How Does a Supervised Exercise Program Improve Quality of Life in Patients with Cancer? A Concept Mapping Study Examining Patients' Perspectives

How does participating in a supervised exercise program affect quality of life during or after cancer treatment? This study used concept mapping to assess patient responses to this question.

Abstract

Background.

Previous systematic reviews and meta‐analyses demonstrated beneficial effects of exercise during or following cancer treatment on quality of life (QoL). Aiming to understand how exercise contributes to a patient's QoL, we examined patients' perspectives via a process called concept mapping. This unique method provides structure and objectivity to rich qualitative data.

Methods.

Patients with cancer who were participating in an exercise program were invited to enroll. Eleven meetings with 3–10 patients were organized in which patients generated ideas in response to the question “How has participating in a supervised exercise program contributed positively to your QoL?” Next, patients individually clustered (based on similarity) and rated (based on importance) the ideas online. The online assessments were combined, and one concept map was created, visualizing clusters of ideas of how patients perceive that participating in a supervised exercise program improved their QoL. The research team labelled the clusters of ideas, and physiotherapists reflected on the clusters during semistructured interviews.

Results.

Sixty patients attended the meetings; of these, one patient was not able to generate an idea in response to the statement. Forty‐four patients completed the online clustering and rating of ideas. The resulting concept map yielded six clusters: personalized care, coaching by a physiotherapist, social environment, self‐concept, coping, and physical fitness and health. Personalized care was rated as most important. Overall, physiotherapists recognized these clusters in practice.

Conclusion.

Patients with cancer reported that participating in a supervised exercise program improved their physical fitness and influenced social, mental, and cognitive factors, resulting in improvements in QoL. These results can be used to increase the awareness of the importance of supervised exercise programs for the QoL of patients with cancer.

Implications for Practice.

According to patients, a supervised exercise program contributes positively to their quality of life by improving physical fitness and health and providing personalized care, coaching by a physiotherapist, and improved social environment, self‐concept, and coping. This knowledge could help to increase physicians' and patients' awareness of the importance of an exercise program during or following cancer treatment, possibly improving referral, participation, and adherence rates to these programs. Furthermore, patients' perspectives may be used to improve supervised exercise programs, taking into account the importance of personalized care, the supervision of a physiotherapist, the social environment, self‐concept, and coping.

Introduction

Patients with cancer often experience fatigue [1], [2], [3], reduced physical fitness and function [4], [5], and symptoms of anxiety and depression [1] as a consequence of cancer or cancer treatment, resulting in a decreased quality of life (QoL) [2], [6]. Several reviews and meta‐analyses have demonstrated that exercise during or following cancer treatment has beneficial effects on QoL [7], [8], [9], [10], [11]. Previous studies indicated that these beneficial effects were mediated by improved physical fitness [12], [13] and reduced fatigue [13], [14], [15]. Improved self‐efficacy [12] and reduced distress [12], [16] may also mediate the beneficial effects of exercise on QoL.

A previous metasynthesis summarizing qualitative studies concluded that patients perceive experiences such as positive distraction, social support, enhanced performance, and overcoming barriers in response to participating in exercise programs during or following cancer treatment [17]. One study included in the metasynthesis examined perceptions of QoL during a preoperative exercise intervention in patients with locally advanced rectal cancer (n = 10) and reported a greater sense of vitality, a positive attitude, enhanced social connections, and a sense of purpose in life as positive experiences during exercise [18]. This study indicates that participating in an exercise program provides physical, psychological, and social benefits possibly associated with increased QoL [18]. To the best of our knowledge, to date, no studies have investigated whether such perceptions may explain the positive effects of exercise on QoL. A better understanding of how patients perceive that participating in a supervised exercise program improves their QoL could help to increase the awareness under patients and care providers of the role of exercise during or following cancer treatment in improving QoL, possibly improving referral, participation, and adherence rates to exercise interventions.

The aim of this study was to identify patients' perspectives of how participating in a supervised exercise program contributes to QoL in patients with different types of cancer using the concept mapping methodology [19]. Concept mapping is a mixed method design in which group processes are combined with statistical methods, thereby allowing patients to provide their unique contribution to QoL research [19], [20]. Concept mapping is a valuable tool to improve patient care by providing a conceptual framework visualizing patients' perspectives [21]. To our knowledge, this is the first study investigating patients' perspectives of exercise during or following cancer treatment using this technique.

