Martha’s Rule: listening as a safety system, not a safety net
What the early evidence tells leaders about patient voice, staff escalation and safer system design
Martha’s Rule asks a harder question than whether a helpline exists
The situation can be a familiar one, where a patient is deteriorating. We often hear through our work with healthcare organizations, that missed opportunities to identify and address the deterioration of a patient isa pervasive, perennial problem. What about when the family can see the deterioration too. There’s that ‘gut’ feeling that something is not right, but as a team, we haven’t landed on what’s causing it yet. It can be for a few reasons: the signals are weak but growing, or we have fallen into the trap of confirmation bias, our pre-held belief is now causing us to interpret the information as confirming our current belief, rather than seeing the alternative picture. This can then result in when concerns are raised, they are dismissed, and the plan does not change.
Martha’s Rule is designed to creates an independent route through that impasse. As of publication, it is available across adult and children’s acute inpatient settings in England, is being expanded through maternity and neonatal services, and a phased national rollout to accident and emergency departments is under way. [1, 2, 3]
The early national data show that the route is being used and that calls are changing care. The independent formative evaluation also reports that patients, families and staff value the rule because it amplifies voice, supports collaborative care and strengthens escalation between ward teams and critical care outreach. [4, 5]
That is encouraging. It is not, however, the end of the governance question.
If a patient or relative has to bypass the normal system to get a deteriorating patient reviewed, what does that tell us about the reliability of the normal system? If staff use the route because their concerns are not heard, what does that reveal about hierarchy, authority gradients and psychological safety? And if the same specialty, ward, pathway, time of day or patient group repeatedly appears in activation data, why would an organisation treat each call as an isolated success?
The leadership test Do not count only how many Martha’s Rule calls were answered. Ask what conditions made the call necessary, what changed because of it, and whether those conditions are recurring elsewhere. |
Martha Mills: the history behind the rule
Martha Mills was 13 when she died in 2021. She had been admitted to King’s College Hospital after sustaining a pancreatic injury in a cycling accident. She later developed sepsis. Her family raised concerns as her condition deteriorated, but those concerns did not lead to the response Martha needed. [1, 6]
The coroner’s Prevention of Future Deaths report recorded that Martha was not referred promptly to paediatric intensive care. It concluded that, had she been referred promptly and treated appropriately, she was likely to have survived. The report also identified concerns about the paediatric early warning system and the relationship between the paediatric hepatology and intensive care services. [6]
Martha’s death therefore cannot be reduced to a single missed number or a family asking the wrong question. It involved clinical deterioration, escalation, communication, professional boundaries, system interfaces and the failure to respond to people who knew that something was wrong.
Martha’s parents, Merope Mills and Paul Laity, campaigned for a formal route through which patients and families could obtain an urgent review when concerns about deterioration were not being acted upon. NHS England subsequently developed Martha’s Rule in response to Martha’s death and other cases involving failures to recognise and respond to deterioration. [1, 7]
The name keeps the policy connected to the person. Patient safety language can become technical very quickly. “Escalation pathways”, “rapid response systems” and “deterioration governance” are useful terms, but Martha’s story reminds us why the route exists. It also puts the patient back at the centre of what we are doing, and why we are doing it.
What Martha’s Rule requires
Martha’s Rule has three connected components. It is not simply a telephone number for families. [1, 7]
Patients are asked at least daily how they feel and whether they are getting better or worse, with the response recorded and acted upon in a structured way.
Staff can request a review from a different team at any time when they are concerned that a patient is deteriorating and their concern is not receiving an adequate response.
Patients, families and carers have access to the same escalation route and can request a rapid review when deterioration is feared and concerns remain unresolved.
The 2026 core standards translate these components into six implementation expectations, including reliable and equitable access, effective patient wellness questions, independent rapid review, a capable workforce, learning and improvement, and governance and assurance. NHS trusts and foundation trusts are required through the 2026/27 NHS Standard Contract to implement the three core components by 31 March 2027. [8]
The strength of the model is that it adds patient, family and staff judgement to physiological monitoring. A score can identify change, but it does not contain every piece of relevant information. The people closest to the patient may notice altered behaviour, appearance, pain, interaction or function before these changes are fully represented in routine observations.
Where Martha’s Rule applies at the time of publication
The position is developing, so organisations should distinguish between services where Martha’s Rule is established, services being expanded, and future possibilities.
