EPIC-Neck delivery fell short of the fidelity needed for a definitive neck pain trial
The EPIC-Neck exercise programme for chronic non-specific neck pain moved through a UK feasibility study with a clear signal: the intervention could be delivered, but not yet with the consistency needed for a larger efficacy trial. In this process evaluation, the programme’s multidimensional fidelity was tested across training, delivery, receipt and enactment inside NHS musculoskeletal outpatient services. The design matters because complex rehabilitation interventions can fail quietly when the intended content is diluted at the bedside, even when the programme looks sound on paper.
EPIC-Neck was built as an individualised, evidence- and theory-informed exercise prescription model. It combined exercise progression with behaviour change techniques aimed at four performance objectives: performing prescribed exercise, progressing or regressing exercise, using exercise to manage flare-ups, and engaging in general exercise. The fidelity work examined whether physiotherapists could deliver those components as intended and whether participants could understand and use them in real life.
Physiotherapist confidence rose, but preparedness did not fully catch up
Seven physiotherapists delivered EPIC-Neck, and five were interviewed about their experience. Their self-rated confidence improved after training, rising from a median of 3/5 to 4/5. On paper, that looks encouraging. In practice, the interview data told a more restrained story.
Physiotherapists still felt underprepared for parts of the intervention, especially the more complex elements that required behavioural coaching rather than straightforward exercise instruction. The gap between confidence and readiness is clinically important. A therapist may feel more comfortable after training, yet still struggle to integrate problem-solving, social support, and nuanced progression decisions into a real consultation with time pressure, mixed patient needs and variable therapeutic skill.
A single training session can lift confidence without guaranteeing competence in complex rehabilitation delivery.
Delivery fidelity missed the acceptable threshold across key components
Delivery was measured using a 17-item fidelity checklist applied to audio-recorded appointments. On a 0–5 scale, with 3 defined as acceptable, mean delivery fidelity reached only 2.4 with a standard deviation of 1.3. That places the programme below the threshold set for acceptable delivery and signals substantial variation between appointments and likely between clinicians.
The weakest elements were social support, problem-solving and the performance objectives tied to PO2 through PO4. Scores for those components were below 2/5, showing that the intervention’s more sophisticated behaviour change features were the first to erode in routine practice. Those omissions matter because progression, flare-up management and general exercise engagement depend on more than simply handing out a set of movements. They require explicit coaching, anticipatory planning and reinforcement of self-regulation skills.
The more complex the rehabilitation task, the more vulnerable it becomes to delivery drift.
What stands out is that the programme’s mechanical core was easier to preserve than its behavioural scaffolding. Physiotherapists could generally deliver the basic prescription, but the higher-order support needed for progression and long-term self-management was not consistently embedded. In chronic neck pain care, that gap may determine whether exercise becomes a durable self-management strategy or a short-lived clinic task.
Patients grasped the exercise work better than the self-management work
Eighteen participants contributed quantitative receipt data, and 12 were interviewed. Patients generally understood and enacted the prescribed neck exercises, which suggests the foundational exercise message was communicated well enough for uptake. The weaker areas appeared when the programme moved from immediate exercise performance to more demanding behaviour change tasks.
Receipt and enactment were less robust for complex or resisted exercises and for longer-term self-management behaviours linked to PO2 through PO4. In plain clinical terms, participants were more likely to understand what to do with the exercise itself than how to adjust it, use it during flare-ups or maintain a broader exercise habit over time. That pattern fits the logic of the intervention: the further the programme moves from a discrete exercise instruction toward autonomous self-management, the more dependent it becomes on skilled coaching and repeated reinforcement.
This is a familiar fault line in musculoskeletal rehabilitation. Exercise prescription is only one part of treatment; adherence, progression and confidence during setbacks usually decide whether the benefits last beyond the clinic.
The process evaluation explains why a full trial would have been premature
The value of a feasibility study lies in exposing the weak links before a definitive randomised trial consumes time and resources. Here, the fidelity data do exactly that. Training improved physiotherapist confidence, but not enough to ensure full delivery of the intervention as designed. Delivery fidelity remained below the acceptable threshold. Patients did better with the immediate exercise tasks than with the more complex self-care components, suggesting the intervention’s intended behaviour change pathway was incomplete.
Mixed-methods integration strengthened the interpretation. Quantitative scores showed where the intervention fell short, while interviews clarified why those gaps appeared. The narrative was consistent across data sources: EPIC-Neck was not yet stable enough in real-world delivery to support a clean test of clinical effectiveness. A full trial at this stage would risk evaluating a diluted intervention rather than the programme that was originally designed.
For chronic non-specific neck pain, that distinction is not academic. If clinicians cannot reliably deliver progression advice, flare-up management or general exercise support, then the programme is unlikely to produce its intended downstream benefits on self-management and symptom control.
Refinement needs to focus on the behaviour change components
The most actionable message from this feasibility work is where refinement should happen. The programme’s simple exercise instruction appears more deliverable than its behaviour change components, and the training model raised confidence without fully closing that gap. Future development will need to strengthen therapist preparation for problem-solving, social support and the practical delivery of PO2 through PO4. It will also need to test whether those components can be made more usable in routine NHS consultations without losing the theory that gives the programme its structure.
Before any definitive trial, both the intervention and its training package need revision. Without that work, the next study would be at risk of measuring incomplete implementation rather than true treatment effect.