Bridging the gap: Enhancing external validity in rehabilitation research through pragmatic and hybrid trial designs

Inclusion criteria and population diversity

Stringent eligibility criteria enhance internal validity but simultaneously erode external validity by filtering out patients with multimorbidity, cognitive impairment, or culturally diverse backgrounds. Even when criteria are broad, homogeneity may persist owing to self-selection bias, regional demographics, and recruitment strategies. Intentional, equity-focused enrolment–backed by community partnerships and culturally tailored materials–is therefore essential to realize the inclusive potential of pragmatic designs.^[7]

Over-reliance on laboratory-based outcome measures

Traditional rehabilitation trials frequently employ objective, laboratory-based outcome measures to assess treatment efficacy. While these measures provide precise, quantifiable data, they may not always capture the patient’s functional improvements in daily life. Patient-reported outcome measures (PROMs), which evaluate usual clinical environments, pain, and quality of life, are often underutilized in explanatory trials. PROMs are valuable along the entire explanatory–pragmatic continuum; their contribution depends less on the instrument itself than on how and when it is implemented. In explanatory trials, PROMs often capture hypothetical or experimental function, whereas pragmatic trials should prioritize PROMs that reflect enacted function–the patient’s actual performance in everyday contexts. Selecting instruments that map onto the enacted tense (Glass 1998)^[8] reduces the evidence–practice gap and helps decision-makers judge typical healthcare contexts. Pragmatic trials, in contrast, integrate both objective and patient-centered outcomes, ensuring greater applicability to clinical practice.

Highly controlled clinical environments

Rehabilitation interventions studied in RCTs are often delivered in highly controlled clinical environments with strict adherence to standardized protocols. Although this approach enhances internal validity, it fails to account for the variability encountered in rehabilitation settings, such as differences in therapist expertise, patient adherence, and available resources. Pragmatic trials accommodate such variability, making their findings more reflective of actual clinical practice.

Small sample sizes and underpowered studies

Many rehabilitation studies involve relatively small sample sizes, leading to insufficient statistical power and limiting the ability to detect clinically meaningful differences. Adequate power is a universal requirement: underpowered studies threaten validity irrespective of where they fall on the explanatory–pragmatic spectrum. Therefore, sample size should be interpreted within the broader methodological rigor rather than as a distinguishing feature between explanatory and pragmatic trials.

By contrast, this limitation is especially pronounced in highly specialized rehabilitation domains, where patient recruitment is inherently challenging. In these “niche rehabilitation fields,” patient populations are often small or very specific, such as those involving rare neurological disorders (e.g., spinal cord injury rehabilitation, rare genetic ataxias) or targeted interventions for subgroups such as amputee rehabilitation or advanced dementia care. These constraints make it difficult to achieve adequately powered studies, increasing the risk of type II errors and potentially leading to underestimation of treatment effects.

Pragmatic trials: A solution for enhancing generalizability

Pragmatic trials differ from traditional RCTs in several key aspects, allowing for a more representative patient population and a more realistic assessment of treatment effectiveness.^[9,10] The pragmatic–explanatory continuum indicator summary (PRECIS-2) framework^[11] categorizes trials along a continuum, with fully explanatory trials at one end and fully pragmatic trials at the other. By employing pragmatic methodologies, researchers can better bridge the gap between research and practice.

Key features of pragmatic trials in rehabilitation

1. Broad inclusion criteria - Unlike traditional RCTs, pragmatic trials aim to recruit a diverse patient population that closely resembles real-world cases. This ensures that findings are applicable across a wider spectrum of patients.

2. Flexible interventions - Rather than rigidly controlling interventions, pragmatic trials permit individualized treatment adjustments based on clinical judgment, thereby mimicking typical healthcare contexts’ practice. While pragmatic trials allow for some flexibility to mirror clinical practice, it is essential that such flexibility operates within the boundaries of a clearly defined intervention protocol. Protocols should specify permitted ranges of individualization, predefined adaptation rules, and structured monitoring mechanisms to ensure intervention fidelity across different providers, settings, and patient populations. Maintaining treatment fidelity is crucial to accurately evaluate the effectiveness of an intervention, even in pragmatic trial designs

3. Clinically relevant outcomes - Emphasis is placed on functional recovery, patient-reported outcomes, and long-term benefits rather than surrogate markers

4. Heterogeneous treatment settings - Interventions are tested across multiple clinical environments, including outpatient, inpatient, and home-based rehabilitation settings

5. Longitudinal follow-up - Pragmatic trials often incorporate longer follow-up periods to assess the sustainability of rehabilitation outcomes over time

6. Adaptability to healthcare system variability - Pragmatic trials account for differences in healthcare infrastructure, therapist availability, and regional variations in care delivery.

