Feasibility and Relevance of an Intervention with Systematic Screening as a Base for Individualized Rehabilitation in Breast Cancer Patients: A Pilot Trial of the ReScreen Randomized Controlled Trial

Introduction

Breast cancer (BC) is the most common cancer in women worldwide¹ and rehabilitation is essential to ensure optimal recovery. Approximately 7600 women annually are newly diagnosed with BC in Sweden.² Developments in BC diagnostics and treatment have resulted in decreased mortality,^3,4 in Sweden, the 5-year and 10-year survival rates are 92% and 86%, respectively.⁵ The increasing number of survivors and the well-known fact that BC is associated with a wide range of persistent, disabling side effects, such as fatigue,⁶ depression,⁷ fear of recurrence,⁸ reduced quality of life,⁹ pain,¹⁰ and distress¹¹ emphasize the need for rehabilitation interventions. It is also known that women with BC have a higher prevalence of psychological symptoms compared to the general population, 5 years after diagnosis,¹² and that between 30% and 40% remain on sick leave 1 year after diagnosis.^13,14 Altogether, this indicates that women are in need of comprehensive rehabilitation following BC treatment.

Cancer rehabilitation aims to prevent and reduce the physical, psychological, social and existential consequences of cancer and its treatment. Rehabilitation efforts should provide support and enable patients and their families to live as well as possible.¹⁵ Rehabilitation research has shown that a variety of different rehabilitation interventions can be effective for known BC-related complications, and that different interventions might be effective for the same problems, depending on how and when the intervention is provided.¹⁶ Previous research has predominantly focused on exercise,^17,18 an important part of cancer rehabilitation. However, as the need for rehabilitation varies greatly between individuals and with the definition of cancer rehabilitation¹⁵ as a foundation, a comprehensive approach to rehabilitation is needed. Also, despite efforts, women report unmet needs.^19–21

Recent research show the need for structures and models to identify individual needs and enable individualized rehabilitation.^22,23 Efforts have been made, for example, a theoretical surveillance model was developed, aiming to facilitate early identification of BC-related physical impairments during the trajectory of medical management.²⁴ Another model was a pilot study testing an individually tailored nurse-navigation intervention, focusing on both physical and psychological symptoms during the BC-trajectory, showing promising results concerning anxiety, depression and distress.²⁵

The patients’ perceived level of distress is one important risk factor for unmet needs following BC treatment.²⁰ Distress, in the cancer context, is a concept that captures the experience of cancer from multiple perspectives. It is defined as a multifactorial experience of a psychological, social, and/or spiritual nature that may interfere with the ability to cope with cancer, its physical symptoms, and treatment.²⁶ About 34–43% of the women newly diagnosed with BC suffer from high distress^27,28 and women newly diagnosed with BC who report high distress may be at risk of developing chronic distress.¹¹ High distress in the initial phase of the BC cancer trajectory is a predictor of chronic distress 8 months later, and persistent high distress at 8 months after surgery has been shown to predict worse psychosocial outcomes as long as 6 years after treatment.²⁹ Distress as a screening tool for rehabilitation needs in a BC context has been tested earlier with promising results.^25,30 However, the level of distress may be culture specific and therefore need to be explored in various contexts.

Research suggest structures for cancer rehabilitation that include a comprehensive assessment of the patients life situation before start of treatment, with an ongoing surveillance during the cancer trajectory, facilitating individualized rehabilitation.³¹ This supports the relevance of early screening for distress at time of primary treatment to ensure individualized rehabilitation in the initial part of the cancer trajectory as well as repeated evaluations throughout the phase of illness to reduce the risk of long-term, BC-related problems. Lack of structures may result in suboptimal recovery, manifested as persistent treatment-related ailments, increased health consumption, long periods of sick leave, and insufficient use of available resources. The ReScreen randomized controlled trial (RCT) was designed to evaluate the effect of screening-based, individualized rehabilitation following primary BC treatment.³² Through pilot and feasibility studies, the likelihood of success of a main study and of valid inference from the research is enhanced.³³ This study aimed to examine recruitment, retention, distribution of distress, relevance of the intervention, and reported problems in a pilot trial of the ReScreen study.

