NHS clinicians describe a long COVID pathway built around uncertainty and fragmentation
In daily practice, long COVID care in the NHS often begins without a clear biomarker, a definitive test, or a single clinic that owns the problem end to end. That absence shapes everything that follows. In this qualitative study from a diverse South London borough, 12 NHS healthcare professionals described a pathway that depends on pragmatic judgment, symptom-led triage, and repeated coordination across services rather than on a stable, unified model of care.
The study used single-timepoint, semi-structured interviews collected between July and October 2024 and analysed them with reflexive thematic analysis. Participants included five general practitioners, two respiratory consultants, three specialist nurses, one respiratory physiotherapist, and one clinical psychologist. Their accounts converged on a central operational problem: national guidance may outline best practice, but the local machinery needed to deliver that care remains uneven, incomplete, and heavily dependent on individual clinicians filling the gaps.
For patients, the result is a pathway that can feel clinically active and administratively inert at the same time.
Diagnostic uncertainty drives pragmatic, exclusionary assessment
The first theme was diagnostic uncertainty. Without a definitive biomarker, clinicians relied on exclusionary testing and subjective assessments of symptom severity to decide whether a patient fit the long COVID pathway and what kind of support was reasonable. In practice, that means the consultation often starts with ruling out alternative diagnoses, documenting the burden of fatigue, breathlessness, cognitive impairment, or other persistent symptoms, and deciding whether the presentation is severe enough to justify specialist input.
This approach is understandable, but it is also fragile. Long COVID remains a syndrome defined by clinical pattern recognition rather than a single laboratory signature, so frontline professionals are forced to weigh incomplete histories, fluctuating symptoms, and limited objective measures in a system that still expects clean referral thresholds. The absence of a biomarker is not a minor technical issue; it affects triage, referral speed, patient reassurance, and service design.
Clinicians were not describing diagnostic nihilism. They were describing a working model of care built around uncertainty management.
General practitioners became the system’s administrative hinge
The second theme placed general practitioners at the center of the pathway. GPs functioned as administrative linchpins, responsible for triaging patients into fragmented, symptom-specific specialist silos. In practical terms, this meant that one patient with persistent dyspnoea, fatigue, and cognitive complaints might be routed through several partially connected services, each focused on a narrow organ system or symptom cluster.
That arrangement increases the burden on primary care. Instead of a single integrated clinic absorbing complexity, GPs often had to keep the pathway moving, interpret service criteria, and decide where each patient should go next. The work is not simply clinical; it is coordinative, repetitive, and time-consuming. A fragmented service structure also raises the risk that patients with multisystem symptoms become stranded between clinics, especially when no one service accepts responsibility for the full syndrome.
A referral is not the same thing as continuity.
The interviews make clear how much hidden labour sits inside a seemingly ordinary referral pathway. When one service defines the problem as respiratory, another as psychological, and another as rehabilitation-related, the patient ends up carrying the burden of translating symptoms into the language each clinic will accept.
Rehabilitation replaced cure, but the care model remained uneven
The third theme centered on treatment pragmatism. Rather than relying on medicalised cures, clinicians described care that leaned heavily on non-pharmacological rehabilitation. That included support intended to improve function, pacing, and recovery rather than a single disease-modifying intervention. The clinical logic is consistent with the current understanding of long COVID as a condition with heterogeneous mechanisms and no universally effective pharmacologic fix.
This is where the pathway becomes especially vulnerable to local variation. Rehabilitation requires staffing, interdisciplinary coordination, and a shared definition of what functional improvement looks like. When those ingredients are missing, care can default to a collection of loosely connected interventions rather than a coherent plan. The study suggests that clinicians were doing what they could with the tools available, but the structure around them was not always built to support sustained recovery work.
The emphasis on rehabilitation also exposes an important tension in long COVID care. Patients may arrive hoping for a medical explanation that ends the search, yet the pathway often leads toward symptom management, graded functional support, and persistent uncertainty. For some patients that is appropriate; for others it can feel like an incomplete answer delivered before the system has fully ruled out other causes or established a shared long-term plan.
Post-discharge follow-up exposed a continuity gap
The fourth theme was the most operationally consequential. Monitoring practices varied across secondary care services, and once patients were discharged, continuity often broke down. The burden of re-initiating care shifted back to the patient, who had to recognize deterioration, navigate access points, and restart the process if symptoms persisted or worsened. This post-discharge discontinuity was not a side issue. It was a structural gap in the pathway.
Secondary care follow-up that varies by service creates uneven expectations for both clinicians and patients. Some services may monitor more closely; others may discharge earlier or with less explicit follow-up planning. In a condition marked by relapsing symptoms and functional fluctuation, that inconsistency can leave patients without a clear route back into care when they need it most. The result is a system that may appear complete on paper while leaving practical surveillance to the patient at home.
An eight-step pathway map points to standardization failures
The investigators synthesized these findings into an eight-step operational care pathway map. The map captured how patients move from presentation to assessment, referral, treatment, and discharge, while also exposing the friction points where the pathway loses coherence. Those friction points include the absence of standard functional assessment tools, the patchiness of interdisciplinary integration, and the lack of robust post-discharge tracking.
A standardized pathway would not erase the biological complexity of long COVID, but it could reduce avoidable variation in how patients are assessed and followed. Functional assessment tools would give clinicians a more consistent way to gauge severity and response. Integrated interdisciplinary clinics could reduce the need for patients and GPs to stitch together separate specialist opinions. Post-discharge tracking would help prevent patients from becoming stranded after an initial consultation or rehabilitation episode.
The study’s value lies in its specificity. It does not claim to solve long COVID. It shows how a national framework becomes harder to deliver when it reaches the clinic, especially in a service environment where no single specialty clearly owns the syndrome. The interview sample was small and drawn from one South London borough, so the findings cannot be assumed to represent every NHS setting. Even so, the pathway problems described here are concrete, recurring, and difficult to ignore: diagnostic uncertainty, fragmented triage, rehabilitation without a stable endpoint, and follow-up that often ends before the patient’s symptoms do.