Adult and paediatric acute inpatient services
Live now. Martha’s Rule is available across adult and children’s acute inpatient services in England. NHS trusts and foundation trusts must have all three core components fully implemented under the NHS Standard Contract by 31 March 2027. This means leaders need to assure not only that an escalation telephone line exists, but also that patients are asked regularly how they feel, and that staff, patients, families and carers can obtain an effective independent review when concerns are not resolved. [1, 8]
Maternity and neonatal services
Expansion is under way. A national expansion was announced in June 2026 following pilots in 15 maternity and neonatal settings, and these services are included in phase 3 data collection. Implementation should reflect the needs of both mother and baby, continuity of care, communication requirements and known inequities, rather than simply copying the acute inpatient model. [2, 9]
Accident and emergency departments
Phased rollout is under way, but it is not yet fully live everywhere. Following a pilot involving seven trusts, national rollout was announced in September 2026 and is due to be completed by March 2028. Local models will need to account for crowding, waiting-room risk, frequent handovers, short clinical encounters and rapidly changing acuity. [3]
Other settings
Not nationally live as a universal model at the time of publication. No equivalent universal national implementation was identified for community, primary care, mental health, ambulance or social care services. Organisations may still apply the underlying principles and develop local escalation arrangements where there is a clear need, but they should describe these accurately and should not present a local approach as the national Martha’s Rule model.
What the evidence says so far
The evidence is promising, but it needs to be described accurately. National management information shows activity, changes in care and transfers to enhanced care. It does not, by itself, establish that Martha’s Rule caused a reduction in mortality across the NHS. The first comprehensive national summative evaluation began in 2026 and will examine outcomes, inequalities, service impact and implementation over three years. [4, 10]
Calls are leading to changes in care
NHS England reported 5,583 escalation calls from September 2024 to July 2025 across phase 1 sites. Seventy-two per cent came through the family or carer route. Of 2,277 calls concerning acute deterioration, 265 led to transfer to intensive or high-dependency care, enhanced care, a tertiary centre, or a specialist service or ward. [5]
By September 2025, 7,257 calls had been recorded across phase 1 and phase 2 sites, again with 72 per cent arising through the family or carer escalation process. [11] These figures demonstrate reach and clinical consequence. They also show why the patient and family route cannot be dismissed as reassurance or customer service.
The rule can amplify voice and improve collaboration
The 2026 NIHR-funded formative evaluation studied three early-adopter hospital sites using observations, interviews, documentary analysis, a systematic review and a public survey. Its interim findings report that patients, families and staff value Martha’s Rule for amplifying their voices, promoting collaborative care and improving escalation between ward and critical care outreach teams. [4]
This is an important positive effect. It places patient and family knowledge inside the safety system rather than outside it. It also gives staff a route when normal lines of escalation are blocked by hierarchy, assumptions or competing priorities.
There are important implementation gaps
The same evaluation identifies variation in how the patient wellness question is asked and recorded, limited awareness among some staff groups, insufficient explanation about what happens after escalation, and barriers affecting those who may be least able to use the route. Public awareness was around one in three in the study survey. [4]
A 2026 peer-reviewed service evaluation in a North Wales hospital found that 93 per cent of 168 surveyed patients and relatives were unaware of its established patient-activated escalation service. It also found confusion about terminology and concern that using the service could feel like “whistleblowing”. [12] The finding comes from one hospital and should not be generalised as a national awareness rate, but it illustrates why posters and telephone numbers are not sufficient controls.
The academic debate is usefully cautious
Peer-reviewed commentary has warned against framing Martha’s Rule only as an individual right that a patient must exercise. Graham and colleagues argue that a system-obligation approach better reflects the purpose of a clinical service intended to protect safety. [13]
A 2026 policy critique similarly supports the principles of listening, flatter hierarchy and improved deterioration response, while cautioning that implementation preceded rigorous prospective evaluation. It highlights unanswered questions about causality, workload, contextual fit and the risk of shifting responsibility towards critical care without addressing ward capacity, capability and governance. [14]
That caution does not weaken the case for Martha’s Rule. It strengthens the case for implementing it as part of a wider safety system rather than presenting it as a complete solution.
Listening is an active safety process
Healthcare organisations often say that they listen, but hearing is only part of it. Its what we do with the information that matters. Listening actively means more than inviting feedback. It means noticing discomfort, making it safe to question, checking understanding, revisiting a working diagnosis, responding proportionately and explaining what happens next. It also means treating disagreement as potentially valuable information rather than as resistance.
This applies to patients, families, carers and staff. Each sees a different part of the system. Patients experience symptoms. Families notice deviations from the person they know. Staff see workflow, deterioration, workload, handover and hidden adaptation. None has the whole picture, but safety is weakened when any of these perspectives is treated as inherently less legitimate.