Materials and Methods

Patient Recruitment

Patients were eligible if they were 18 years or older, able to speak Dutch, and participated in a supervised exercise program during or following cancer treatment (e.g., chemotherapy, radiotherapy, immunotherapy, hormone therapy) for at least 4 weeks. In the Netherlands, first‐line physiotherapists can become specifically educated to supervise patients with cancer and are provided with knowledge on evidence‐based exercise interventions [22] and on how to tailor exercise interventions to comorbidities and treatment side effects [23]. Patients can be referred to these physiotherapists via their treating physician or general practitioner, most often at patients' own initiative. Twenty‐three physiotherapists who were educated to supervise exercise programs for patients during or following cancer treatment were contacted and asked to invite their patients to participate. Because we did not want to exclude patients from participation who were following an exercise program at the time of data collection, two patients were included who followed the exercise program for 2 weeks. All patients received an information letter and signed informed consent forms. According to the certified medical ethics committee that reviewed the study, this study did not fall within the scope of the Medical Research Involving Human Subjects Act.

Procedure

The current study is a concept mapping study combining group processes with multivariate statistical analyses [19]. In concept mapping, qualitative data (i.e., ideas in response to a focus statement and the importance of these ideas) are collected during structured brainstorm sessions and analyzed quantitatively, resulting in a concept map summarizing the results. In preparation for the brainstorm sessions, a focus statement was formulated that was designed to obtain information to answer the research question [20]. Because the aim of the current study was to identify patients' perspectives of how participating in a supervised exercise program improved their QoL, the focus statement focused only on the beneficial effects of exercise on QoL. Obtaining information on negative experiences was not part of the current study. Two‐hour brainstorm sessions were organized directly before or after an exercise session at physiotherapy practices in October 2016. Group size ranged from 3 to 10 patients, depending on the number of patients that exercised at the practice and were willing to participate. At the end of the brainstorm session, patients filled out a short questionnaire on sociodemographic and clinical characteristics (i.e., age, gender, education, cancer type, presence of metastasis, time since diagnosis, treatment type, and exercise duration). Patients received an invitation for an online assignment within 4 weeks after the brainstorm session. Patients who did not complete the online assignments received a reminder by e‐mail after 2 weeks and again 2 weeks later.

Concept Mapping

Brainstorm Session.

During the brainstorm sessions, the researchers supervising the session (M.S. and W.V.) introduced the aim of the study and explained the steps of concept mapping by means of a PowerPoint presentation. After the presentation, the focus statement was projected and presented in two ways:

“How has participating in an exercise program contributed positively to your quality of life?”

“Participating in an exercise program has contributed positively to my quality of life because…”

Patients were provided with pen and paper and were asked to write down as many ideas in response to the focus statement during a 15‐minute individual brainstorm, after which a group brainstorm started. Patients were asked to share all their ideas one by one. All ideas were projected on a screen and patients were asked whether the ideas were defined properly and unambiguous. After patients finished the group brainstorm, the online assessment was explained by means of an example that was projected on a screen. After completion of 11 brainstorm sessions, researchers (M.S., L.B., W.V., M.C., T.A.) discussed the content of all ideas and removed identical ideas. Subsequently, the collected ideas were imported into the software program Ariadne (Wilmslow, England, http://www.minds21.org/), which was used for the clustering and rating session.

Clustering and Rating Session.

Patients received an e‐mail with instructions for the online assignment, a list of all the ideas, and a personalized link to the online tool Ariadne. Patients were instructed to first cluster the ideas based on their similarity by creating two to ten clusters with at least two ideas per cluster and to provide titles for their clusters. Next, patients were asked to rate all ideas on importance on a five‐point Likert scale from “relatively unimportant” to “very important.” Patients could contact one of the researchers (M.S./W.V.) by phone or e‐mail if they needed help or clarification.

Statistical Analysis

Concept mapping data were analyzed using the software program Ariadne. Based on the input, Ariadne created a concept map which presented the mean rate of importance of all ideas and the relationship between the individual ideas. The concept map was based on completed cluster assignments. First, the software created a binary symmetric similarity matrix for every patient visualizing which ideas were clustered together. Binary symmetric similarity matrices of all patients were combined in a square group similarity matrix [24], [25], visualizing the number of patients that clustered a particular pair of ideas. Next, this number was used as proximity value for the multidimensional scaling analysis, which resulted in a point map representing each idea by a point. The relationship between ideas was reflected by the distance between the points on the map [25], [26]. Finally, the clusters were constructed using hierarchical clustering. The coordinates from the multidimensional scaling analysis were used to form a tree structure, starting with one single cluster including all ideas and moving to clusters that included each idea on its own [25]. The mean importance for each idea was calculated, based on patients' individual ratings of the importance of each idea. Subsequently, the mean importance of each cluster was calculated based on the mean importance of the ideas in that particular cluster.

