Abstract
Background. In 2002, the Danish Cancer Society opened a rehabilitation centre in which cancer patients were offered a free, six-day, multidimensional residential course. Our previous studies of the effects of this course at one and six months of follow-up showed no positive effect on distress. We investigated long-term effects at 12 months of follow-up and whether subgroups with fewer psychosocial resources received more benefit from the intervention than patients with better resources.
Material and methods. In two Danish counties, 507 patients with breast, prostate, colon or rectum cancer diagnosed within the past two years who had completed primary treatment were randomised to a six-day, multidimensional residential rehabilitation course or to standard care. Of these, 208 patients received the allocated intervention and 244 received the allocated control condition and were included in the analyses. Patients in both groups completed questionnaires at baseline and at one, six and 12 months of follow-up, including the ‘Profile of Mood States short form’, the ‘General Self-efficacy’ scale and a question on emotional support. At 12 months of follow-up, 179 participants in the intervention group and 195 in the control group provided data.
Results. No effect of the intervention was found on distress at 12 months of follow-up, even in subgroups with fewer psychosocial resources at baseline, i.e. greater baseline distress, poorer self-efficacy and less emotional support.
Conclusion. Multidimensional rehabilitation programmes may not be effective in the treatment of distress. During the past few decades, studies of psychotherapy or psycho-education in cancer patients have shown small to moderate effects. More focused rehabilitation programmes may be more effective.
For decades, there has been growing interest in supporting cancer patients to improve their quality of life and in rehabilitation to meet the demands of life as cancer survivors [Citation1,Citation2]. Therefore, increasing numbers of trials have investigated various support programmes. With the exponential increase in the number of cancer survivors and the introduction of chronicity as a term for cancer, the provision of adequate rehabilitation has become more important [Citation2]. The formal definition of rehabilitation is “the process by which physical, sensory and mental capacities are restored or developed in (for) people with disabling conditions” [Citation3]. As not all cancer patients experience disability, in 1989, the Oncology Nursing Society in the USA more broadly defined cancer rehabilitation as “a process by which individuals within their environments are assisted to achieve optimal functioning within limits imposed by cancer” [Citation3].
In 2002, the Danish Cancer Society, a non- governmental organisation of which 10% of the population (N∼ = 500 000) are paying members, opened a rehabilitation centre at a fully restored mediaeval castle, where cancer patients are offered a free, six-day, residential psychosocial course that includes daily teaching sessions conducted by professionals. Our previous comparisons of the effect of this course with standard care in patients with breast, prostate or colorectal cancer at one and six months of follow-up [Citation4–8] did not show an effect on distress or quality of life [Citation4] or on health behaviour, body mass index or self-rated health [Citation5].
The purpose of the present study was two-fold. First, we evaluated the effect of the psychosocial cancer rehabilitation course on distress, the primary outcome at 12 months of follow-up. Second, we conducted exploratory secondary analyses to determine whether subgroups of patients with fewer psychosocial resources (i.e. greater distress, poorer self-efficacy and less emotional support) benefited from the programme. Further, we explored gender, cancer type, age and education as possible effect modifiers.
Material and methods
Study participants
Patients were recruited from six hospital departments in two Danish counties, Aarhus and Frederiksborg. The eligibility criteria included a primary cancer of the breast, prostate, colon or rectum diagnosed within the past two years, completion of primary treatment (except for hormonal treatment) and physical ability to participate in intervention activities. The exclusion criteria were acute need of treatment and a life expectancy of less than six months. The sampling has been described in more detail elsewhere [Citation4].
Study design
Consenting patients were randomised to a rehabilitation intervention programme or usual care. The intervention was a multi-focus psychosocial empowerment intervention at the Dallund Rehabilitation Centre, run by the Danish Cancer Society, and consisted of a free six-day retreat (20 cancer patients participated each week, including non-randomised patients from other Danish counties) combining lectures, discussions and patient group work on the treatment of cancer, psychological reactions, spirituality, working life and lifestyle. A central part of the intervention was sharing experiences with peers. In collaboration with a health professional, all participants made a personal “action plan” to support lifestyle changes upon their return to daily life. The intervention has been described in detail elsewhere [Citation4,Citation6]. Data management, storage and security with regard to this study were approved by the Danish Data Protection Agency (File number: 2007-41-1596).
