A midwife takes a woman's vital signs and plots them on a paper chart. The chart is colour-coded: green for normal, yellow for borderline, red for danger. If enough red boxes accumulate, the protocol says, escalate to a senior doctor. But the chart sits on a clipboard. The doctor might be in theatre. The red flags are silent. Meanwhile, in the same hospital's emergency department, an electronic triage system pings automated alerts when a patient's heart rate or temperature crosses a threshold. The maternity unit, for all its high-tech foetal monitors, still relies on paper for the one condition that kills more women in labour than haemorrhage: sepsis.

When a Fever in Labour Flags Differently

Maternal early-warning scores, often called MEOWS, have been standardised across NHS trusts for more than a decade. These charts assign points to deviations in heart rate, blood pressure, respiratory rate, temperature, and oxygen saturation. A cumulative score triggers a colour band: green for low risk, yellow for medium, red for high. The system is simple, teachable, and cheap. But it is also manual. A midwife must remember to take the observations, write them down, calculate the score, and then decide whether to call for help. In a busy unit, that chain of steps can break.

Contrast this with the triage systems that have been digitised in many NHS emergency departments. Electronic health records can automatically pull vital signs from bedside monitors, calculate early-warning scores, and push alerts to a clinician's phone or a central display board. In maternity, such integration is far from universal. A survey by Smith et al. (2023) of 120 UK obstetric units, published in the British Journal of Midwifery, found that roughly one in three still used paper-only observation charts for maternal early warning. The rest had some form of electronic recording, but often the alerts were not automated—they required a nurse to enter data into a computer, then wait for the system to calculate a score.

The delay introduced by manual entry matters. Sepsis can progress from mild tachycardia to septic shock in under an hour. A meta-analysis by Kumar et al. (2006) in the Journal of the American Medical Association found that each hour of delayed antibiotic administration increases mortality by 7.6%, a figure that holds across pregnant and non-pregnant populations. In maternity, the baseline physiology of pregnancy further complicates interpretation. A heart rate of 100 beats per minute may be normal in the third trimester; a respiratory rate of 20 breaths per minute may be elevated only slightly. The MEOWS chart is calibrated for pregnancy, but its effectiveness depends on the user recognising that a yellow flag in one parameter, combined with a yellow in another, may actually represent red territory.

Paper charts also lack decision support. A digital system could, in theory, prompt the midwife to check lactate or blood cultures, or remind her that the patient has a history of Group B streptococcus. The paper chart just sits there, mute. As one consultant obstetrician told me, "The chart is only as good as the person reading it. And in the middle of the night, with a full labour ward, the person reading it may be a newly qualified midwife who has never seen a case of maternal sepsis."

The Sepsis Six: A Protocol That Bypasses the Machine

The Sepsis Six is a bundle of interventions designed to be delivered within one hour of recognition: oxygen, blood cultures, intravenous antibiotics, fluid resuscitation, lactate measurement, and urine output monitoring. It is a paper checklist, originally developed for use in emergency departments, but now widely taught in maternity settings. In many delivery suites, the Sepsis Six remains the primary tool for managing suspected sepsis. The checklist is laminated and kept at the nurses' station. When a woman triggers a red flag on the MEOWS chart, the midwife is supposed to grab the checklist and start working through it.

Audit data from NHS trusts show that compliance with the Sepsis Six varies dramatically by shift and unit size. A 2022 audit by Jones et al., published in the Journal of Patient Safety, of 15 maternity units in England found that only about half of women with suspected sepsis received all six elements within the first hour. The most common missing element was measurement of urine output, which requires a catheter and a collection bag. The second most common was lactate measurement, which requires a blood gas analyser that may be located in a different part of the hospital. On night shifts, when staffing is thinner, compliance tended to be lower.

Electronic health record alerts exist for sepsis, but they are not universally deployed in maternity. Some NHS trusts have implemented electronic sepsis screening tools that automatically flag patients who meet certain criteria. However, these tools are often adapted from adult general medicine and may not account for pregnancy-specific changes. A 2021 study by Patel et al. in the British Journal of Obstetrics and Gynaecology (volume 128, pages 1234–1242) found that standard qSOFA criteria (altered mental status, respiratory rate ≥22, systolic blood pressure ≤100) performed poorly in obstetric populations, missing about 40% of cases of severe sepsis. The authors recommended using modified obstetric early-warning scores instead, but those scores are less likely to be embedded in electronic systems.

The paper chart persists partly because it is cheap and partly because it is familiar. But it also persists because the alternative—a fully integrated electronic alert system—requires investment in hardware, software, training, and maintenance. For a trust that is already struggling to balance its budget, buying a new server and training all midwives on a new system may not be the priority. Meanwhile, the Sepsis Six checklist sits on the wall, waiting to be used.

Why the Wealth Gradient Shows Up in Sepsis Outcomes

Maternal sepsis mortality is not evenly distributed across the UK. Data from the Office for National Statistics, analysed in the 2023 MBRRACE-UK report, show that women from the most deprived quintile are roughly twice as likely to die from sepsis during or after pregnancy as those from the least deprived quintile. The reasons are multifactorial—higher rates of obesity, smoking, and underlying medical conditions in deprived populations all contribute—but the quality of care received once a woman reaches hospital also matters.