Interpretation of the Concept Map

By default, the software program Ariadne created eight clusters based on the distance between points on the map. Two researchers (M.S., W.V.) discussed whether each cluster optimally represented patients' ideas (e.g., two clusters may include ideas related to a similar construct or a cluster may include ideas related to different constructs), and determined the final number of clusters. Next, the researchers (M.S., L.B., W.V., H.B., M.C., T.A.) discussed the content of the concept map and formulated cluster names that optimally covered the content of each cluster based on patients' cluster names. As statistical analyses may not always result in the best representation of qualitative data, the researchers critically reflected on the computer‐generated clusters. If an idea fit better in another cluster, the researchers relocated that idea.

After the concept map was constructed, the results were discussed with the physiotherapists in a semistructured interview. The goal of the interview was to evaluate whether the physiotherapists recognized the clusters representing patient‐reported experiences in their practice and to discuss how these results could help to optimize exercise programs for patients with cancer in order to improve their QoL. An interview guide was constructed by five researchers (M.S., L.B., W.V., M.C., T.A.) based on questions related to recognition of patients' individual ideas, the clusters of these ideas, and the importance of these ideas and clusters. Furthermore, questions regarding whether physiotherapists would change any aspect of the exercise program after knowing the results were included. Two researchers (T.A. and W.V.) conducted the first interview, and one researcher (W.V.) conducted all remaining interviews according to the interview guide (supplemental online Appendix 1). First, the results of the concept map were explained and physiotherapists were instructed to indicate anything they considered relevant for interpreting patients' perspectives of an exercise program. The interviews were audio‐recorded and transcribed. Subsequently, through thematic content analysis, two researchers (M.S., W.V.) independently extracted quotes supporting the answers to the questions defined in the interview guide using the software ATLAS.ti. Disagreements in extracted quotes were resolved by discussion. All ideas and quotes were translated by a native speaker of the English language.

Results

Thirteen (57%) physiotherapists invited their patients to participate in this concept mapping study. Three physiotherapists indicated that there were no patients with cancer following an exercise program at their practice at the time of the study, and seven physiotherapists could not be contacted in our specific time frame. Two physiotherapy practices had less than three patients who were interested in participation, and these patients were invited to join a brainstorm session at another practice. Eleven sessions were organized at different physiotherapy practices. Sixty patients participated in the sessions, of which 59 (98%) generated one or more ideas in response to the focus statement. Two patients were unable to attend a brainstorm session but performed the online assignments and provided sociodemographic and clinical characteristics via e‐mail. There were no patients who were invited to participate but who declined participation. Patient characteristics are presented in Table 1. Mean age was 57 (SD 11.2) years, 73% were female, and 63% were diagnosed with breast cancer. Sixteen patients (26%) reported that their cancer was metastasized to the lymphatic system, seven patients (11%) had metastases in one distant organ, two patients (3%) reported metastases in multiple organs, and two patients (3%) reported that they were diagnosed with metastatic disease but did not specify the location of metastases. The median duration of the exercise program was 20 (interquartile range, 12.0–40.0) weeks. In total, 44 patients (71%) completed the online cluster assignment, and 52 patients (84%) rated at least one idea on importance.

Table 1. Patient characteristics.

Abbreviation: IQR, interquartile range.

Concept Map

The number of ideas generated during each brainstorm session ranged from 10 to 39, resulting in a total of 297 ideas. After deleting duplicates, 186 ideas remained. As a maximum number of 98 ideas could be imported in Ariadne, overlapping ideas were combined to generate the final list of 98 unique ideas (Table 2). The derived concept map included six clusters: personalized care, coaching by a physiotherapist, social environment, self‐concept, coping, and physical fitness and health (Fig. 1). One idea, “because by participating in research I do other exercises, which stimulates me to exercise more and to do my best” (88, Table 2) did not fit in any cluster and was placed in the middle of the concept map based on the multidimensional scaling analyses. Two ideas (4 and 63, Table 2) were relocated by the researchers from cluster 5 to 4 and from cluster 6 to 5 (Fig. 1).