Methods

Design and Brief Description of the ReScreen Main Study

A study protocol of the ReScreen study has been published.³² To ensure transparency and quality, the pilot trial is reported according to the CONSORT extension for randomized feasibility and pilot trials.³⁴

In brief, the ReScreen RCT³² was inspired by the UK Medical Research Council framework for developing and evaluating complex interventions³⁵ and was designed as a three-armed trial. Women who agree to participate are screened for distress using the Distress Thermometer (DT),³⁶ 1–2 weeks following BC diagnosis. The DT is validated in different countries and cultures,³⁷ and has good predictive validity in the context of oncology care in Sweden,³⁸ and is a promising screening tool for identifying extended needs in a BC context.^27,39 The DT includes a 0–10 visual analog scale, where 0 is equivalent to no distress, and 10 is the worst imaginable distress, and a 39-item problem list (PL) with the purpose of identifying problems in five categories: practical, family, emotional, spiritual/religious, and physical.³⁶ The PL has been validated in a Swedish colon cancer context.⁴⁰ Those who report high distress are randomized to either the intervention group (IG) or the control group (CG), while women with low distress form an observational group (OG). All groups (the IG, CG and OG) receive “care as usual.” In addition, the IG is monitored proactively by a research nurse (nurse navigator), for 1 year to facilitate evidence-based individualized rehabilitation supported by a decision support tool (DST), described in the study protocol,³² developed within this project to guide the research nurse. The DST focuses on areas of commonly reported problems by BC patients as well as health-related behaviours (physical activity, tobacco and alcohol consumption) and is based on research¹⁶ and national documents,¹⁵ and contains both general advice, and advice depending on identified individual needs. The dedicated research nurse will use the DST in conversation with the participant when structuring an individual rehabilitation plan, with explicit goals for the rehabilitation process. Evaluation and follow-up of the rehabilitation plan depends on the participant’s individual needs, but as a minimum at 2 weeks after start of treatment and once a month during the first 3 months.³² The primary outcome for the main study is distress.

Participants and Randomization

In Sweden, cancer rehabilitation is organized on various levels: 1) basic rehabilitation needs, 2) special rehabilitation needs, 3) advanced rehabilitation needs, and 4) very advanced rehabilitation needs. On the basic level, rehabilitation and information are provided by the contact nurse and a physician in a cancer unit, sometimes assisted by a social worker. Special rehabilitation needs could include counselling with social worker during crisis response. For patients with advanced rehabilitation needs a specialized multi-professional cancer rehabilitation team can be utilized.¹⁵ The patient will remain on the first level during the cancer trajectory, unless extended needs are identified.

Participants for this feasibility study were recruited between November 2018 and April 2019 at a BC unit at a university hospital in the south of Sweden. Inclusion criteria were 1) newly diagnosed primary BC; 2) age ≥18; 3) ability to communicate in Swedish; and 4) cognitive ability to participate. Exclusion criteria were 1) pregnancy; 2) psychological disease or drug abuse; 3) previously treated for BC (ie, relapse); and 4) endocrine therapy as only treatment.

All eligible women with newly diagnosed BC were consecutively invited to participate in the study. A contact nurse approached them about participation during a pre-planned visit to the outpatient clinic, approximately 1–2 weeks following the BC diagnosis. In line with Danish³⁹ and Dutch²⁷ studies on BC, a cutoff score of 7 was set to identify patients with moderate/high distress. Patients scoring ≥7 were randomized to the IG or CG. Patients scoring ≤6 were followed longitudinally for comparison (OG).³²

Data Collection

Data were collected at baseline, and 2 weeks and 3, 6, 9, and 12 months after start of treatment. Clinical and treatment-related data was collected through the patients’ medical journal and was manually registered in an electronic case report form. Demographics and outcome data were collected via self-reported questionnaires containing validated instruments and single items focusing on physical, psychological, and behavioural aspects of rehabilitation.³² The questionnaires were sent to the patients by mail and returned in a pre-addressed, stamped envelope. Up to two reminders were sent. The questionnaires were scanned using a system with a control function for identifying missing data.

Feasibility Outcomes

Data Collection Rate and Frequency

“Recruitment rate” was defined as the number and percentage of participants who accepted to participate in the study out of all eligible patients. Reasons for non-participation were not systematically collected for ethical reasons. “Answer frequency” was defined as the number and percentage of patients returning the questionnaires at each given follow-up point for the sample as a whole and divided into subgroups (IG, CG, and OG). “Retention rate” was defined as the number and percentage of patients remaining in the study at each follow-up point for the sample as a whole, and divided into subgroups (IG, CG and OG).

