Clinical implications
To increase the effectiveness of interventions, various recommendations were made by the authors to guide either clinical practice or future studies (see Table 3). Some recommendations were made regarding important elements that should be included or considered, including supporting participants’ self-efficacy, which is defined as an individual’s self-judgment of their own abilities to execute certain behaviours that determine the likelihood of a person engaging in or avoiding certain behaviours [38]. Indeed, interventions that address self-efficacy elements may increase children or survivors’ confidence in their abilities to initiate and maintain healthy behaviours [28]. For example, it was found that with increased self-efficacy, children with cancer are more likely to adopt and maintain regular PA [30]. Also, psychological variables are deemed important intervention targets to improve quality of life [25]. For example, cognitive deficits such as information processing problems should be addressed [29].
The engagement of families, especially parents is particularly important [34]. Indeed, caregivers serving as role models of healthy eating and exercise facilitate family changes [33]. Positive social interactions and encouragement with the mentor was found to be associated with greater adherence to healthy behaviours in one of the studies; the authors therefore recommended enlisting the support of parents or friends to provide additional social support to the CCS [31]. Furthermore, authors support more frequent contact with the interventionists as participants seem to need the continuing support on the lifelong commitment to healthy behaviours to moderate the late effects of therapy [27].
Another popular recommendation was to use targeted, individualized programs and age-appropriate approaches, as they might be better than standard or generic programs to increase the effects of the program [21, 23, 25, 32]. Making the messages specific to the needs of young cancer survivors would likely enhance their relevance as well as increase the engagement and satisfaction of survivors with the content [24].
Multimodal approaches have contributed to compliance and overall success in reaching targeted outcomes [29]. Indeed, multimodal seemed preferable to education only, as provision of information related to adverse outcomes is necessary, but was found to be not sufficient to produce significant changes in participants’ health outcomes [22]. Furthermore, the use of technology offers a feasible, relatively low-cost alternative to more in-person intensive interventions in this at-risk but sparse population because it can be distributed across time and geography; however, personal contact also appeared to help compliance with protocol and follow-up [29].
Finally, regarding the point over the course of treatment and survivorship that interventions might be best implemented, Cox et al. [27] concluded that it may not be feasible during early treatment owing to the child’s responses to the disease and treatment. Indeed, shortly after the diagnosis and following the participants through the end of therapy may have resulted in the study participants and/or parents being too sick and/or overwhelmed to complete the intervention with the frequency and intensity necessary to improve function. They suggested offering study participation at maintenance, in contrast to within 10 days after diagnosis, so that parents and children will have had a considerable period of time to adjust to the diagnosis and to have hurdled the worst treatment and disease effects. Fear, anxiety, symptoms, and overall well-being will no longer be the significant threats that they were at diagnosis, thereby potentially facilitating greater participation and commitment to the intervention. Another early lifestyle intervention concluded that it was feasible, but that the results did not show significant changes in the targeted outcomes [34]. For survivors, it was found that trying to recruit after treatment was difficult as families are often trying to forget their cancer and hospital experiences and, similarly, too long after (e.g. more than 3 years) was also difficult as families are likely to have created a new normal. The average time since treatment ended for participants recruited in their study was 1.96 years [33].
DISCUSSION
The present scoping review sheds light on the dominant areas of research in terms of target population, intervention type, and outcomes for complex interventions or multimodal programs targeting healthy behaviours in children with cancer. Due to a paucity of studies, and the heterogeneity of the studies included in this review, identifying the best health promotion interventions or program is challenging. The interventions typically focused on PA alone or in combination with other health behaviours, the majority had a short duration (<6 months), and few conducted long-term assessments. Furthermore, no studies included families of children younger than three years of age and few recruited patients who were still undergoing treatment. There are a lot of challenges that can be encountered in developing and administering interventions with children who are severely ill, which could explain why so little intervention effort has been made.
