Rethinking Scale-Up of Rehabilitation for Chronic Disease in Low-Resource Settings: Embracing Complexity for Contextual Impact

Identification of literature

We used conventional systematic methods to identify literature in the niche area of exercise-based rehabilitation AND chronic disease populations (NCDs specifically) AND low-resourced settings.

To identify literature, we pragmatically utilized the literature identified through two systematic scoping reviews our team previously published within the context of NCDs, rehabilitation, and low-resourced settings (910). Firstly, a scoping review of 60 exercise-based rehabilitation studies for patients with NCDs (Heine et al. 2019). Second, a scoping review on 48 rehabilitation studies in self-reported, low-resourced settings. This review aimed to unravel the concept of low-resourced settings when rehabilitation is concerned (9). Studies included in the latter review (van Zyl et al. 2021), which focused on NCDs (using the same criteria as outlined in Heine et al. 2019), were eligible for inclusion in this review and its qualitative analysis. Subsequently, both reviews were updated (11 March 2022) using conventional procedures for study selection (i.e., screening of title, abstract, and full text by two independent reviewers).

Qualitative synthesis

We then conducted a reflexive thematic analysis by deductively and inductively coding the text of the selected articles (11), including the introduction up to the discussion section (i.e., entire text body), using the socioecological model (SEM) to guide the deductive analysis. The SEM is particularly useful when determinants of an outcome are nested within different levels of influence (12). Once all articles were coded, we merged all codes into content categories and themes (i.e., key actors) through a series of team meetings. Importantly, this reflexive qualitative synthesis was conducted twice. First, on the set of articles identified within the two previous reviews (i.e., the main analysis). Subsequently, the analysis was replicated in a subset of articles (20%) purposefully selected from the eligible articles identified when updating the search for these two reviews (i.e., validation). Selection was based on potential gaps (e.g., geographical representation, type of chronic condition) in the original analysis and supplemented with a random selection if applicable. This second analysis in a new set of articles allowed to appreciate the level of thematic saturation (i.e., no new were identified) achieved. If no saturation was achieved, a subsequent new set (20%) of articles would have been selected to repeat the analysis for a third time (and so forth) until saturation was achieved (i.e., no new themes identified); however, this was not applicable.

Finally, we developed a preliminary conceptual synthesis of the themes derived from the thematic analysis by reflecting on the interplay between the variables using concept mapping (13). Within the qualitative paradigm of the project, there was less emphasis on the ‘accuracy’ of the data. The research team embraced their diversity (age, gender, profession, and nationality). Both the reflexive thematic analysis and the development of the concept map were completed through a series of team meetings and robust discussions.

Concepts

Scale-up

Scale-up is the ‘efforts to increase the impact of health interventions so as to benefit more people and to foster policy and programme development on a sustainable basis’ (15). We’ve tailored the scale-up framework for an integrated care package (16) to a three-dimensional scale-up framework that applies to rehabilitation:

1. Strengthening integration within primary care structures, including integrating rehabilitation care for chronic disease within existing programs for communicable disease management (e.g., integrated care for diabetes in patients with HIV/AIDS) and community-based rehabilitation models. Primary care is a key model of care that supports first-contact, accessible, continuous, comprehensive, and coordinated person-focused care (17), often at community level. Strengthening the integration of rehabilitation into primary care has been considered a key strategy for increasing access to rehabilitation (7);

2. Increasing the comprehensive nature of existing programs to ensure more core components of rehabilitation are included, including linkage of healthcare and social care to address social determinants of health; and

3. Set-up of new programs or program optimization to increase the number of people that have access to rehabilitation services (Figure 1).

Figure 1

The on-the-ground realities in low-resource settings (9) add real layers of complexity to the scale-up of rehabilitation as to mitigate the burden of chronic disease.

Study selection for data saturation

A bespoke PRISMA-inspired flowchart is provided in online supplement 1 to showcase the selection process. A full reference list and consolidated overview of all 144 papers (Original [81] and updated [63]) can be found in online supplements 2 and 3, respectively. Thirteen articles (20%) were purposefully selected from articles identified during the search update. These articles were selected based on key areas underrepresented in the main analysis being conducted in a low-income country (n = 5) and patients with cancer (n = 4), supplemented by a final random selection (n = 4).

Reflexive Thematic analysis

The initial analysis resulted in 362 codes, drawn from 81 articles. Subsequently, these codes were merged into 168 categories and 40 themes. In the analysis of the purposefully selected 13 articles, another 51 codes and one additional category were identified (i.e., seasonal fluctuations in resource availability). The single additional category did not warrant additional themes, and therefore the team accepted that data saturation had likely been reached. For the total set of article analyses (n = 81 + 13), six were conducted in low-income countries, 55 in lower-middle income countries, 17 in upper-middle income countries, and 15 in explicit low-resourced settings within high-income countries. A description of each theme, linguistically drawn from the underlying codes and categories, can be found in Table 1. For each theme, the main spheres of influence, according to the socioecological model, are provided (Figure 2). The themes are clustered into ‘sub-systems’ (inner circle, Figure 2) based on the SEM model. The interplay between the nine subsystems that directly or indirectly may inform program optimization to promote program integration, comprehensiveness, and/or reach is depicted in a concept map (Figure 3).