Data Management Outcomes

To identify potential systematic errors, 1) all scanned questionnaires were manually verified and compared to the original data and 2) data were manually compared with the medical journal by two research team members, one of whom was not involved in the data management process.

Distribution of Distress

The distribution of distress was explored in several steps. Firstly, the mean value for distress was calculated in the total sample at each follow-up point. Secondly, the number and percentage of participants scoring ≥7 (cutoff) at baseline was calculated. Finally, alternative cutoffs were explored by calculating the number and percentage of patients scoring ≥4, 5, 6, or 8.

Adherence to Intervention Protocol and Frequently Reported Problems

To ensure adherence to the intervention, a review of the study documentation of the IG was conducted, focusing on 1) the number of contacts with the research nurse; 2) whether all patients had an established rehabilitation plan; 3) whether the rehabilitation goals stated by the patient were clearly documented; and 4) whether the goals were followed up by the research nurse.

To evaluate if the areas in the DST were consistent with the most common patient-reported problems, the PL was used to explore patient-reported problems in the total sample at each follow-up point. The PL was also used to investigate the number of reported problems per patient in the total sample.

Statistical Analyses

Descriptive statistics in terms of frequencies, mean, and standard deviation (SD) were used to evaluate outcome data using IBM SPSS Statistics for Windows, version 25.0 (IBM Corp., Armonk, NY, USA). Data on all participants were included in the analyses, except for data giving the number of problems, where only questionnaires with complete data on the PL at baseline were included.

Results

A total of 117 patients were found to be eligible and were invited to participate in the study. Eighty-five patients agreed to participate and signed informed consent, resulting in a recruitment rate of 73% (Figure 1). The mean age in the total sample was 62.8 years and the majority were married/cohabiting (n=57; 67.1%). Half of the participants had two children (n=43; 50.6%) and almost 86% were born in Sweden. About 9.5% were on sick leave at baseline. The majority reported having a good/very good financial situation (n=69; 82.1%) and almost half of the participants had a university degree (n=39; 45.9%). More than half of the participants used no or only small amounts of alcohol (n=49; 61.3%), and almost everyone was non-smoking (n=78; 94%). A large proportion of the participants stated that they had less than 30 minutes of exercise/week (n=40; 47.1%), but the majority reported other physical activity of more than 90 minutes/week (n=53; 62.4%) at baseline (Table 1).

Table 1 Demographic Characteristics of the Participants

Figure 1 Flowchart of the recruitment process, randomization and answer frequency.

A total of 425 forms were sent out during the 1-year follow-up, of which 392 were returned, resulting in an answer frequency of 92% (Figure 1). The answer frequency decreased over time from 85 questionnaires at baseline to 81 (95%) at 2 weeks, 79 (93%) at 3 months, 81 (95%) at 6 months, 76 (89%) at 9 months, and 75 (88%) at 12 months. Divided into the groups IG, CG, and OG, 44 out of 45 (98%) forms were sent back in the IG, while 29 out of 45 (64%) and 319 out of 335 (95%) were sent back for the CG and the OG, respectively. The retention rate for the total sample was 99% at 2 weeks (n=78), 98% at 3 months (n=75), 95% at 6 months (n=76), 93% at 9 months (n=72), and 88% at 12 months (n=72). The retention rate at 12 months was 100% in the IG (n=9), 56% in the CG (n=9), and 91% in the OG (n=67).

Data Management

Three systematic errors were identified. Two were related to the scanning procedure, namely 1) the scanning system in some cases interpreted handwritten numbers inaccurately, resulting in incorrectly recorded patient identification number; and 2) the system did not interpret commas; for example, 78.9 kg was interpreted as 789 kg. 3) The third error was related to the timing of data extraction from the medical record as the treatment regime could change during the process, for example, when a patient declined recommended adjuvant treatment after visiting the oncologist.

Distribution of Distress

The mean value for distress in the total sample (n=85) was 3.99 (SD 2.67) at baseline declining to 2.78 (SD 2.40) at 12 months (Figure 2).

Figure 2 Mean distress scores and standard deviation (whiskers) at the pilot trials data collection points.

