Nature Medicine July 28, 2026

Integrated care pathway in individuals with Long COVID: STIMULATE-ICP, a cluster-randomized, phase 3 trial

Amitava Banerjee et al.

Bottom Line

The STIMULATE-ICP phase 3 cluster-randomized trial found that while multidisciplinary specialist care improved fatigue in Long COVID patients, adding multi-organ MRI or digital rehabilitation yielded no significant additional benefit.

Key Findings

1. Across all arms, the primary outcome was achieved with mean Fatigue Assessment Scale (FAS) scores improving significantly by 4.5 points from a baseline of 35.8 to 31.3 (s.d. 9.31) at 12 weeks.

2. The addition of multi-organ MRI to usual care did not significantly improve 12-week fatigue scores compared with usual care alone (adjusted difference -0.18; 95% CI -0.72 to 1.09; P=0.69).

3. Digital rehabilitation failed to demonstrate a statistically significant reduction in fatigue at 12 weeks versus usual care (adjusted difference -0.53; 95% CI -1.42 to 0.36; P=0.25).

4. The combination of both multi-organ MRI and digital rehabilitation yielded no significant interaction benefit on 12-week FAS scores (adjusted difference -0.17; 95% CI -1.06 to 0.72; P=0.71).

5. Although 24% of patients (105 of 437) undergoing multi-organ MRI were found to have site-reported organ impairment, this diagnostic information did not translate into measurable symptomatic or quality-of-life improvements.

Study Design

Design

Cluster-randomized phase 3 trial

Open-label

Sample

1,152

Patients

Duration

12 wk

Median

Setting

Multicenter, UK

Population Adults (18 years and older) with Long COVID referred to National Health Service (NHS) specialist clinics

Intervention Multi-organ MRI (Coverscan), digital rehabilitation (Living with COVID Recovery), or a combination of both

Comparator Usual care (multidisciplinary integrated care pathway without additional MRI or digital rehabilitation)

Outcome Mean Fatigue Assessment Scale (FAS) at 12 weeks

Study Limitations

• The pragmatic, open-label design could introduce performance or reporting biases, particularly for subjective self-reported endpoints like fatigue [1.1.2].

• The trial was conducted exclusively within the UK National Health Service (NHS) specialist clinics, which may limit the external validity and generalizability of the findings to different healthcare systems.

• The baseline standard of usual care evolved during the pandemic and varied slightly across primary care network clusters, potentially diluting the comparative effect size of the tested interventions.

• The modest intervention effects suggest that app-based digital rehabilitation adherence and engagement factors may have influenced the efficacy observed in that specific arm.

Clinical Significance

The STIMULATE-ICP trial provides high-quality evidence supporting the utility of multidisciplinary specialist care for Long COVID, as demonstrated by the clinically meaningful reduction in fatigue across all study groups. Crucially, the trial challenges the routine use of multi-organ MRI for these patients; while MRI identified organ abnormalities in nearly a quarter of individuals, it did not alter management in a way that improved fatigue or quality of life. Consequently, health systems can focus resources on holistic, integrated supportive care and targeted symptom management rather than broad, unselected advanced imaging, which adds cost without clear symptomatic benefit.

Historical Context

In the wake of the COVID-19 pandemic, a substantial subset of patients developed persistent, debilitating symptoms termed Long COVID. To manage this complex condition, the UK National Health Service established dedicated multidisciplinary Long COVID clinics starting in 2021. Advanced diagnostic tools such as multi-organ MRI received emergency use authorization in the UK to identify potential occult organ damage. However, despite rapid clinical adoption, there was a lack of high-quality randomized trial data demonstrating whether such resource-intensive investigations or standardized digital rehabilitation programs improved patient-centered outcomes. The STIMULATE-ICP trial was launched to rigorously evaluate these integrated care pathways.

Guided Discussion

High-yield insights from every perspective

Med Student

Medical Student

What pathophysiological mechanisms of Long COVID make multidisciplinary care more effective than single-system interventions, and why might a multi-organ MRI not change the immediate clinical outcome for fatigue?

Key Response

Long COVID is a multisystem condition involving endothelial dysfunction, immune dysregulation, and autonomic nervous system imbalances. Multidisciplinary care addresses overlapping somatic and psychological symptoms. Multi-organ MRI, while detecting subclinical organ damage, does not currently translate to specific fatigue-alleviating therapies, highlighting the gap between anatomical findings and functional symptom management.

Resident

Given the STIMULATE-ICP findings, how should you approach the diagnostic workup and initial management of a patient presenting with severe Long COVID fatigue who requests advanced imaging to find the cause?

