Repeated use of rich pictures to explore changes in subjective experiences over time of patients with advanced cancer

Abstract Background The combination of verbal and visual tools may help unravel the experiences of advanced cancer patients. However, most previous studies have focused on a specific symptom, at only one moment in time. We recently found that a specific visual tool, originating from systems thinking, that is, rich pictures (RPs), could provide a more comprehensive view of the experiences of patients with advanced cancer. Aims To examine whether the repeated use of RPs can make changes in subjective experiences of patients living with advanced cancer visible over time. Methods and results We performed a prospective study with a generic qualitative approach that was mostly informed by the process of grounded theory. We invited patients to make an RP twice, at the start of the study, and again after 2 months. Both RP drawing sessions were directly followed by a semi‐structured interview. Patients with all types of solid tumors, above the age of 18, and with a diagnosis of advanced, incurable cancer, were eligible. Eighteen patients participated and 15 patients were able to draw an RP twice. In eight RP‐sets, considerable differences between the first and second RP were noticeable. Two patterns were distinguished: (1) a change (decline or improvement) in physical health (five patients), and/or (2) a change in the way patients related to cancer (three patients). Conclusion RPs are a valuable qualitative research method that can be used to explore the experiences of patients with advanced cancer, not only at a single point in time but also over time.


| BACKGROUND
Having advanced incurable cancer has a profound impact on the quality of life of cancer patients. [1][2][3][4] Patients may not always be able to find words to express what they feel, particularly in such confusing and distressing circumstances. 5,6 Hence, current methods to capture subjective experience, such as questionnaires and in-depth interviews, possibly only uncover part of the experiences of patients with advanced cancer. [7][8][9] This is also important because patient's experiences can change over the course of their illness. 10 For example, a systematic review and meta-analysis by van Kleef et al. shows that cancer patients experienced an improvement in some physical symptoms and emotional functioning during treatment, but a deterioration in role functioning, fatigue, and negative feelings about hair loss. 11 If we could invite these patients to tell their story through visuals, in addition to spoken language, we might gain a more complete understanding of what they are experiencing during the whole course of their illness. 6,[12][13][14][15] In previous studies, the combination of verbal and visual tools was found to help unravel the experiences of patients. [15][16][17][18][19] For instance, reviews by Broadbent et al. and Cheung et al. provide an overview of studies that explored patients' illness perceptions using drawings and conclude that drawings can aid in understanding the illness experience. 18,19 However, it should be noted that in most of the studies reviewed, the drawing instruction had a narrow focus-asking patients to draw themselves, the illness, the part of their body that is affected by the illness, or how the illness made them feel. 18,19 In addition, in all but one of these studies, drawings were made at only one moment in time. [15][16][17][18][19] We recently found that a specific visual tool, originating from systems thinking, that is, rich pictures (RPs), could provide a more comprehensive view of the experiences of patients with advanced cancer. 20 RPs are drawings that research participants create to express their experiences, including different elements, for example, people, materials, processes, interactions, and feelings. 5,6,12,13,21 RPs were found to combine different aspects of patients' lives, communicating these in one single snapshot. 20 This promising finding encouraged us to address the subsequent question: How can the repeated use of RPs make changes in subjective experiences of patients with advanced cancer visible over time and what are the most optimal research conditions to achieve this? We explored these questions in a qualitative study among the same patients with advanced cancer who had drawn an RP previously. 20 2 | METHODS

| Study design and patients
Reporting of the study was done according to the consolidated criteria for reporting qualitative research. This study adopted a generic qualitative approach that was mostly informed by the process of grounded theory. 22 We applied purposive sampling and started data analysis alongside data collection to inform subsequent sampling and data collection. We invited patients to make an RP twice, at the start of the study, and again after 2 months. Both RP drawing sessions were directly followed by a semi-structured interview. The findings of the first round of RP interviews were previously published. 20 The present study will specifically focus on the repeated use of the RPs. We adopted a constructivist research paradigm acknowledging that knowledge is co-created between researchers and patients. 23 Patients constructed their own reality when making an RP, which was