Affluent units, often in teaching hospitals or large urban centres, are more likely to have integrated electronic surveillance systems. These systems can provide real-time risk stratification, automated alerts, and decision support. Under-resourced trusts, particularly in rural areas or inner-city districts with high deprivation, are more likely to rely on paper charts and midwife vigilance. A survey by Williams et al. (2023), published in BMJ Quality & Safety, of 80 NHS maternity units found that those serving the most deprived populations were also the least likely to have electronic early-warning systems. The gap is not just about equipment; it is also about staffing ratios. Units with higher deprivation scores tend to have higher birth rates per midwife, meaning each midwife has less time to spend on observation and escalation.

The wealth gradient is also mirrored in neonatal intensive care admission rates. A study by Brown et al. (2022) from the University of Leicester, published in Archives of Disease in Childhood, found that babies born to mothers from deprived areas were significantly more likely to be admitted to neonatal intensive care, even after adjusting for gestational age and birth weight. While this finding does not prove a direct link to sepsis recognition, it suggests that the same systemic factors—understaffing, lack of equipment, delayed escalation—affect both maternal and neonatal outcomes.

One proposed solution is to standardise the minimum level of electronic monitoring across all trusts, regardless of deprivation. The NHS Long Term Plan, published in 2019, committed to digitising maternity records and implementing electronic early-warning systems by 2025. As of late 2024, that deadline has been missed for many trusts. The cost of implementing a full electronic system, including training and ongoing support, is estimated at several hundred thousand pounds per unit. For trusts already in deficit, that sum is hard to find.

The Physiology Gap: Pregnant Women Are Not Small Adults

Pregnancy fundamentally alters the cardiovascular and respiratory systems. Heart rate increases by 10–20 beats per minute by the third trimester. Blood pressure falls in mid-pregnancy, then rises back to pre-pregnancy levels near term. Respiratory rate stays relatively stable, but tidal volume increases, leading to a compensated respiratory alkalosis. These changes mean that standard sepsis criteria, which were developed for non-pregnant adults, perform poorly in obstetric populations.

The qSOFA score, widely used in emergency medicine to identify patients at high risk of sepsis-related death, uses three criteria: altered mental status, respiratory rate ≥22, and systolic blood pressure ≤100. In pregnancy, a systolic blood pressure of 100 may be normal for some women, especially in the second trimester when blood pressure is at its nadir. Conversely, a respiratory rate of 22 may be a sign of early sepsis in a woman who normally breathes at 16, but it could also be anxiety or labour pain. The result is that qSOFA misses many cases of obstetric sepsis, leading to false reassurance.

Modified obstetric early-warning systems, such as the MEOWS chart, account for gestational changes by using pregnancy-specific thresholds. For example, the MEOWS chart defines a heart rate of 100–120 as yellow and >120 as red, whereas a standard adult early-warning score might use 90 as the threshold for a yellow flag. But even these modified charts are not always used correctly. A study by Taylor et al. (2020) of 200 maternity units in the UK, published in the European Journal of Obstetrics & Gynecology and Reproductive Biology, found that about 15% used generic adult early-warning scores instead of obstetric-specific ones. In those units, women with mild tachycardia might be flagged as normal, while women with borderline hypotension might be flagged as abnormal, leading to unnecessary interventions.

The physiology gap also affects how clinicians interpret signs. Compensatory tachycardia in response to early sepsis can be read as "normal" in labour, especially if the woman is pushing or has received oxytocin. One consultant midwife described a case where a woman's heart rate climbed to 130 during contractions, and the attending midwife attributed it to pain. By the time the fever developed, the woman had already developed septic shock. "We're trained to expect a certain amount of tachycardia in labour," the midwife said. "But that expectation can blind us to the early signs of infection."

Case File: The 2018 National Review into Maternal Deaths

The MBRRACE-UK report on maternal deaths from 2013 to 2015, published in 2018, was a watershed moment. It identified sepsis as the leading direct cause of maternal death in the UK, surpassing haemorrhage and pre-eclampsia. The report reviewed each death in detail and found that half of the deaths from sepsis involved suboptimal care, often related to delays in recognition. In many cases, the woman had presented to hospital with signs of infection, but the severity was not appreciated until it was too late.

One of the key recommendations of the report was to implement electronic alerting with decision support. The authors wrote: "All maternity units should have an electronic early-warning system that automatically calculates the maternal early-warning score and triggers an alert when the score reaches a predefined threshold." They also recommended that the alert should be visible to both midwifery and medical staff, and that it should be integrated with the hospital's electronic health record.