The few studies published demonstrate that it is feasible to implement these types of interventions or programs and that it can potentially improve patient outcomes or healthy behaviours practises. Furthermore, popular recommendations include the individualization of interventions, the use multiple modalities and the engagement of families. The results of this study are consistent with other reviews; one concluding that diet and exercise interventions are feasible [39] and another one that lifestyle technology-based interventions demonstrated high feasibility and acceptability rates in the pediatric oncology and survivors of childhood cancer populations [40]. Moreover, a recent review of health behaviour in CCS interventions confirmed a gap in interventions designed for younger cancer survivors (<8 years of age) and that engaged parents, who are critical to child behaviour change [41].
A crucial aspect in the evaluation of complex behavioural interventions is the choice of outcome measures. Precise and reliable measures of both behaviours targeted by the intervention and the health outcomes should be included for a comprehensive evaluation of behavioural interventions. Esbenshade and Ness [39] have suggested to pick out outcomes that are reliable and reproducible as well as to standardize outcome measures so that studies can be compared and combined. Furthermore, long-term follow-up may be needed to determine whether the downstream effects on the health outcome predicted by the change in behaviour occurred or whether the short-term changes, such as the change in behaviour, persisted [42]. Thus, implementing programs of longer duration should be encouraged. Collecting data over an extended period of time (e.g. more than 12 months after starting the intervention) would allow clinicians or researchers to evaluate the long-term effects and benefits of the interventions or program on outcomes such as quality of life, which requires an extended period to respond the intervention compared with other outcomes, for example levels of PA.
Consistent with the recommendations from Stern [33] and Zhang [34] to engage families in behavioural interventions, a systematic review that examined parental involvement in diet and PA studies in CCS concluded that adding a parental component may improve health promotion interventions for CCS [36]. Indeed, in their review, studies with no or indirect parental involvement had lower amounts of success than studies with direct parental involvement. Furthermore, as some research has shown that parents remain involved in the healthcare of survivors even into adulthood [43], parental involvement should be considered throughout the continuum of care and regardless of the survivors’ age. Parents of CCS are also known to experience psychosocial issues related to the child’s cancer and treatment that can affect their family life [44], such as being overprotective and restricting children from participating in physical activities [45] or adopting a parenting style that is associated with increased junk food consumption [46]. Therefore, addressing parenting practises and offering psychosocial support or training to families may be beneficial for both parents and children with cancer.
According to Short et al. [47], programs addressing the promotion of health behaviours such as PA should be theory-based in order to maximise effectiveness. Indeed, there is a growing recognition that theory-based behaviour change interventions are more effective than approaches that do not rely on theory [48]. Among other reasons, theory-based interventions can help pinpoint what factors need to be considered before developing a program or intervention, it can provide insight into how to shape strategies to address these factors, and can provide information on how or why the intervention works. The theoretical framework that has been used the most in the field of pediatric oncology is the Social Cognitive Theory (SCT) [49], which is a theory that has shown particular promise also in the field of PA promotion for breast cancer survivors [47]. However, the application of SCT or any other theoretical framework to promote healthy behaviours in oncology is still in its infancy and further research is needed to optimize their use.
Interventions to change health behaviours such as adopting a healthy diet and practicing frequent PA have huge potential to make a positive impact on both the individual themselves and the overburdened healthcare system [50]. Furthermore, there is growing evidence that multiple-behaviour interventions promise to have much greater impact on public health than single-behaviour interventions [51]. The key idea that underpins behavioural change is to deliver interventions comprised of various change-generating techniques in order to promote a change in behaviour, which will subsequently have an effect on health outcomes. The same rigorous experimental methods that have characterized drug-based clinical treatments should be applied to those studies, which is unfortunately not always the case. Indeed, methods must be devised to measure what was previously thought to be unmeasurable and assessments conducted in a reproducible and valid fashion, even for the measures of subjective states [52].
Limitations from this review include the fact that this is a developing field of research, with the oldest article published in 1999, with few studies addressing the research question. The high variability in intervention type and outcome measures across studies made comparison of results difficult and meant we were not able to identify attributes of studies that were effective. Furthermore, many of the included studies had small sample sizes, biases and short follow-up duration. This review underlines the need for further well-designed trials using standardized outcome measures to be implemented in this population as well as the gaps in the evidence base.