This reflects the wider argument made across Promethean HD’s work on human factors and PSIRF: people are not external observers of the system. They are part of it. Their knowledge, limitations, relationships and conditions of work shape what the system can detect and how it responds.
Related Promethean HD reading: Human Factors in Healthcare | What is Human Factors? | PSIRF, Human Factors and Patient Safety Governance
The patient and family must not become the safeguard
Martha’s Rule is a valuable additional defence. It must not become the defence on which the system depends.
A patient may be too unwell, sedated, frightened, cognitively impaired or unfamiliar with the service to challenge care. A family may not be present. A carer may face language, communication or accessibility barriers. Some people will worry that speaking up could damage relationships with staff. Others may assume that the absence of an invitation means they should remain quiet.
If safety depends on patients or families recognising deterioration, understanding the escalation route, feeling confident enough to use it and persisting after an unsatisfactory response, the organisation has transferred part of its control system to those with the least formal authority and, often, the greatest emotional burden.
|
The primary controls remain competent clinical assessment, reliable observation, effective handover, multidisciplinary communication, manageable workload, clear escalation, appropriate senior review and a culture in which staff can challenge. Martha’s Rule adds resilience when those controls do not work as intended. It does not replace them.
Every activation is safety intelligence
An activation has two lives. The first is immediate and clinical: review the patient, understand the concern and act. The second is organisational: learn why the ordinary pathway did not provide sufficient confidence or response.
This creates a valuable source of weak-signal intelligence. A single call may relate to uncertainty or a communication gap. Repeated patterns may reveal a system that is drifting into danger.
What organisations should monitor
Boards and operational teams need more than a monthly total. Useful analysis should examine:
where calls originate, including site, service, ward, specialty and pathway;
when they occur, including time of day, day of week, shift transition and seasonal pressure;
who activates the route, including patient, family, carer or staff member;
what prompted the concern, including acute deterioration, pain, medication, delayed investigation, communication, discharge or unmet basic care needs;
what happened before activation, including previous concerns, response times, changes in observations and escalation attempts;
what the independent review changed, including treatment, diagnostics, monitoring, senior review, transfer or communication;
whether the call exposed a failure in a known control, such as handover, staffing, escalation or critical care outreach;
whether particular groups encounter barriers or appear under-represented relative to activity and risk;
whether repeat patterns involve the same service, professional interface, diagnosis, transition or operational condition;
whether improvements made after earlier calls reduced recurrence.
This analysis should connect with complaints, incidents, clinical audit, deterioration reviews, mortality review, staff speaking-up information, patient experience, staffing, flow and operational pressure. Martha’s Rule data should not sit in a separate patient-safety silo.
Questions for boards and leaders
A board that receives Martha’s Rule information should be able to answer practical questions, not simply note activity.
Are all three components implemented reliably and equitably?
Do patients, families, carers and staff know how to use the route?
How quickly are calls answered, and is the reviewer sufficiently independent and skilled?
What proportion of calls lead to a change in care, and what type of change?
What do calls tell us about the reliability of ward-level recognition, communication and escalation?
Where are repeat patterns occurring, and who owns the improvement response?
Which groups may be least able to access the route, and how do we know?
How are staff protected and supported when they use the escalation route?
How does learning from activations alter training, staffing, pathway design, handover or assurance?
What evidence shows that the underlying system is becoming safer, rather than simply more capable of recovering after failure?
Where the approach could be beneficial next
The principle behind Martha’s Rule is wider than acute deterioration: when credible concern is not receiving an adequate response, people need a safe and independent route to review. Expansion should still be evidence-led and adapted to context rather than copied mechanically.
Emergency care and waiting environments
The announced A&E expansion is a logical but demanding next step. Waiting rooms create different hazards from inpatient wards: crowding, incomplete information, frequent handovers, uncertain ownership and patients whose acuity can change while they wait. The escalation model must fit that work rather than assume an inpatient pathway. [3]
Maternity and neonatal care
Maternity and neonatal services are another clear application because the condition of a woman or baby can change rapidly, families may notice subtle change, and national reviews have repeatedly raised concerns about listening, hierarchy and escalation. The 2026 national expansion follows pilots in 15 maternity and neonatal settings. [2]
Implementation must reflect the distinctive context, including two linked patients, changing risk across pregnancy and birth, continuity, trauma, inequity and communication across obstetric, midwifery, neonatal and critical care teams.