Key Response

The trial showed multi-organ MRI added no significant benefit to fatigue outcomes over standard multidisciplinary care. Residents should practice shared decision-making, validating the patient's symptoms while explaining that advanced imaging is unlikely to alter management or improve fatigue, and instead focus referrals on integrated rehabilitation and symptom-guided therapies.

Fellow

The study found no additional benefit from the digital rehabilitation app. How might the concepts of post-exertional malaise (PEM) and digital fatigue explain this lack of efficacy in this specific patient population?

Key Response

Many Long COVID patients experience PEM, where structured or unpaced exertion (often prescribed in standardized rehab apps) can exacerbate symptoms. Furthermore, cognitive fatigue and screen intolerance are common, meaning a digital intervention could paradoxically increase cognitive load, offsetting any potential rehabilitation benefits compared to personalized, in-person pacing strategies.

Attending

How does the failure of multi-organ MRI and digital rehab to improve outcomes in this trial reframe our resource allocation strategy for Long COVID clinics, and what does it teach us about the limitations of medicalizing functional recovery?

Key Response

This trial reinforces that more testing (MRI) and scalable tech (digital rehab) do not inherently improve patient-centered outcomes in complex syndromes. Attendings should use this to teach the value of investing clinical resources into human-centered multidisciplinary teams rather than expensive diagnostics or automated therapeutics that lack targeted, disease-modifying pathways.

Scholarly Review

Critical appraisal through the lens of expert reviewers and guideline development

PhD

The STIMULATE-ICP trial utilized a cluster-randomized design. What are the primary methodological challenges regarding intra-cluster correlation and contamination in evaluating complex, multi-component interventions like integrated care pathways across different healthcare trusts?

Key Response

In cluster randomized trials, patients within a clinic (cluster) share similarities, reducing statistical power (design effect). Furthermore, multi-component interventions face fidelity issues; if control clinics adopt multidisciplinary practices (contamination) or if the intervention clinics struggle to implement the MRI/digital rehab uniformly, the treatment effect size is diluted, complicating the causal attribution of specific components.

Journal Editor

Considering fatigue is a subjective primary endpoint, how does the unblinded nature of the MRI and digital rehabilitation arms in this cluster-randomized trial introduce performance and detection biases, and how should these be weighed when evaluating the null result?

Key Response

Lack of blinding in a trial with a subjective endpoint usually biases towards a positive result (placebo effect of receiving advanced care/imaging). The fact that this trial found a null result despite these inherent biases strengthens the conclusion that MRI and digital rehab are ineffective for fatigue. As an editor, this makes the null finding highly robust and clinically significant, preventing healthcare systems from adopting costly interventions.

Guideline Committee

Current NICE and WHO guidelines emphasize comprehensive assessment and rehabilitation for Long COVID. Based on the STIMULATE-ICP trial, how should guidelines formally recommend against the routine use of multi-organ MRI and digital rehabilitation apps for fatigue management?

Key Response

Guidelines should issue a strong recommendation against routine multi-organ MRI for Long COVID unless specific focal symptoms warrant it, as the evidence shows no functional benefit and high cost. For digital rehabilitation, guidelines should adopt a conditional recommendation, noting that standardized digital platforms may not adequately account for post-exertional malaise and should not replace individualized, multidisciplinary pacing protocols.

Clinical Landscape

Noteworthy Related Trials

2011

PACE Trial

n = 641 · Lancet

Tested

Cognitive behavioural therapy (CBT) and graded exercise therapy (GET)

Population

Patients with chronic fatigue syndrome (CFS)

Comparator

Adaptive pacing therapy or specialized medical care alone

Endpoint

Fatigue severity and physical function

Key result: Reported that CBT and GET moderately improved outcomes for CFS, though the methodology and safety of GET for post-exertional malaise remain heavily debated.

2023

REGAIN Trial

n = 585 · BMJ

Tested

8-week online structured rehabilitation program

Population

Adults discharged from the hospital after COVID-19

Comparator

Usual care (single session of advice)

Endpoint

Health-related quality of life (PROMIS-29) at 3 months

Key result: The structured rehabilitation program significantly improved health-related quality of life, primarily driven by reductions in fatigue and improvements in psychological well-being.

2023

COVID-OUT Trial

n = 1,126 · Lancet Infect Dis

Tested

Metformin titrated to 1500mg daily for 14 days

Population

Overweight or obese adults with acute COVID-19

Comparator

Placebo

Endpoint

Severe acute COVID-19 and 10-month incidence of Long COVID

Key result: Outpatient treatment with metformin within 3 days of symptom onset reduced the 10-month incidence of Long COVID by 41 percent.

Tailored to your role

Want this tailored to you?

Add your specialty or training stage to get role-specific takeaways and more questions.

Personalize this analysis