Table 2. List of 98 unique ideas generated by patients during concept mapping meetings.

Figure 1.

Concept map of patients perspectives of how supervised exercise programs during or following cancer treatment improves their quality of life: cluster name (rate of importance). Note that each point on the concept map represents one of the ideas that the patients generated in response to the focus statement. Points closer to each other were clustered more often together by the patients and are therefore related. The number of each point corresponds to the ideas presented in Table 2. The size of each point represents the mean rate of importance of the corresponding idea, with a larger point indicating a higher mean rate of importance. The line width of a cluster represents the mean rate of importance of all the ideas included in the corresponding cluster, with thicker lines representing a higher mean rate of importance. The arrows represent the ideas that were relocated by the researchers.

The cluster personalized care was rated as the most important (score 3.7) cluster contributing to QoL, followed by coaching by a physiotherapist and physical fitness and health (score 3.6). Self‐concept was rated as the least important cluster (score 2.5). The mean rate of importance of individual ideas ranged from 1.7 to 4.4 (Table 2), and the SD of the rate of importance of the individual ideas ranged from 0.8 to 1.5.

Semistructured Interview with Physiotherapists

Seven physiotherapists were interested in reflecting on the results during a semistructured interview. The remaining physiotherapists were unable to plan an interview because of absence or logistical reasons. All interviewed physiotherapists mentioned that they were familiar with the ideas generated by the patients and the clusters that were displayed in the concept map. Three physiotherapists mentioned in particular that they were familiar with the importance of the physiotherapist's knowledge of and experience with patients with cancer (supplemental online Appendix 2, quote 1). In addition, two physiotherapists were familiar with the importance of regaining physical fitness and strength (supplemental online Appendix 2, quote 2), and two were familiar with the contact with fellow patients with cancer and the possibility of exchanging experiences and information (supplemental online Appendix 2, quote 3). Three mentioned in particular that the concept map addressed all aspects that play a role during the exercise program (supplemental online Appendix 2, quote 4).

Regarding future perspectives to optimize the effect of exercise on QoL, three physiotherapists mentioned that an individualized program is very important for patients with cancer (supplemental online Appendix 2, quote 5). One physiotherapist was intrigued by the cluster coping. The physiotherapist hoped that participating in an exercise program helps patients to better cope with their disease but was unaware of how exercise may help patients to cope with their disease or how changes in coping can be measured (supplemental online Appendix 2, quote 6). Finally, three physiotherapists indicated that they could not think of anything that should have been done differently and that they supported the program as it is (supplemental online Appendix 2, quote 7).

Discussion

This study investigated patients' perspectives of how supervised exercise programs during or following cancer treatment improved their QoL using concept mapping. According to patients, personalized care, coaching by a physiotherapist, and improved physical fitness and health were the most important benefits of participating in a supervised exercise program that contributed to increased QoL. Furthermore, patients reported that the social environment, self‐concept, and coping contributed to the beneficial effects of supervised exercise on QoL.

The cluster physical fitness and health matches the results from previous quantitative studies that found improved physical fitness to mediate the exercise effects on QoL [12], [13], [14], [15]. In addition, previous studies suggested that psychological variables, such as self‐efficacy and distress, may mediate the exercise effect on QoL [12], [16]. Although self‐efficacy is not directly represented by one of the clusters, it may relate to ideas in the cluster self‐concept. Reduced distress was not directly reported as a factor contributing to improved QoL by patients in the current study. Possibly, patients in our sample did not experience distress and, consequently, may not have perceived reduced distress as a result of exercise, or patients perceived distress as a component of QoL, thereby not reporting it as a separate construct influencing their QoL.

The primary aim of a supervised exercise intervention is to provide guidance in preventing a deterioration in, or improving, muscle strength and aerobic fitness during or following cancer treatment [27]. The clusters physical fitness and health, personalized care, and coaching by a physiotherapist are therefore constructs that correspond directly to the aim and content of the exercise program. In addition, patients perceived that exercise resulted in an improved social environment, better coping, and self‐concept, contributing positively to their QoL. These experiences are in line with experiences reported in other qualitative studies, such as positive distraction, social support, and enhanced performance [28], [29], [30], [31], [32], [33], [34], despite different study aims. Future studies should confirm the potential mediating role of social environment, coping, and self‐concept in the exercise effect on QoL in patients with cancer. The items included in these clusters can inform the development of tools to measure changes in the social environment, self‐concept, and coping in this particular patient group.

