Martha’s Rule: a policy critique

Tragedy to policy: a narrative-driven change

High-profile adverse events have frequently shaped healthcare policy. Whilst such events can catalyse improvement, implementation science emphasises the importance of distinguishing between circumstances of individual cases, factors contributing to harm and the mechanisms targeted by national interventions [6].

In the case that inspired Martha’s Rule, the public and media attention surrounding the case has powerfully shaped the framing of the policy response [1, 2]. Independent reviews of the case, including the external trust-commissioned review and the coroner’s inquest, identified systemic failures extending beyond the absence of a patient-activated escalation pathway [1, 4, 2]. These included fragmented responsibility between teams, unresolved inter-specialty conflict, rigid hierarchies and communication failures. Mechanisms for escalation already existed locally, including the recognition that parental concern should prompt review [1, 2]. The failure lay not in the absence of a pathway, but in the inability of teams to act on available information within a hierarchical and constrained clinical culture.

From a policy perspective, this distinction matters. Creating new escalation routes may be insufficient to address failures arising from how existing pathways are used, respected or bypassed. Interventions are unlikely to succeed if they do not target the true causal mechanisms of harm.

The structure of Martha’s Rule

NHS England describes three core components [7]:

1. Patients will be asked, at least daily, how they are feeling, if they are getting better or worse, and this will be acted on in a structured way.

2. All staff will be able to ask for a review from a different team if they are concerned that a patient is deteriorating and is not being responded to.

3. This escalation route will be available to patients, their families and carers and advertised across the hospital.

In practice, public discourse has focused predominantly on the third component. This imbalance has implications for both implementation and evaluation.

Asking patients how they feel: signal–noise context

The inclusion of patient-reported wellness within routine assessment is intuitively attractive and already common in perioperative care. Emerging literature suggests that caregiver concern may correlate with deterioration, particularly in paediatric populations [8, 9]. However, correlation does not equate to independent predictive value. Associations between patient concern and subsequent deterioration may be confounded by underlying illness severity. Without evidence that such concerns add any incremental value beyond the existing assessment, its contribution to risk discrimination remains uncertain.

There is limited standardisation in how wellness questions are framed, recorded or interpreted in relation to clinical risk. The questioning, understanding and responses to such questions are influenced by factors including language, cultural norms, health literacy, anxiety and expectations of care. Some patients may minimise symptoms, whereas others may report distress, not correlating with physiological deterioration. Consequently, the same question may generate fundamentally different signals across patients and settings, reducing reliability as a quantifiable safety measure. Without careful framing, the risk is that wellness questions become either performative or over-interpreted, generating false positives and defensive escalation [9].

This variability is compounded by the differences in how Martha’s Rule has been implemented across organisations. Without standardisation, it becomes difficult to determine what is being measured, or to compare outcomes across organisations. Without a clearer definition of both intervention and intended mechanism of action, any evaluation risks conflating heterogeneous practices, ranging from structured assessment to ad hoc escalation, under a single policy label, obscuring which components drive observed effects.

The key question is therefore not whether unwell patients feel unwell, but whether patient-reported concern adds an actionable signal beyond the existing warning systems, early enough to change outcomes [9]. The national rollout of Martha’s Rule has preceded definitive evaluation of this mechanism.

Patient-activated escalation and critical care capacity

The most operationally consequential element of Martha’s Rule, and the most publicised, is the provision of a direct escalation route to rapid review teams, typically CCOT. Pilot data indicate low call volumes [10], although this may not reflect steady-state demand, particularly as public awareness increases. Nevertheless, the most recent data show that, of those patients who did acutely deteriorate, 81% had EWSs that would not otherwise have triggered escalation. Many of these episodes were associated with reported changes in management or transfers of care [10]. Importantly, however, the nature and clinical significance of these management changes remain poorly defined, as heterogeneous interventions are reported collectively, again limiting interpretation of their true impact. Critical care outreach teams are configured to assess and manage acute clinical instability and physiological deterioration. However, national data suggest that a substantial proportion of calls relate not to physiological deterioration but instead to poor communication, delayed investigations or discharge planning [10]. Whilst important, these issues may be more appropriately addressed through ward-based senior review and care coordination. The routine diversion of such concerns to outreach risks blurring roles and misaligning expertise, potentially reducing the capacity for assessment and management of acute clinical instability. Preserving these functions is essential if patient-activated escalation pathways are to function effectively and remain sustainable within perioperative systems.