Measures
At baseline (before randomisation) and at one, six and 12 months of follow-up, the patients in both groups filled out a questionnaire on self-reported measures of distress, self-efficacy and emotional support.
The ‘Profile of Mood States short form’ (POMS-SF) [Citation9], the short version of the ‘Profile of Mood States’ scale [Citation10], consists of a list of 37 adjectives; respondents indicate the degree to which each adjective describes their mood during the previous week on a five-point Likert scale (from “not at all” to “extremely”). Six subscales are derived: tension-anxiety, depression-dejection, anger-hostility, fatigue-inertia, vigour-activity and confusion-bewilderment. A score for total mood disturbance (TMD) is calculated by subtracting the score for the only positively scored subscale, vigour-activity, from the sum of the other five subscales. Thus, a higher score for TMD represents greater mood disturbance. The scale has been shown to be reliable and valid for cancer patients [Citation11].
Self-efficacy was measured by the Danish version of the General Self-efficacy scale [Citation12], which consists of 10 statements; respondents indicate the extent to which they consider each statement to characterise them as a person on a four-point Likert scale (from “not at all true” to “exactly true”). A higher sum score represents greater self-efficacy. The scale has been shown to be valid across cultures [Citation13].
Emotional support was measured from answers to the question: “Can you have a good talk with the following persons if you need their support?” This question was answered with regard to the partner by the response categories: “Don't have any”, “Always”, “Often”, “Sometimes”, “Rarely” or “Never”. This item is part of a series of items measuring various aspects of social networks and has been used and tested in other studies [Citation14].
Self-reported information on education and employment status was obtained from the baseline questionnaire.
All residents of Denmark are assigned a unique 10-digit personal identification number that includes date of birth and which permits linkage of information between registers. Cancer diagnosis, date of diagnosis, treatment modalities and tumour stage were obtained from nationwide clinical cancer databases for breast [Citation15] and colorectal cancer [Citation16] and from medical records for prostate cancer. If information was not available in these data sources, the information was retrieved from the Danish Cancer Register, when possible [Citation17].
Analyses
First, unadjusted analyses with the t test were used to estimate the association between group assignment (i.e. intervention or control group) and changes between baseline and 12 months on POMS TMD. The analyses were subsequently stratified on gender, age, diagnosis and length of education.
Adjusted analyses included the following potential predictors: group assignment, distress score at baseline, gender, time since diagnosis, age at baseline, cancer diagnosis and length of education as well as all interactions involving group assignment and the above-mentioned predictors. Analysis of the longitudinal observations for each patient (at one, six and 12 months of follow-up) was conducted by applying a random effect for each patient.
The secondary analysis of possible effect modification on TMD was a repetition of the adjusted analysis but including self-efficacy and emotional support from the partner. In this analysis, baseline distress was entered first as a continuous variable and subsequently as a dichotomised variable, where high baseline distress was defined as values above the median value 2.
Effects of the intervention are reported on an absolute scale as estimated differences in score reduction between the two groups, with 95% confidence intervals. A significance level of 5% was applied, and all statistical tests were two-sided. The statistical software R [Citation18] was used for all analyses.
Results
Of the 507 participants who were randomised, 259 patients were assigned to the intervention and 248 to standard care. Of these, 208 patients received the allocated intervention and 244 received the allocated control condition and were included in the analyses. At 12 months of follow-up, 179 participants (86%) in the intervention group and 195 (80%) in the control group provided data ().
Figure 1. Randomisation and follow-up of cancer patients participating in the study of cancer rehabilitation in Denmark, 2004–2008.
Most of the patients who did not receive the allocated intervention had colorectal cancer, and the smallest proportion was breast cancer patients (). Those who did not receive the intervention were less likely to have high-risk disease and to have received radiotherapy or hormone treatment, and they reported less baseline distress than those who received the intervention. No other substantial differences between the three groups (i.e. those who received the allocated intervention, those who never received the allocated intervention and the control group) were observed at baseline.