Figure 2

Figure 3

Integrated Rehabilitation

The first dimension of scale-up is integration of care within existing health structures. Within the rehabilitation space, we could discern three forms of ‘integration’ that are particularly relevant to address the rehabilitation need. Firstly, integration of rehabilitation care for multiple conditions simultaneously to address multimorbidity or converging burdens of disease (18). Secondly, integration of rehabilitation care with other social domains that therefore can target social determinants of health (19). Thirdly, integration of rehabilitation within primary care structures (20). The value of each of these three becomes apparent when reflecting on the critical factors and subsystems identified in this paper. In many low-resourced settings, multiple burdens of disease collide with human resources (i.e., quantity, adequate training) and health system governance unequipped to tackle the convergence of, for example, communicable and non-communicable disease or diabetes and hypertension, as in our case example (21222324). Hence, within rehabilitation care for chronic disease, too, we may need to move beyond disease-siloed programs (e.g., cardiac rehabilitation, pulmonary rehabilitation) towards person-centered and goal-oriented care models and provide real-world evidence in their support (25). In providing such support, some of the key questions for rehabilitation in low-resourced settings remain with respect to implementation, adequate monitoring and evaluation to inform policy, cost-benefits, and impactful implementation (2627).

Furthermore, there is clear evidence that factors outside of healthcare delivery greatly affect an individual’s health and well-being; this is particularly true for low-resourced settings with a high prevalence of adverse social determinants of health. In addition to the impact of social determinants of health on health and well-being, these determinants also directly impact the implementation of rehabilitation care (2829). Some critical factors and social determinants of health identified in this reflexive thematic analysis were, for instance, health literacy, health-seeking behavior, poverty, productivity, and access to transport. Such social determinants of health may inform both the burden of disease through underlying risk factors ((30)), financial safety net to engage with rehabilitation services, and willingness or ability to make lifestyle changes that can contribute to reduced risk. Hence, rehabilitation-led models of integrated rehabilitation and social care (e.g., case managers) that comprehensively address social determinants of health can make a remarkable difference in the burden of disability and health inequities (19). Awareness of individuals’ social needs, adjustment of care provided, assistance for the individual to have needs met, alignment of efforts with community and social care organizations, and advocacy for changes to infrastructure and policy (both health and social) may facilitate such rehabilitation-led models (19).

Primary care is a model of care that supports first-contact, accessible, continuous, comprehensive, and coordinated person-focused care. The World Health Organization has specifically called for strengthening and scale-up of interventions within primary care in their draft 2023–2030 roadmap on the prevention and control of NCDs (31). Various health systems have approached the integration of rehabilitation into primary care differently, reflecting differences in the organization of care and health governance (32). Six different ways of integrating rehabilitation in primary care have been described, including clinic-based, outreach, self-management, community-based rehabilitation, shared care, and case management (20), Each of which emerges as possible models in relation to the subssystems identified in this reflexive analysis, including the geopolitical context (e.g., rural context informing an outreach model), human resources (e.g., task-shifting, self-management), and health governance (e.g., resilient health systems, referral pathways). The work presented here, in conjunction with various models for integrating rehabilitation into primary care, including social care where applicable, and cognizant of multimorbidity, could help inform program optimization and scale up initiatives.

Comprehensiveness

Integrated models of rehabilitation and social care already speak to a more comprehensive offering of rehabilitation beyond those common core components (e.g., physical inactivity, nutrition) strictly related to the underlying risk factors for chronic disease. In scaling the comprehensive nature of rehabilitation, one could argue that human resources are central, including both quantity and quality (e.g., adequate training). Considering social determinants of health and integration within primary care, such human resources may not be restricted to allied health professionals like physiotherapists and also include non-medical human resources (e.g., community champions, community health workers, caregiver-led models). To optimize the human resources required to provide quality rehabilitation for chronic diseases in low-resourced settings, one needs to ask a couple of questions. First, what risk factors are underlying the burden of disease in my specific context, and how may that inform the human resources required (i.e., training needs, composition of the rehabilitation team, integration with social care)? Second, what non-medical human resources are available in my specific context, and how can we optimize program delivery by actively engaging these resources (e.g., task-shifting, self-management)? Importantly, the process of task-shifting in itself may be complex, with important roles for health literacy in self-management models (33) and the community’s social fabric in the feasibility of training lay community workers (34). Thirdly, how can digital innovation be used equitably to promote access, uptake, and/or adherence of technology-driven rehabilitation models, and how does this affect the competencies required of those overseeing such rehabilitation services? The implications thereof, for example, may be that a single physiotherapist oversees a cadre of community health workers for coordinated physiotherapy using a digital telehealth platform rather than solely providing face-to-face clinical services. In this light, we argue that it is pivotal that while we train the next cadre of rehabilitation professionals (35), we embrace the competencies needed to set up and coordinate such task-shifting models as a means to optimize rehabilitation delivery.

Increasing access

This paper provides an overview of the various critical factors that may be considered in strengthening existing and scale-up of new rehabilitation programs. However, addressing a single factor is unlikely to result in the provision of quality rehabilitation care, as each factor should be recognized as equally important yet equally inadequate if addressed in isolation. We argue that the critical factors, the nine subsystems, in conjunction with existing frameworks (e.g., WHO rehabilitation competency framework) that relate to these subsystems can assist researchers and policymakers in making informed decisions in scaling-up rehabilitation or in conteptualizing the research and information needs. What becomes apparent from this work is that upscaling rehabilitation is complex and requires intersectoral and interdisciplinary impetus across the continuum of care.