What the Nurse Saw First

The nurse saw it first.

Not in the way that phrase is usually meant. Not as a credit, not as a courtesy. In the literal, temporal sense — the early signs of clinical deterioration were visible at the bedside hours before the deterioration arrived. That is not anecdote. That is what three decades of literature on in-hospital cardiac arrest has consistently shown.

In 1990, a study in Chest examined the records of patients who arrested on hospital wards. Most had documented clinical deterioration in the eight hours before the arrest. The pattern has held in every replication since. Most hospital-floor cardiac arrests are not sudden events. They are slow deteriorations that ended at a recognizable threshold.

The nurses saw the deterioration. The vital signs that crossed the eventual threshold had been documented by nursing staff at regular intervals all night. The system did not respond because the system was not designed to respond to information held by nurses.

This essay is about what that costs.

What Continuous Observation Catches

The clinical assessment that physicians perform is structured around intermittent encounters. The morning round. The afternoon check. The new consult. The patient is observed for a span of minutes, evaluated against a clinical picture, and signed off until the next encounter. The next encounter is hours away.

The clinical assessment that nurses perform is structured around continuous presence. Vital signs at intervals. Hourly checks. Continuous monitoring of mental status, mobility, breathing, pain. The patient is observed across hours, evaluated against a baseline that exists in the nurse's own running model of the patient, and reassessed every time the nurse enters the room — which is many times more often than the physician does.

These are not two different ways to do the same task. They are two different sampling rates of the same clinical signal. The physician's sampling rate cannot detect changes that occur between samples. The nurse's sampling rate can.

The Modified Early Warning Score, developed in the UK in the late 1990s, formalizes what continuous observation catches. It assigns numeric values to respiratory rate, heart rate, blood pressure, temperature, and level of consciousness, and triggers escalation when the aggregate crosses a threshold. The National Early Warning Score, published by the Royal College of Physicians in 2012 and updated as NEWS2 in 2017, is the version now standard across the NHS. The scores work because the parameters they monitor are the parameters nurses already document.

A meta-analysis of early warning score studies shows that aggregate scoring identifies deteriorating patients earlier than physician judgment alone, with high sensitivity across most validated systems. The clinical signal is in the data nurses already collect. The systems that perform well are the systems that take that data seriously.

Failure to Rescue

The clinical research literature has a name for what happens when a deteriorating patient is not recognized in time. Jeffrey Silber and colleagues at the University of Pennsylvania introduced the term failure to rescue in 1992. It refers to in-hospital death following a complication — the system's failure to recognize and intervene when something goes wrong.

Failure to rescue is now a standard quality metric in American hospitals. It is reported to the Centers for Medicare and Medicaid Services. It is used in nursing-staffing research, in surgical-outcomes research, and in hospital performance comparisons.

Linda Aiken's work — the same research program cited in Issue 002 — uses failure to rescue as a primary outcome. In a series of papers, Aiken and her colleagues have shown that hospitals with more BSN-prepared nurses and better nurse-to-patient ratios have lower failure-to-rescue rates, with effect sizes large enough to imply that staffing decisions made at the administrative level affect mortality at the patient level. The nurses are not making the patients sick. The nurses are catching the deterioration that determines whether the system rescues the patient or doesn't.

The implication is structural. The hospital's capacity to intervene in deterioration depends on the hospital's capacity to detect deterioration. The capacity to detect deterioration is held by nursing staff, organized by nursing observation protocols, and bottlenecked by the system's willingness to act on what the nurses see.

The Data Physicians Don't See

A physician on rounds in the morning has access to overnight nursing notes, flowsheet vitals, and the patient at the moment of evaluation. What the physician does not have is the nurse's running model of the patient — the synthesis built across twelve hours of presence that produces a clinical impression no chart entry fully captures.

The mismatch is not informational. The information exists. The nurse has documented it. The mismatch is in what is operationally noticed.

A nurse who has been with a patient since 7 PM and saw them shift restlessly at 11, refuse the late snack at 1, and stop answering the call light at 4 has a different prior probability for sepsis than the physician arriving at 7 AM with a flowsheet showing only that the patient slept poorly and ate less. The flowsheet has the data. The flowsheet does not have the model. The model is in the nurse.

Hospitals that perform well on deterioration detection have built systems that surface the nurse's model, not just the nurse's data. Structured handoff protocols. Nurse-initiated rapid response teams. Bedside huddles. The MERIT trial of medical emergency teams, published in the Lancet in 2005, was famously mixed in its results — but subsequent meta-analyses have shown consistent benefit when activation criteria are nurse-driven and when nurses are empowered to call without physician permission.

The intervention that works is not adding more physician monitoring. It is acting on what the nurses already see.

What a Hospital Built Around This Would Look Like

A hospital that took continuous observation seriously would not need new technology. It would need new authority lines.

It would empower nursing staff to escalate without intermediate physician approval. It would treat the MEWS or NEWS score as a clinical order, not a suggestion. It would build nurse-led rapid response into the standard care pathway rather than as an after-the-fact reactive system. It would compensate nursing leadership in proportion to the clinical outcomes their staff produces, rather than in proportion to administrative span of control.

It would also, in residency education, teach physicians to read the nursing data as primary signal. Not "did the nurse call you" but "what has the nurse been documenting." The physician trained to scan the eight-hour nursing flowsheet before evaluating a deteriorating patient catches things the physician who relies on the moment-of-encounter exam will miss.

The nurse saw it first. The literature is unambiguous on this. The question is whether the system is built to listen.

About the author

Peter Schindler, MD, PhD is an Assistant Professor of Medicine and Associate Program Director of the Community Health Center Family Medicine Residency Program at the University of Nebraska Medical Center. He practices at Winnebago Comprehensive Health System, OneWorld Community Health Center, and Nebraska Medicine. He completed a Primary Care Research Fellowship at McGill University and holds a BSN from the University of Wisconsin-Oshkosh, an MS and PhD in nursing from Emory University's Laney Graduate School, a Diploma in Tropical Medicine from the Liverpool School of Tropical Medicine, and an MD from the Medical College of Wisconsin-Green Bay. He is board certified by the American Board of Family Medicine. The Interprofessional publishes new essays every week at the intersection of medicine, nursing, and the clinical knowledge that lives between them.

Disclaimer. The views in The Interprofessional are Peter Schindler's own and do not represent the official positions of the University of Nebraska Medical Center, Winnebago Comprehensive Health System, OneWorld Community Health Center, Nebraska Medicine, or any other affiliated institution.