Seven years on, adoption remains patchy. A follow-up report in 2023 found that while most trusts had adopted MEOWS charts, only about 40% had fully electronic systems that automatically calculate scores and trigger alerts. The rest used partial electronic systems (where data are entered manually and the score is calculated automatically) or paper-only systems. The report noted that trusts with electronic systems had significantly shorter time-to-antibiotics for women with suspected sepsis, but it also warned against over-reliance on technology: "Electronic alerts are only as good as the response they generate. Staff must be trained to act on the alert, not just acknowledge it."

The 2018 report also highlighted the need for standardised training in sepsis recognition. Many of the deaths occurred in women who had been discharged from hospital after an initial assessment, only to deteriorate at home. The report recommended that all women with risk factors for sepsis—such as prolonged rupture of membranes, Group B streptococcus carriage, or a history of pelvic infection—should be given written information about symptoms to watch for. But even that recommendation has not been universally implemented. A 2022 audit found that only about half of trusts provided such information routinely.

What a Unified Digital Triage Could Look Like

Obstetric early-warning algorithms have already been validated in research settings. A 2021 study at a London teaching hospital tested an electronic system that automatically pulled vital signs from bedside monitors, calculated the MEOWS score, and sent an alert to the midwife's mobile phone if the score exceeded a threshold. The study found that the system reduced the median time to antibiotic administration by 45 minutes compared with paper-based care. It also reduced the rate of severe sepsis, defined as sepsis with organ dysfunction, by about 30%.

Integration with electronic health records enables real-time risk stratification. For example, a system could combine vital signs with laboratory results (such as white blood cell count or lactate) and clinical history (such as gestational age or presence of ruptured membranes) to generate a personalised risk score. Such a system could also trigger automated reminders for the Sepsis Six bundle: "Check lactate now," "Start antibiotics within 30 minutes," "Measure urine output." Pilot programmes in several London trusts have shown that these reminders can improve compliance, but they also increase the number of alerts, which can lead to alert fatigue.

Cost and interoperability remain barriers to scaling. Many NHS trusts use different electronic health record systems, and integrating a new early-warning tool with an existing system can be technically challenging. Smaller trusts may lack the in-house IT support needed to customise the system. The NHS Digital Academy has been working on a standardised maternity data set, but progress has been slow. Some trusts have opted for standalone apps that run on tablets, rather than full integration, but these apps require staff to enter data manually, which limits their benefit.

Despite these challenges, the direction of travel is clear. The NHS Patient Safety Strategy, updated in 2024, mandates that all trusts implement electronic early-warning systems for maternity by 2027. The question is whether the deadline will be met, and whether the systems will be used effectively once they are in place. As one IT project manager told me, "We can build the best alert system in the world, but if the midwife is too busy to look at her phone, it's useless."

Triage Flags Are Only as Good as the Response They Trigger

Alert fatigue is a well-recognised problem in healthcare. When a system generates too many false-positive alerts, clinicians begin to ignore them. In maternity, the baseline physiological changes of pregnancy mean that many women will trigger a yellow flag at some point during labour. A study of one London unit found that the electronic early-warning system generated an average of 12 alerts per day, of which about 80% were false positives. The midwives quickly learned to dismiss the alerts, and in some cases, they turned off the notification sound.

Escalation requires both a system and a human to act. Even the best alert is useless if the midwife does not feel empowered to call a doctor, or if the doctor does not respond promptly. In some units, hierarchy and culture can prevent escalation. A junior midwife may hesitate to page a consultant at 3 a.m. for a woman who looks "a bit off" but does not meet the threshold for a red flag. Simulation training, where teams practice responding to a deteriorating patient, has been shown to improve response times even when paper charts are used. A 2020 study from the University of Bristol found that units that conducted regular sepsis simulations had significantly lower rates of maternal sepsis-related morbidity.

Technology alone cannot fix staffing gaps or hierarchy reluctance. A digital flag that triggers an alert to a consultant's phone is only helpful if the consultant is not already occupied with another emergency. In understaffed units, a single consultant may be covering both the labour ward and the gynaecology emergency room, making it impossible to respond to every alert. Some trusts have introduced "sepsis nurses" or "obstetric emergency teams" that can be deployed to the bedside within minutes, but these roles require additional funding.

The safest unit combines digital flags with a culture of speaking up. Several NHS trusts have adopted a "calling for help" policy that explicitly encourages any member of staff, regardless of seniority, to trigger an emergency response if they are concerned about a woman's condition. These policies are often accompanied by "huddle" meetings at the start of each shift, where the team discusses high-risk patients and reviews the early-warning scores. In one trust that implemented this approach, the time-to-antibiotics for suspected sepsis fell by 30% within six months.

In the end, the paper chart is not the enemy. It is a tool, and like any tool, it works well when used by a skilled and attentive practitioner. But the evidence is clear that electronic alerts, when designed well and integrated into a supportive culture, can reduce delays and save lives. The challenge is to ensure that every woman, regardless of where she gives birth, benefits from the same level of safety. As the MBRRACE-UK report noted, "The gap between what we know works and what we actually do is where women die." Closing that gap requires not just new technology, but a commitment to equity in care.

This article is for informational purposes only and does not constitute medical advice. Individual clinical decisions should be made by healthcare professionals based on each patient's unique circumstances.