| Data collection
During the first phase of the research, we included 18 patients, as this allowed for some loss-to-follow-up enabling a sufficiently large sample size to obtain meaningful findings. 20,23 Clinical data, such as performance status, were obtained from the medical chart. From February till June 2018, we conducted two RP sessions of approximately 1 hour with each patient, with an interval of 2 months, in a setting of their preference: at the hospital or their own home, and with or without a family member or friend being present. In both sessions, we asked patients to make an RP that depicted how they experienced living with cancer at the moment in time. Patients were encouraged to include all aspects of their life that they felt were relevant to their experience. When the RP was finished, the interviewer invited the patient to explain the content of the drawing, which was followed by more specific questions about the colors, shapes, and specific elements used. All interviews were audio recorded. During the whole study period, Z.M.B. kept a notebook to collect field notes. A detailed description of the procedure of the interviews has been described previously. 20

| Repeated use
To explore different options for instructing patients for drawing the second RP, we divided the patients into three conditions at the time of the second interview. In condition one, patients were asked to make a completely new RP. In condition two, patients were presented with a copy of their RP from the first interview and subsequently asked to make a new RP. We left it to the patient's own discretion whether the new RP would be related to the first, for example, by expanding or deleting elements, or would be unconnected. In condition three, patients were given a copy of their RP from the first interview and invited to make adjustments on the copy itself, rather than making a new RP. For example, patients could adjust elements, erase elements, add new elements, change colors, or even leave the RP as it was when they felt it still represented their current experience.
Patients were assigned to the three conditions alternatingly, in order of appearance at the second interview, such that the first available patient was assigned to condition one, the second patient to condition two, the third to condition three, the fourth to condition one, etc. However, patients could change conditions during the RP session when they were not able or willing to follow the instructions specific to those conditions. For instance, when patients expressed a wish to make a new RP, they were allowed to switch to condition three, or when patients opposed to making a new RP, they could switch to condition two.

| Data analysis
Interviews were transcribed verbatim and used to support the RP analysis. Transcripts were not returned to patients for additional comments. Data analysis of the combined "Rich Picture/interview" data included both inductive and deductive approaches. The first set of RPs was analyzed in an inductive way, as described in detail before. 20 During so-called "RP viewing sessions" one single RP was analyzed in detail for an hour by six to eight researchers who had experience with RP research, using content analysis. 24,25 In gallery walks, all RPs were pinned to a wall in random order. Attendees (researchers with backgrounds in medicine, theology, arts, qualitative research, and RP research, and also cancer patients) walked along to study the RPs and subsequently discussed the RPs together, looking for patterns, disparities, and similarities across the entire data set. The RP viewing ses-

| Ethical considerations
No formal ethical approval was needed for the study, as confirmed by to withdraw from the study at any moment. Patients were informed that their RPs would be used for research purposes and could be displayed to others for data analysis. All patient data were treated confidentially and stored in a protected database in the hospital.

| Patients
Eighteen patients participated in the first session of RP drawing and interviews (Table 1). 20 Five patients were lost to follow up during the study. Two of them passed away before the second session, one declined further participation, and two could not find time due to busy private schedules. In the other eight RP-sets, considerable differences between the first and second RP were noticeable. Two patterns could be distin- Lost-to-follow-up. changed more often preferred to make a new RP (condition one or two). Two patients expressing considerable change, however, were assigned to condition 3 and were able to show changes by building on/changing their previous RP. Below we will describe the results of five patients as an example.

| RPs illustrating changes in physical health
A decline in physical health was visible when comparing the first and second RP of patient P6 (Figure 1(A,B)). In his first RP, this patient drew a snowflake to represent the neuropathy he experienced when the weather was cold. He also drew an empty balloon symbolizing the emptiness of his days due to his inability to work. Whereas he put a cross through the snowflake in his second RP, as the weather got better, he added a drawing of his hand and feet hurting because of neuropathy ( Figure 1(B)). He also added a drawing of himself sleeping on a pillow, having stomach pains, with his ribs showing through his skin, and his hair falling out. In sum, he was more tired and in pain, and suffered from weight loss and alopecia due to chemotherapy. This visualized decline was confirmed by his medical record that indicated an elevated WHO performance status (from one to two), indicating deteriorating physical health.
The RPs of patient P14 showed improved physical health. In the first RP, the central element was the patient lying in bed, in close proximity of the toilet and medication, whereas in the second RP only a small bed was drawn (Figure 1(C,D)). In addition, in the first RP, the patient specifically wrote down that she could not cycle anymore and could not join others in activities. In the second RP, she drew pictures of cycling and hiking. The increasing physical health of the patient was in line with the improvement of her WHO performance status (one to zero).