Table I. Baseline characteristics of patients with breast, prostate or colorectal cancer (n = 507) in a randomised controlled trial of a cancer rehabilitation intervention, by receivers and non-receivers of the intervention although allocated, and receivers of the allocated control condition.
Primary aim
In unadjusted analyses, group assignment (intervention or control group) was not associated with changes in distress between baseline and 12 months (p = 0.42), nor was there an effect of group assignment in analyses stratified by gender, age, diagnosis or length of education (data not shown).
In adjusted analyses of the effect on POMS TMD, there were no significant interactions between group assignment and baseline TMD score, gender, time since diagnosis, age, diagnosis or length of education (). The analyses in which the correlation between repeated observations for each patient was accounted for yielded the same conclusion ().
Table II. Adjusted analysis of changes between baseline and 12 months of follow-up. Included as predictors are: intervention, distress score at baseline, gender, time since diagnosis, age at baseline, cancer diagnosis and education as well as all interactions with intervention status (i.e. intervention or control group). The estimate is reported as point changes in POMS TMD.
Table III. Adjusted analysis of changes between baseline and 12 months of follow-up, where measurements at 1- and 6-months follow-up are included. Included as predictors are: intervention, distress score at baseline, gender, time since diagnosis, age at baseline, cancer diagnosis and education as well as all interactions with intervention status (i.e. intervention or control group).
Secondary aim
The effect of the intervention was not modified by self-efficacy, emotional support or baseline TMD (p-value for the interaction between the intervention and baseline TMD as a continuous variable: 0.47, data not shown), however when baseline distress was dichotomised into high (> 2) vs. low (≤ 2) distress, the interaction was significant (p = 0.02) such that the intervention had a stronger effect on distress among those with low baseline distress ().
Table IV. Adjusted analysis of changes between baseline and 12 months of follow-up including indicators of few psychosocial resources as possible effect modificators. As measurements at 1- and 6-months follow-up are included, the correlation between repeated observations (within patient) is accounted for. Included as predictors are: intervention, follow-up time, gender, time since diagnosis, age at baseline, cancer diagnosis and education as well as all interactions with intervention status (i.e. intervention or control group).
Discussion
Consistent with an earlier report from this study that showed no positive effect on distress at one and six months of follow-up [Citation4], no effect on distress was found at 12 months of follow-up, nor were there positive effects of the intervention in subgroups with fewer psychosocial resources at baseline. However, when baseline distress was dichotomised, the intervention had a higher effect among those with low baseline distress. This effect was not seen in analyses including baseline distress as a continuous variable and should be interpreted with caution taken the multiple testing into account. This may be a chance finding. It has been argued that lack of effect of psychosocial intervention could be caused by inclusion of non-distressed cancer patients that hindered improvement on distress. In the present study, secondary analyses were done to test whether the most distressed patients improved more than those with low baseline distress. This was not the case.
The intervention used was designed to meet various potential rehabilitation needs such as late effects, impaired physical functioning and distress and offered education about the disease and treatment, diet, sexuality, return to work etc. In addition to the educational sessions, healthy meals were offered and participants were encouraged to try similar recipes at home; the beautiful surroundings offered the opportunity for walking and enjoying nature; and being with peers in a residential setting allowed exchanges of experiences and mutual support. Although the content of the rehabilitation courses was modified for certain groups of participants (e.g. courses for patients with certain diagnoses and young patients), the content may not have been specific enough to cover individual patients’ needs. A multidimensional one-week residential rehabilitation course for Swedish breast cancer patients found a positive effect on anxiety at 12 months’ follow-up but no effect on depression [Citation19]. A Cochrane review of the effect of multidimensional rehabilitation programmes showed no effect on psychosocial measures and suggested that programmes with a narrower focus were more successful [Citation20].
