospital because I have only a few days to live!” (Participant 20). In addition, patients who believed they were receiving advanced medical treatment were less likely to judge the disease as life-threatening. “My aunt was diagnosed with the same disease I have. She has lived for a long time. Now, I am receiving better healthcare than her. No worries!” (Participant 5). In addition to comparisons with others, a patient’s physical change over time could inspire their perception of the risk of death. “Before I got sick, I could walk for a long time and did not feel uncomfortable. Now I often feel tired. My health is deteriorating…” [crying] (Participant 2).

Discussion

This qualitative study explored the experiences of patients with advanced cancer with regards to their perceived risk of death. We found that professional communication was the main factor affecting a patient’s perception of death risk. Objective medical information, affective influences, inherent cognition, and comparisons to others or oneself also played an important role in formulating the perception of death risk among patients with advanced cancer.

According to the TRIRISK model, deliberative risk perception, which is based on the understanding of traditional logic and rules of evidence, is the main predictor of preventive health behavior [15, 19]. The subjects in the present study believed that the information about death risk obtained from health professionals and medical data was objective and deliberative. Thus, professional communication is key to a patient’s perception of their risk of death. Previous studies have supported the notion that doctors’ disclosure of disease information such as prognosis and death risk estimation helped advanced cancer patients accurately perceive their own risk of death [20, 21].

In our interviews, patients often expressed affective influences on their perception of their own risk of death. These findings are consistent with the affective risk perception dimension of the TRIRISK model. When individuals respond to risk events with important affective significance, strong affection interferes with their rational thinking about risk perception [22]. In this study, patients with negative affective reactions often expressed fear or avoidance of death. Patients who were afraid to die expressed that they were living under the constant threat of the risk of death. However, patients with avoidant attitudes emphasized that they were dedicated to treatment and willing to go to any lengths to delay or avoid death. As one patient said, “hold on to the last straw.” This result is consistent with those of previous studies. For example, Bergqvist and Strang [8] found that patients with advanced breast cancer failed to admit a high risk of death, even if they realized the terminal and incurable nature of their disease. A longitudinal study confirmed that patients facing imminent death had strong affective reactions to death risk events (e.g., outright rejection) that affected their perception of a poor prognosis [23]. However, in our interviews, certain patients showed natural/positive affective reactions to death and were willing to face their situation and talk about death. They calmly discussed treatment plans, end-of-life arrangements, and other death preparations. This is important for professional communication on death risk and prognosis. Improving patients’ emotional state could be a way to enhance the effectiveness and accuracy of communication.

Additionally, our study revealed that patients with different past experiences varied in their inherent cognition, which was mainly reflected in their trust in contemporary medical technology and the professional level of their doctors. This is consistent with the experiential dimension of the TRIRISK model, which reflects the consequences of previously acquired or learned associations¹¹. Alhakami and Slovic [24] suggested that individuals’ assessments of activities or technology are based on their emotions toward them. For example, if people have a positive emotional perception of an activity or technology (e.g., trust), they tend to assess the risk of the activity or technology as low and the return high. This phenomenon was also confirmed in the present study. Patients who exhibited a firm sense of trust in contemporary medical technology and the professional level of their doctors were more willing to accept aggressive treatment. They believed that if they received medical treatment, death would be postponed.

Interestingly, we found another unique aspect of perception of death risk among patients with advanced cancer that is not included in the TRIRISK model. The model emphasizes directly evaluating risk from the individual’s perspective and risk event itself. However, in our interviews, patients with advanced cancer indirectly perceived their own risk of death by comparing themselves to their patient peers or past selves. This phenomenon has also been reported in the relevant research. A questionnaire survey of breast cancer patients found that patients used social comparisons to indirectly judge their disease condition, physical condition, and risk of death level [25]. In addition, a phenomenological study revealed that patients formed individual perceptions of risk of death by comparing their current physical condition with their previous one [26].

The findings of this study will inform health providers caring for patients with advanced cancer. The results offer valuable insights to help providers understand patients’ perceptions of death risk. For those with deliberative risk perception, health providers should focus on objective disease information to enhance patients’ understanding of their risk of death. For patients whose perceptions of death risk are influenced by emotions, psychological support should be prioritized.

Limitations

This study had some limitations. First, the patients with advanced cancer who served as subjects for this study were only interviewed once. Further research should adopt a longitudinal design with multiple interviews to provide a more detailed description of death risk perception among patients with advanced cancer. Second, our study data were generated from two hospitals and one home-based hospice in China, which may limit its transferability to other countries, although the theory-based analysis moderates this limitation. Additionally, due to the sensitivity of the topic, some patients were unwilling to participate in the research, which could potentially threaten the applicability of the results of this study.

Conclusion

This TRIRISK model-informed qualitative study elucidates how patients with advanced cancer perceive their risk of death is conditioned by professional communication about death risk, data-driven risk perception, affective influences, inherent cognition, and comparisons to others or oneself. The findings not only further validate the TRIRISK model, but also create a conceptual extension of death risk perception in palliative care, providing a reference for the development of a death risk-related instrument and interventions.

Electronic supplementary material

Below is the link to the electronic supplementary material.

Author contributions

Conception and design of the study: HMX; Data acquisition: GJC, ZXC, JWZ, SWY, HW; Analysis and interpretation of data: GJC, ZXC, JWZ, SWY, HW; Drafting the article or revising it critically for important intellectual content: GJC, HMX.

Funding

This study was funded by the Natural Science Foundation of Fujian Province (Grant no. 2023J01665).

Data availability

The data are available on request from the corresponding author.

Declarations

Ethics approval and consent to participate

The study was approved by the Ethics Committee of Fujian Medical University of Biomedical Research (No. FMU202272). All methods were carried out in accordance with relevant guidelines and regulations of the Declaration of Helsinki. The purpose of the interview was explained to the participates prior to the interviews, and verbal informed consent that were taped was obtained from the participants.

Consent for publication

Before conducting the research, CGJ have explained the purpose of this research to the interviewees, ensured that the research materials will not be released, and verbal informed consent for publication has been obtained from the participants in this study.

Competing interests

The authors declare no competing interests.