Independent clinical review is often valuable, creating opportunity for reflection in evolving situations and supporting communication and decision-making, particularly when concerns are raised by patients or families. When clinical opinions differ, governance around the decision-making authority may help to ensure that escalation supports collaboration rather than defensiveness.

Retrospective justification and evidentiary standards

Some commentators have raised concerns that Martha’s Rule has been implemented nationally ahead of the accumulation of robust supporting evidence, with subsequent evaluation occurring alongside an already-established policy direction [1, 10]. This sequencing places scrutiny on how evidence is generated, interpreted and weighed in large-scale safety interventions. The approach is not inherently illegitimate: adaptive policies are sometimes necessary in complex systems, particularly when ethical imperatives favour action. However, it requires transparency regarding uncertainty and careful interpretation of process measures. Retrospective justification risks overstating benefit and effect size. Reported outcomes, such as call volumes, reviews triggered or management changes, may be conflated and not necessarily reflect patient benefit. It can attribute causality when only associations exist and do not adequately account for confounding factors, including staffing variation, case mix, seasonal pressure or parallel safety initiatives.

Whilst the ethical impulse to act following tragedy is understandable, differences between evidentiary standards in policy implementation and those typically applied in clinical practice may influence professional confidence and engagement with safety initiatives. Post hoc attribution of lifesaving interventions or implied mortality reduction without clear counterfactuals risks inflating perceived impact and may undermine confidence.

Structural context: resources, responsibility and drift toward critical care

Evaluation of Martha’s Rule must be situated within the wider NHS context. Many of the challenges it seeks to address – delayed recognition of deterioration, communication failures and reluctance to escalate – reflect not only culture but sustained workforce and resource constraints within an increasingly overextended system.

Much responsibility for clinical risk has gradually migrated towards critical care services, which function increasingly as the final common pathway for uncertainty and system strain. Whilst outreach models may improve outcomes [11], expansion of their remits risks normalising underinvestment upstream.

In perioperative medicine, this dynamic is especially visible. Surgical wards now manage patients with greater medical complexity and with variable senior presence. Although the escalation to critical care may be appropriate for individual patients, reliance on this response at scale may reduce the focus on strengthening the ward-level ability to recognise and manage deterioration. This inversion places additional expectation on critical care to identify, arbitrate and absorb risk generated elsewhere. A sustainable safety strategy must, therefore, prioritise resourcing and ward-level capability, with critical care providing support for deterioration, rather than as a compensatory mechanism for systemic deficit.

Implications for perioperative medicine

Perioperative services offer a unique opportunity to implement patient-centred escalation mechanisms more thoughtfully, within established multidisciplinary frameworks. The specialty was developed to address fragmentation between teams and stages of care, emphasising shared responsibility, robust communication and coordinated decision-making. Preoperative assessment allows an anticipatory discussion of risk and escalation preferences, whilst postoperative pathways already integrate structured monitoring and senior review.

Rather than adopting Martha’s Rule as a blunt universal mandate, perioperative medicine could develop context-specific models that:

• further integrate patient-reported concern into existing assessment frameworks;

• define clear thresholds and responsibilities for independent review;

• protect outreach capacity for physiological deterioration;

• evaluate outcomes prospectively, including workload, staff morale and unintended effects [12].

Research ethics statement

Not applicable to this article.

Consent for publication statement

This narrative article is based on a fictional clinical vignette and does not describe a real patient or clinical encounter so consent for publication is not necessary. Images are used solely for illustrative purposes and are not related to the vignette. Where real patient images are included, the authors confirm that written informed consent for publication has been obtained for this purpose in accordance with the journal’s patient consent requirements.

Conflicts of interest statement

David Walker is Editor-in-Chief of Advances in Perioperative Care. The authors declare no further conflicts of interest.