With the cutoff set at ≥7, 18 participants (21.2%) were identified as having high distress at baseline and were randomized to either the IG (n=9) or the CG (n=9). Exploring distribution of distress at baseline using different cutoffs showed that 49 participants (56.9%) scored 4 or higher, 38 (44.2%) scored ≥5, 29 (33.7%) scored ≥6, and nine (10.4%) scored ≥8.

Adherence to Intervention

Of the nine participants in the IG, eight accepted to take part in the intervention. The eight participants had at least the minimum amount of planned contact (one occasion per month during the first 3 months). All of the participants had a rehabilitation plan documented in the medical journal, with clearly stated and followed-up rehabilitation goals. Advice concerning, for example, physical activity or sleep was documented, followed up, and evaluated.

Frequently Reported Problems

The most frequently patient-reported problems in the total sample (n=85) were in the emotional and physical domains of the PL. Five problems were among the most frequently reported problems at all time points (Figure 3): Almost 68% of the participants reported “Worry” as a problem at baseline, decreasing to 38.5% at 3 months, and increasing to 48% at 12 months. “Fatigue” increased from 32.5% at baseline to 67.5% at 3 months, but decreased to 44% at 12 months. “Sleep”, “Sadness” and “Memory/concentration” remained stable over time (Figure 3). The most common problems were consistent with the areas in the DST.

Figure 3 The most common problems in percentage (%) reported at the pilot trials data collection points.

Some problems were more frequently reported at different follow-up points. “Nervousness” and “Fear” were frequently reported in the beginning of the trajectory (baseline) (41% and 38.6%, respectively) and at 2 weeks (38.6% and 28.7%); “Dry skin/itchy” was reported by 44.7% at 3 months. “Getting around” was reported by 30% at 9 months. “Pain” was more frequently reported at 2 weeks (26.9%) compared to baseline, but was most frequently reported at 9 months (29.7%).

The number of reported problems at baseline ranged from 0 to 27. Having three problems were most frequently reported (n=15; 20.5%). Fifty-five participants (75.3%) reported having no problems to 5 problems, 14 (19.2%) reported 6–12 problems and 4 (5.5%) reported 19 problems or more.

Discussion

This feasibility study examined the data collection and data management process of a pilot trial of the ReScreen RCT. The study also explored the distribution of distress for comparison with international studies and assessed the relevance of the intervention. The results show that the processes are working well, with satisfactory recruitment rates, answer frequency, and retention rates. Few systematic errors were identified, and they were easily corrected. The results also showed that fewer women in this sample scored ≥7 at baseline compared to international studies. After adjustment of the cutoff, the result facilitates the decision to move on to the full-scale study.

Recruitment to RCTs is known to be difficult and the recruitment rate can affect the results.⁴¹ Previous studies highlight the importance of organization, training, and support to facilitate recruitment to RCTs.⁴² The recruitment rate of 73% in this study is in line with earlier studies in a similar context,³⁰ which indicates well-functioning procedures for recruitment. In the initiation of this study, several preparatory meetings were held with the contact nurses at the outpatient clinic recruiting potential patients. The meetings aimed to enhance compliance, by clarifying inclusion and exclusion criteria and emphasizing the importance of inclusion. It is known that the way recruiters communicate about a trial is important for recruitment, as is a person-centered approach.⁴³ Therefore, the meetings discussed how the study was to be presented to the patients, in order to facilitate recruitment. A dedicated research nurse was available on a daily basis to support recruiters whenever they had questions about recruitment or the study. In other studies, the importance of support^44,45 and the presence of research staff⁴⁵ for recruitment has been highlighted as an important factor to enhance recruitment.

The comprehensive support to the recruiting contact nurses may also be a possible explanation for the high answer frequency and retention rate. Research on retention has highlighted the importance of good strategies within the trial plan, and also the importance of a good relationship with the participants.⁴⁶ Since the recruitment situation was the only occasion when the participants in the CG and OG formally received information and had time to ask questions about the study this situation became very important, which was emphasized at the preparatory meetings. In this study, 93% of the questionnaires were returned. The recruitment procedure, together with the procedures for data collection that included pre-addressed, stamped envelopes and up to two remainders, contributed to the answer frequency and facilitated retention, which indicates good strategies and well-functioning procedures.