Examples of pragmatic trials in rehabilitation

Recent studies demonstrate the benefits of pragmatic designs in rehabilitation research. For instance, a pragmatic cluster RCT conducted by Ford and Norrie^[12] compared different stroke rehabilitation strategies in multiple clinical settings, enrolling patients with a broad range of comorbidities. The study found that the pragmatic approach provided a more accurate reflection of rehabilitation effectiveness compared to previous highly controlled RCTs. Similarly, a stepped-wedge trial in post-operative rehabilitation for total knee arthroplasty patients demonstrated superior external validity by integrating standard rehabilitation protocols with patient-specific modifications.^[6]

Implications for clinical practice and future research

The adoption of pragmatic methodologies in rehabilitation research has several implications:

• Enhanced translation of research into practice - By aligning study conditions with clinical environments, pragmatic trials facilitate the implementation of evidence-based interventions.

• More inclusive evidence base - With broader inclusion criteria, pragmatic trials ensure that findings apply to diverse patient populations, including older adults and those with comorbidities.

• Guidance for policy and resource allocation - Health policymakers can rely on pragmatic trial data to inform guidelines, resource distribution, and rehabilitation program development. Moreover, recent adaptations of the Reach, Effectiveness, Adoption, Implementation, Maintenance (RE-AIM) framework have emphasized contextual factors, stakeholder engagement, and equity considerations, offering valuable guidance for embedding implementation outcomes into rehabilitation trials.^[13]

• Integration of artificial intelligence (AI) and Digital Health - Future pragmatic trials should explore the integration of AI-driven rehabilitation tools and telehealth interventions to maximize patient engagement and adherence. The integration of AI and digital health technologies in rehabilitation research should be framed within the context of evaluating—not promoting—these tools under routine care settings. Pragmatic trials offer a methodological advantage by allowing researchers to assess the scalability, usability, and impact of digital interventions across heterogeneous clinical environments. For example, AI-based gait analysis or remote adherence monitoring can be tested within the variability of routine practice, revealing barriers to implementation and real-life effectiveness. Thus, pragmatic designs are well-suited to support the critical evaluation of digital health innovations as they are actually used in clinical contexts.

Future research should focus on hybrid trial designs, which integrate elements of both explanatory and pragmatic approaches to balance methodological rigor with external validity. Hybrid trial designs specifically refer to effectiveness-implementation studies, where the primary aim is to simultaneously evaluate both the clinical effectiveness of an intervention and key aspects of its implementation. Hybrid effectiveness-implementation designs aim to simultaneously assess clinical outcomes and implementation processes. According to Curran et al.,^[14] these are classified into three types: Type 1 (focus on clinical effectiveness with limited implementation data), Type 2 (equal emphasis on both), and Type 3 (primary focus on implementation outcomes while monitoring effectiveness). These typologies provide a structured framework to evaluate the effectiveness of an intervention and how it can be adopted, sustained, and scaled in routine care settings. More detailed application of hybrid designs has been documented in rehabilitation contexts, such as the implementation of community-based stroke care or remote physical therapy in underserved areas.

Such designs accelerate the translation of evidence into practice by addressing both clinical efficacy and adoption challenges.^[15]

In addition, the use of evidence Real-World Evidence (RWE) and big data analytics in rehabilitation research may further improve the generalizability of findings by leveraging large-scale, observational data. Analytically, pragmatic trials often tolerate more flexible study designs and may incorporate observational elements such as registry-based follow-up or data integration. Causality in pragmatic trials is frequently addressed using advanced statistical modeling techniques – including propensity score analysis, mixed-effects models, or network meta-analysis – to account for variability while still estimating the intervention’s effects on clinical outcomes.

To accelerate this methodological shift, we propose three concrete actions

Researchers should use frameworks such as PRECIS-2 prospectively and report trial flexibility transparently. Funders should prioritize designs that incorporate implementation outcomes such as fidelity, adoption, and sustainability. Policy-makers should support infrastructure for integrating data Real-World Data (RWD), AI-based trial logistics, and pragmatic evidence dashboards into health systems. These coordinated efforts can bridge the gap between evidence and practice and improve patient outcomes across diverse rehabilitation contexts.