Clinical Implications

Patients perceived that exercise not only improved physical fitness but also influenced social, mental, and cognitive factors, which resulted in QoL benefits. This supports the results from a previous qualitative study in patients, health care providers, and policymakers, reporting social, mental, and physical dimensions as well as QoL as indicators of health [35]. The constructed concept map provides a conceptual framework visualizing patients' perspectives, which can help to improve patient care [21]. In particular, knowledge of constructs influencing QoL could help to increase physicians' and patients' awareness of the importance of a supervised exercise program during or following cancer treatment, thereby possibly improving referral, participation, and adherence rates to exercise interventions. Furthermore, patients' perspectives reported in the current study may be used to further optimize supervised exercise programs taking into account the importance of personalized care, the supervision of a physiotherapist, the social environment, self‐concept, and coping. The mediating role of these constructs in the exercise effect on QoL should be confirmed in a future intervention study.

Strengths and Limitations

To the best of our knowledge, our study is the first concept mapping study identifying patients' perspectives of how exercise during or following cancer treatment improves their QoL. In concept mapping, patients have a unique contribution to research, and the input is analyzed using sophisticated multivariate statistics. The perspectives of a large group of patients are visualized and can be used to increase awareness of the importance and potential of exercise programs for patients during or following cancer treatment to increase QoL.

There are some limitations associated with this study. The focus statement is driven by the aim of the study [36] and therefore focused on participants' experiences of the beneficial effects of exercise on their QoL. From previous literature, it is known that exercise can improve QoL [7], and in the current study, only one patient did not perceive positive effects on QoL and was therefore not able to provide an idea in response to the focus statement. However, from this study, no conclusions can be drawn on the negative effects of participating in a supervised exercise program on QoL. Some caution is warranted with generalizing the results of the current study to all patients with cancer. Most patients were women with breast cancer and patients treated with curative intent, and it is unclear whether patients' perspectives differ between patients with different types or stages of cancer. Furthermore, patients voluntarily participated in an exercise program, and it is unclear whether patients not participating in a supervised exercise program would identify similar constructs to contribute to improvements in QoL when exercising. Because of software constraints, the total number of unique ideas had to be reduced to 98 by combining overlapping ideas. Despite thorough discussion, this sometimes resulted in ideas with a long description, and some ideas may have been ambiguous to some patients. Furthermore, some patients reported difficulties with the online software used for the clustering and rating assignment. Therefore we recommend future research to organize a second face‐to‐face meeting for the clustering and rating assessment instead of the online assessment at home, in order to facilitate help or clarification. In addition, clustering and rating assessment on paper is also recommended for participants with limited computer skills.

Conclusion

According to patients, a supervised exercise program contributes positively to QoL by improving physical fitness and health and providing personalized care, coaching by a physiotherapist, and an improved social environment, self‐concept, and coping. This information shows that cancer rehabilitation programs improve patients' QoL, not only via improvement in physical fitness but also via social, mental, and cognitive factors associated with participating in a supervised exercise program. Knowledge of the wide range of factors influencing QoL can help to inform patients and care providers of the potential of exercise during and following cancer treatment for improving QoL, possibly improving patient referral, awareness, and adherence to exercise programs.

See http://www.TheOncologist.com for supplemental material available online.

Author Contributions

Conception/design: Maike G. Sweegers, Laurien M. Buffart, Johannes Brug, Mai Chin A Paw, Teatske M. Altenburg

Provision of study material or patients: Maike G. Sweegers, Wouke M. van Veldhuizen, Edwin Geleijn

Collection and/or assembly of data: Maike G. Sweegers, Laurien M. Buffart, Wouke M. van Veldhuizen, Edwin Geleijn, Teatske M. Altenburg

Data analysis and interpretation: Maike G. Sweegers, Wouke M. van Veldhuizen, Henk M.W. Verheul, Johannes Brug, Mai Chin A Paw, Teatske M. Altenburg

Manuscript writing: Maike G. Sweegers, Laurien M. Buffart, Wouke M. van Veldhuizen, Edwin Geleijn, Henk M.W. Verheul, Johannes Brug, Mai Chin A Paw, Teatske M. Altenburg

Final approval of manuscript: Maike G. Sweegers, Laurien M. Buffart, Wouke M. van Veldhuizen, Edwin Geleijn, Henk M.W. Verheul, Johannes Brug, Mai Chin A Paw, Teatske M. Altenburg

Disclosures

The authors indicated no financial relationships.