Although the intervention was relatively lengthy it could be seen as a break from everyday life and as such, it was not easily transferred to daily living. In an attempt to transfer the intervention to life lived after the retreat, each patient was asked to make an individually tailored “action plan” to follow after returning home. Unfortunately, no information is available on adherence to these “action plans”, and a large group of participants may never have transferred the intervention to their everyday lives. Inability to adhere to the “action plan” might have engendered feelings of guilt and insufficiency. Furthermore, the exchanges with peers might have been upsetting and resulted in negative downwards comparisons (e.g. feeing anxious when connecting with someone who is worse off) or might have created an identity as a cancer patient that hampered a return to normal life. This hypothesis is supported by findings from a randomised study of Internet groups formed by participants at the Dallund Rehabilitation Centre, in which participants in the peer-support groups reported less reduction in anxious preoccupation, helplessness, confusion and depression than participants who were not allocated to Internet groups after the retreat [Citation8]. A qualitative study based on observations at nine courses at the Rehabilitation Centre, in-depth interviews with patients participating in the courses and written sources demonstrated the complexity of the experiences. While participants reported feeling unique and cared for by dedicated staff at the beautiful castle, some felt that the programme was too demanding and was emotionally upsetting [Citation21].
A major strength of this study is its randomised design, as few residential rehabilitation courses have been evaluated in randomised studies. Also, the sample of cancer patients was relatively large. A limitation is the large (20%) group of patients randomised to the intervention group who did not receive the intervention. As these patients were not sent questionnaires, intention-to-treat analyses could not be performed; however, diagnosis, baseline distress and other baseline characteristics of completers and non-completers were adjusted for in the multiple analyses. It is difficult to estimate the extent to which the results can be generalised, as patients who participate in a rehabilitation course may not be representative of all cancer patients. This should not, however, compromise the internal validity of the study.
In this randomised study, we did not find an effect on distress of a six-day multidimensional residential course among patients with breast, prostate or colorectal cancer. Earlier reports also showed no effect on quality of life [Citation4], health behaviour, BMI or self-rated health [Citation5], in line with the findings of a Cochrane review [Citation20]. Thus, multidimensional rehabilitation programmes may not be the best way to reduce distress in cancer patients. During the past decades, studies of varying quality have shown small to moderate effects of psychotherapy or psycho- education [Citation22]. More recently, promising results have arisen from studies in which yoga [Citation23] and mindfulness [Citation24,Citation25] were used mainly for breast cancer patients. Just as in the case of introducing psychosocial interventions some four decades ago, it is crucial that it is determined whether such interventions appeal to other groups and whether the results apply to all cancer patients. Also, the effect should be compared to that of other group activities or relaxation. We nevertheless find it interesting that a structured intervention with up to six hours of dedicated education every day for six days in combination with interaction with peers and full reimbursement of transport, accommodation and meals had no effect on major outcomes of a priori interest for planning cancer rehabilitation.
Acknowledgements
We thank the participating cancer patients for their valuable contributions. We also acknowledge the important collaboration of the staff at the Dallund Rehabilitation Centre and the participating hospital departments.
Declaration of interest: The authors report no conflicts of interest. The authors alone are responsible for the content and writing of the paper.
The study was funded by the Danish Cancer Society (Grant number 22506001).
References
- Dalton SO, Bidstrup PE, Johansen C. Rehabilitation of cancer patients: Needed, but how? Acta Oncol 2011;50: 163–166.
- Dalton SO, Johansen C. New paradigms in planning cancer rehabilitation and survivorship. Acta Oncol 2013;52:191–4.
- McQuellon R, Danhauer S. Psychosocial rehabilitation in cancer care. In: Ganz P, editor. Cancer survivorship: Today and tomorrow. New York: Springer; 2007. pp 1942–54.
- Rottmann N, Dalton SO, Bidstrup PE, Wurtzen H, Hoybye MT, Ross L, et al. No improvement in distress and quality of life following psychosocial cancer rehabilitation. A randomised trial. Psychooncology 2012;21:505–14.
- Ibfelt E, Rottmann N, Kjaer T, Hoybye MT, Ross L, Frederiksen K, et al. No change in health behavior, BMI or self-rated health after a psychosocial cancer rehabilitation: Results of a randomized trial. Acta Oncol 2011;50:289–98.
- Hoybye MT, Dalton SO, Christensen J, Larsen LR, Kuhn KG, Jensen JN, et al. Research in Danish cancer rehabilitation: Social characteristics and late effects of cancer among participants in the FOCARE research project. Acta Oncol 2008;47:47–55.