Mental health and learning disability services
The underlying principle may also be valuable where behavioural or physical change is dismissed as part of a diagnosis, where diagnostic overshadowing is possible, or where families and carers hold essential knowledge about a person’s usual presentation. Any model would need clear clinical ownership, safeguards for autonomy and a response team appropriate to the setting.
Community, primary and virtual care
Potential value also exists across fragmented pathways, remote monitoring and hospital-at-home services, where deterioration may be distributed across organisations and no single team sees the whole picture. The design challenge is different: there may be no critical care outreach team and the risk may arise from coordination, access or delay rather than ward response. The principle of independent review may transfer, but the operating model must be designed locally and evaluated.
Digital systems, discharge and transitions of care
Patients and staff may identify that a discharge is unsafe, information has not followed the patient, medication is wrong, or a digital process is preventing review. These concerns are not always acute physiological deterioration, yet they can precede harm. Organisations should consider whether an independent escalation principle is needed for high-risk transitions, while avoiding the temptation to label every concern “Martha’s Rule”.
Martha’s Rule should help leaders see the system
The most visible success is the patient whose care changed after a call. The deeper success would be an organisation that learns enough from calls that the same bypass is needed less often for the same reasons.
That requires governance which is curious about the work behind the metric. A rising call rate might represent better awareness and access. It might represent worsening reliability in ordinary escalation. A low call rate might indicate safe care. It might indicate poor awareness, inaccessible information or fear of speaking up. The number has no stable meaning without context.
This is where human factors, patient safety and governance meet. Human factors asks how information, workload, hierarchy, technology, environment and team interaction shaped the response. Patient safety asks what risk was created and what controls failed. Governance asks whether leaders receive enough reliable intelligence to make informed decisions and can demonstrate that action follows insight.
Martha’s Rule gives organisations another way to hear. The leadership responsibility is to make sure they are listening before, during and after every call.
Conclusion: an exceptional route, not ordinary work
Martha’s Rule exists because Martha and her family were not heard when it mattered. The early evidence shows that independent escalation can alter treatment, prompt higher levels of care and improve collaboration. That is a significant patient-safety intervention. [4, 5]
It should remain exceptional. Patients, families and carers are essential partners in care, but they cannot become compensating controls for poor observation, weak communication, rigid hierarchy or unsafe clinical decision-making. Nor should staff have to rely routinely on a bypass because ordinary professional challenge is ineffective.
The practical test is not whether an organisation has launched Martha’s Rule. It is whether people can access it equitably, whether concerns lead to timely independent review, whether activations are analysed as system intelligence, and whether leaders act on the patterns they reveal.
A helpline can create a route around a problem. A listening organisation uses that route to understand and redesign the problem itself.
References
NHS England. (2026). Martha’s Rule.
Department of Health and Social Care, & NHS England. (2026, June 24). Martha’s Rule extended to all maternity services. GOV.UK.
NHS England. (2026, September 22). NHS to expand “life-saving” Martha’s Rule to all A&Es.
Thana, L., Prus, S., Fylan, B., Zondo, S., & Lawton, R. (2026). The formative evaluation of the implementation of Martha’s Rule: Interim report. NIHR Policy Research Unit in Quality, Safety and Outcomes of Health and Social Care.
NHS England. (2025). Martha’s Rule Programme July 2025.
Hassell, M. E. (2022, February 28). Regulation 28: Report to prevent future deaths, Martha Poppy Mills (2022-0063). Courts and Tribunals Judiciary.
NHS England Digital. (2026). Martha’s Rule Programme data collection.
NHS England. (2026, March 2). Martha’s Rule: Core standards.
NHS England. (2026). Martha’s Rule Programme July 2026.
National Institute for Health and Care Research. (2026, June 24). NIHR funds in-depth evaluation of Martha’s Rule.
NHS England. (2025). Martha’s Rule Programme September 2025.
Jeffrey, A., Edwards, E., Joishy, M., & Subbe, C. (2026). Patient and relative understanding of Martha’s Rule: Identifying barriers to patient-activated escalation in a mature system. Frontiers in Health Services, 6, Article 1812956. https://doi.org/10.3389/frhs.2026.1812956
Graham, M., Hanson, I., Hart, J., Young, P., Lignou, S., Parker, M. J., & Sheehan, M. (2024). Getting rights right: Implementing “Martha’s Rule”. Journal of Medical Ethics. Advance online publication. https://doi.org/10.1136/jme-2023-109650
Sara, T., Walker, D., & Welch, J. (2026). Martha’s Rule: A policy critique. Advances in Perioperative Care, 1(1). https://doi.org/10.14324/111.444/apoc.3730
Comments