| RPs illustrating differences in the way patients relate to cancer
The repeated use of RPs allowed us to identify distinct ways of relating to cancer over time. Patient P2 drew himself in his first RP thinking about death and the loss of his future, and explicitly depicted the sadness and anger this invoked (Figure 2(A)). He acknowledged he had cancer, but did not accept it. For his second RP, however, he just drew one element: an ostrich sticking its head into the sand (Figure 2 (B)). He explained: "I bury my head in the sand, I just pretend it is not there." In other words, after initially acknowledging cancer, he now decided to try to deny the existence of cancer.
A change from acknowledging to acceptance was found in patient P5. In her first RP, she depicted great sadness and anger about the loss of her life and drew, among other things, her grave and a peace sign with a question mark over it (Figure 2(C)). She had added the latter sign because the question "is this supposed to bring me peace?" kept popping up in her head. Hence, this patient acknowledged her situation, but steps toward acceptance were not yet taken. In her second RP, she put a cross through the peace sign and the question mark and colored the grave with orange and green, because her immanent death was less haunting. She also added three elements to the RP: flowers, leaves, and an Ohm-sign (Figure 2(D)). These, respectively, represented the happiness about being able to exercise her hobbies, the inner peace she had now found, and the regained connection with her religion. About finding inner peace she said: "I think the acceptance is just way bigger. A certain calmness, more inner peace than the first time." Patient P18 showed a transition from acknowledging to receiving.
In her first RP, she depicted the cancer in her body and the mental struggles she was having (Figure 2

| What this study adds
To the best of our knowledge, this was the first study that examined the potential of RPs as a method for capturing changes in the subjective experience of living with cancer over time. Previous research has applied RPs to gain insight into subjective experiences at one time only. 20 Since we found that the repeated use of RPs revealed meaningful changes over time, we suggest that the repeated use of RPs has potential as a longitudinal follow-up tool or as an outcome measurement before and after an intervention, for example, a mental health intervention for advanced cancer patients.
In comparison to other studies that used drawings to explore patient experience, 16,18,19, , the added value of RPs lies in the ability to represent all important themes of a patient's life in one single snapshot. When making a RP, patients are invited to include everything that they consider relevant to the situation in their drawing, from events, people, and objects to feelings and emotions, instead of just focusing on one element such as pain or anxiety. 16,18,19 Hence

| Strengths and limitations
A strength of our study is that it explored the repeated use of RPs in a rigorous way. We built upon previous RP research, examined three different interview conditions, and performed an interim analysis. The interim analysis allowed us to make the best possible use of our 13 patients, as we did not waste interviews on interview conditions that were not working.
It should be noted that we did not intend to make a formal comparison in terms of outcome between the three modes of instruction, but rather these findings, we argue that when patient preferences for the drawing of RPs is taken into account, repeated use of this interview format is feasible.
Next, researchers from different backgrounds, but also patients, an artist, and doctors, gave us the opportunity to incorporate many backgrounds in the data interpretation, enriching the results. In this study, we aimed for consensus in data interpretation between the different actors, while the level of agreement between coders was not of main interest. A final strength of this study is the use of the theoretical model of Kruizinga et al. 26 to support the interpretation of how patients relate to the cancer at the two moments of time.
Generalizability of our findings, however, might be limited due to practical factors. RPs may not be suitable for all patients with advanced cancer, as engaging in an RP session of approximately an hour may be too burdensome. In addition, not all patients may feel comfortable with expressing themselves using visual methods. This may be patient, age, and culture dependent. Finally, as is the case with almost all measurements, drawing of an RP may by itself induce a change in one's experience of living with cancer, because a reflective process is brought into motion. The sensitivity of RPs to pick up changes after an intervention compared to a control group needs further investigation.

Conclusions
RPs were found to provide insight into changes in physical condition as well as how patients relate to living with cancer. RPs are a valuable qualitative research method that can be used to explore the experiences of patients with advanced cancer, not only at a single point in time but also over time.

ACKNOWLEDGMENTS
This work was supported by The Young Academy, Royal Dutch Academy of Sciences, and the Dutch Cancer Society (grant number 11507). We thank the patient organization SPKS (Stichting voor Patiënten met Kanker aan het Spijsverteringskanaal) for their participation in the analysis of the data of the first rich pictures set.