The lower answer frequency in the CG compared to the OG probably relates to the fact that the participants who were randomized to the CG were more stressed by their situation than those in the OG, as they reported ≥7 on the DT. The answer frequency was also lower in the CG compared to the IG group, which can be related to the fact that they did not receive the enhanced support as did those who were randomized to the IG. The IG was followed continuously by the dedicated contact nurse, and the good outcome of the individualized rehabilitation support in this group reflects a good relationship and the participants’ satisfaction with the support, both of which were described as important in previous research.⁴⁶ In this study, the low number of participants in the IG and CG groups prevents any definitive conclusions, but the outcome of the pilot trial is important to take into account in the main study.

The cutoff of 7 was initially set in line with Danish²⁸ and Dutch²⁷ studies using the DT in a similar context, where 43% and 34%, respectively, scored ≥7 on the DT, indicating moderate/high distress for women with newly diagnosed BC. In this study, 21.2% of the women scored ≥7 on the DT, a lower percentage than expected. International research shows a mean distress score of 5.4²⁸ and 5.5⁴⁷ for women newly diagnosed with BC, compared to 3.99 in this study. It is evident that women in this Swedish context had a lower distress score than women in other BC contexts, and the reason for this is unknown. When exploring different cutoffs and lowering the cutoff to 5 or 6, the same frequency (34% and 44%, respectively) as in other studies^27,28 was seen. This indicates that a lower cutoff is feasible in this context, which will lead to an increased number of participants in the IG and CG. To enable evaluation of the optimal cutoff in a Swedish context, the cutoff will be reviewed in the main study.

To ensure the relevance of the DST, the most common patient-reported problems in this study were compared with earlier research. The most reported problems in this sample are known problems in patients with BC, such as worry,²⁸ fatigue,⁶ sleep and concentration problems⁴⁸ and sadness.²⁸ Over time, some problems fluctuated while others remained stable. Previous research has shown that some patients with BC have increased levels of distress and associated problems in the re-entry to the survivorship phase.¹¹ This study showed a marked increase in worry at 12 months, in agreement with a longitudinal study exploring anxiety, indicating that almost 30% of participants were suffering from persistent or newly developed anxiety at 12 months after surgery.⁴⁹ For the majority of the patients in the Swedish BC care context, the treatment phase is completed at 12 months and no more physical follow-ups are planned. This is a likely cause of their increased perceived distress, which indicates that they may benefit from increased support at this stage. However, this needs to be explored in future studies.

Long-term follow-up studies report that fatigue, and cognitive and sleep problems can lower the quality of life for BC patients for as long as 5 years after diagnosis,²¹ and high distress during the first year after diagnosis can predict worse psychosocial outcomes 6 years later.²⁹ This knowledge clarifies the need for early identification and long-term support and rehabilitation to avoid persistent treatment-related ailments. Together with the finding in this study showing a wide range of reported problems, the pathway for BC patients seems to be complex, with different problems appearing at different time points during the trajectory. Women with BC report unmet needs,^8,19 indicating that even if some problems are reported more frequently on a group level, the need for a complex intervention is evident, with the possibility to identify needs and act upon them individually, where the patients’ preferences are combined with evidence-based interventions. The individualized approach is also important considering the known fact that one problem can have different origins and that different interventions can be effective depending on individual preferences.¹⁶ This feasibility study shows that the reported problems are in line with earlier research, indicating that the RCT is feasible and relevant.

Strengths and Limitations

According to the CONSORT statement,³⁴ a feasibility study aims to investigate if and how a future trial should be performed. The final sample was deemed adequate for assuring evaluation of procedures, and relevance of the intervention. The large number of participants in this pilot trial is a strength and validates the results. A limitation is the exclusion of certain groups of women with BC, such as those with relapse and those unable to communicate in Swedish. Since people with language difficulties, such as migrated women, are often excluded from research, future studies are needed to explore the BC care trajectory also in these women.

Conclusion

The satisfactory recruitment rate, answer frequency, and retention rate, together with few systematic errors, show that the data collection and management in the ReScreen randomized controlled trial is feasible, with well-functioning procedures. The answer frequency and retention rate in the control group were lower than in the intervention group and observational group. This will be monitored actively in the main study. The difference in distribution of distress in this sample compared with international studies, leading to few participants in IG and CG, indicates a need to adjust the cutoff to identify patients in need of extended support in the main ReScreen randomized controlled trial and to explore the optimal cut-off in this context. Perceived problems reported by the participants are in line with previous studies and show the relevance of the planned complex intervention.