- Kjaer TK, Johansen C, Ibfelt E, Christensen J, Rottmann N, Hoybye MT, et al. Impact of symptom burden on health related quality of life of cancer survivors in a Danish cancer rehabilitation program: A longitudinal study. Acta Oncol 2011;50:223–32.
- Hoybye MT, Dalton SO, Deltour I, Bidstrup PE, Frederiksen K, Johansen C. Effect of Internet peer-support groups on psychosocial adjustment to cancer: A randomised study. Br J Cancer 2010;102:1348–54.
- Shacham S. A shortened version of the Profile of Mood States. J Pers Assess 1983;47:305–6.
- McNair D, Lorr M, Droppleman L. EITS manual for the profile of mood states. San Diego, CA: Educational and Industrial Testing Service; 1871.
- Baker F, Denniston M, Zabora J, Polland A, Dudley WN. A POMS short form for cancer patients: Psychometric and structural evaluation. Psychooncology 2002;11:273–81.
- Schwarzer R, Jerusalem M. Generalized self-efficacy scale. In: Weinman J, Wright S, Johnston M, editors. Measures in health psychology: A user's portfolio. Windsor, England: NFER-NELSON; 1995.
- Luszczynska A, Scholz U, Schwarzer R. The general self-efficacy scale: Multicultural validation studies. J Psychol 2005;139:439–57.
- Ross L, Thomsen BL, Boesen SH, Frederiksen K, Lund R, Munk C, et al. Social relations and smoking abstinence among ever-smokers: A report from two large population-based Danish cohort studies. Scand J Public Health 2013;41:531–40.
- Moller S, Jensen MB, Ejlertsen B, Bjerre KD, Larsen M, Hansen HB, et al. The clinical database and the treatment guidelines of the Danish Breast Cancer Cooperative Group (DBCG); its 30-years experience and future promise. Acta Oncol 2008;47:506–24.
- Harling H, Nickelsen T. [The Danish Colorectal Cancer Database]. Ugeskr Laeger 2005;167:4187–9.
- Storm H, Michelsen E, Clemmesen I, Pihl J. The Danish Cancer Registry – history, content, quality and use. Dan Med Bull 1997;44:549–53.
- R: A language and environment for statistical computing. R Core Team, R Foundation for Statistical Computing: Vienna, Austria. Available from: www.R-project.org/ [cited 2014 Aug 12].
- Bjorneklett HG, Lindemalm C, Rosenblad A, Ojutkangas ML, Letocha H, Strang P, et al. A randomised controlled trial of support group intervention after breast cancer treatment: Results on anxiety and depression. Acta Oncol 2012;51:198–207.
- Scott DA, Mills M, Black A, Cantwell M, Campbell A, Cardwell CR, et al. Multidimensional rehabilitation programmes for adult cancer survivors. Cochrane Database Syst Rev 2013;3:CD007730.
- Hansen HP, Tjornhoj-Thomsen T, Johansen C. Rehabilitation interventions for cancer survivors: The influence of context. Acta Oncol 2011;50:259–64.
- Faller H, Schuler M, Richard M, Heckl U, Weis J, Kuffner R. Effects of psycho-oncologic interventions on emotional distress and quality of life in adult patients with cancer: Systematic review and meta-analysis. J Clin Oncol 2013; 31:782–93.
- Buffart LM, van Uffelen JG, Riphagen II, Brug J, van Mechelen W, Brown WJ, et al. Physical and psychosocial benefits of yoga in cancer patients and survivors, a systematic review and meta-analysis of randomized controlled trials. BMC Cancer 2012;12:559.
- Wurtzen H, Dalton SO, Elsass P, Sumbundu AD, Steding-Jensen M, Karlsen RV, et al. Mindfulness significantly reduces self-reported levels of anxiety and depression: Results of a randomised controlled trial among 336 Danish women treated for stage I-III breast cancer. Eur J Cancer 2013;49:1365–73.
- Piet J, Wurtzen H, Zachariae R. The effect of mindfulness-based therapy on symptoms of anxiety and depression in adult cancer patients and survivors: A systematic review and meta-analysis. J Consult Clin Psychol 2012;80: 1007–20.