Registered nurses

Three parts of your work where AI is already doing real lifting, and what stays yours.

• Nightingale sanitary reform + systemati…1854–1900

  Nightingale sanitary reform + systematic observation

  Florence Nightingale's great insight was that the nurse's primary tool was observation and environmental control (handwashing, clean linen, ventilation, lighting, and diet) rather than heroic intervention. Her rose diagrams (polar-area charts showing causes of mortality) were among the first statistical visualizations used to change hospital policy. The 1860 Nightingale school at St Thomas' codified this as systematic training: nurses learned to record vital signs in writing, track fever curves, and communicate findings to physicians in structured terms.

  Effect on the work

  Mortality at Scutari dropped following Nightingale's sanitary reforms (1854-55), a data point that launched nursing as a respected profession rather than domestic service.

• Hypodermic syringe + sphygmomanometer1895–1940

  Hypodermic syringe + sphygmomanometer (BP cuff)

  Two instruments became inseparable from the trained nurse by 1900-1920: the hypodermic syringe (widely commercialized from the 1880s) and the sphygmomanometer, whose routine use in clinical settings took hold in the early 20th century. Taking and recording blood pressure became a nursing task: the first piece of continuous physiologic monitoring nurses performed systematically. Subcutaneous injections of morphine, digitalis, and camphor moved from experimental to standard nursing procedure.

• IV therapy + blood transfusion + penici…1943–1960

  IV therapy + blood transfusion + penicillin administration

  WWII transformed bedside nursing from primarily palliative care to active pharmacological intervention. Penicillin's mass production reduced mortality from wound infections among soldiers; in civilian hospitals, infections that had been near-certain death sentences (septicemia, pneumonia, puerperal fever) became treatable with a nurse-administered course of injections. IV fluid therapy and blood transfusions moved from surgical suites into general wards. Nurses went from monitoring patients who were dying to administering treatments that cured them: a profound shift in clinical authority.

  Effect on the work

  The leading causes of in-hospital death began shifting from infectious disease to chronic illness (cardiac disease, cancer), permanently restructuring which conditions required long nursing observation vs. short acute intervention.

• Coronary Care Unit1960–1975

  Coronary Care Unit (CCU) + continuous ECG telemetry

  The coronary care unit, pioneered by Hughes Day at Bethany Medical Center (Kansas City, 1962) and Desmond Julian in Sydney (1961), was the first clinical environment in which nurses were trained to interpret continuous physiologic data, specifically ECG waveforms, and act on them without waiting for a physician. The four pillars of the CCU were: continuous ECG monitoring, defibrillation, clustering of MI patients in one unit, and, critically, a policy change permitting trained nurses to initiate resuscitation. This was the first time nurses had standing orders to make life-and-death decisions autonomously.

  Effect on the work

  Hospital mortality in CCU settings by the late 1960s, with nurse-initiated defibrillation and CPR as the primary mechanism. The CCU model spread across US hospitals through the 1960s, requiring a new tier of specialty-trained 'coronary care nurses' and seeding the modern ICU profession.

• CPR1963–1985

  CPR: AHA standardized training for nurses

  The AHA formally endorsed closed-chest cardiac resuscitation in 1963 (building on Kouwenhoven, Jude, and Knickerbocker's 1960 landmark JAMA paper) and began training physicians, then nurses. Standardized CPR certification transformed nursing: every RN in an acute-care setting was now expected to initiate cardiac resuscitation, operate a defibrillator, and manage a code. For the first time, an explicit expectation of clinical authority in emergencies was built into RN licensure and hospital hiring standards.

• Hospital information systems1980–2000

  Hospital information systems (HIS) + early clinical computing

  The first hospital information systems (HIS) arrived in large academic medical centers in the 1970s-1980s, handling order entry and lab results. By the late 1980s, nurses at hospitals with early HIS were printing medication administration records (MARs) rather than hand-writing them. The transition was partial and uneven: most community hospitals remained paper-based through the 1990s. The key effect was not automation but the first bifurcation of "patient-facing time" vs. "computer time" in the nursing day.

• Barcode Medication Administration1995–2010

  Barcode Medication Administration (BCMA)

  BCMA was invented by nurse Glenna Sue Kinnick and first deployed at the Colmery-O'Neil VA Medical Center in Topeka, Kansas in 1995. The VA promoted it to all facilities between 1999-2001. BCMA requires nurses to scan a patient wristband and each medication unit before administration, verifying the five rights (right patient, drug, dose, route, time) electronically. Early studies found a sizeable share of medication errors occurred at the administration phase; BCMA cut near-miss administration errors in VA deployments.

  Effect on the work

  Adds several minutes per medication pass but eliminates a major source of preventable patient harm. Widely cited as the single largest patient-safety technology improvement in nursing practice of the pre-EHR era.

• EHR/EMR mandates2009–2018

  EHR/EMR mandates: HITECH Act + Meaningful Use (Epic / Cerner / Meditech)

  The Health Information Technology for Economic and Clinical Health (HITECH) Act of 2009 established financial incentives and eventually penalties to drive EHR adoption across US hospitals. By 2014, nearly all non-federal acute care hospitals had adopted at least a basic EHR. For nurses, this was the most disruptive technology change since penicillin, but in the opposite direction: instead of making work faster, it made it slower. Studies published by 2016 documented nurses spending between a sizeable share of each shift on EHR documentation, more time than on direct patient contact. Clinicians overall were spending an estimated one-third to one-half of their workday interacting with EHR systems.

  Effect on the work

  The documentation burden created the commercial opportunity for AI nursing scribes a decade later.

• Epic Sepsis Model2017–2022

  Epic Sepsis Model: first AI clinical decision support for nurses

  Epic Systems deployed an AI-based sepsis prediction model across hundreds of hospitals starting around 2017. Nurses received EHR alerts when the model flagged high-risk patients. A 2021 study in JAMA Internal Medicine (University of Michigan cohort, n=27,697 hospitalizations) found the model correctly sorted sepsis vs. non-sepsis patients a large share of the time, and missed two-thirds of actual sepsis cases while generating high alert-fatigue volumes. The paper became a landmark in AI skepticism for clinical decision support: well-intentioned prediction models, deployed at scale without rigorous external validation, can train clinicians to ignore alerts rather than respond to them.

  Effect on the work

  Alert fatigue from poorly calibrated AI is now recognized as a patient-safety hazard in its own right. The Epic Sepsis Model controversy prompted JAMA Network, NEJM, and hospital accreditation bodies to call for external validation before AI clinical alerts go live.

• COVID-19 pandemic2020–2022

  COVID-19 pandemic: telehealth surge + travel nursing crisis

  COVID-19 compressed a decade of telehealth adoption into eight weeks: CMS waived telehealth restrictions, and nurses began conducting remote video assessments at scale for the first time. A large share of the RN workforce was considering leaving the profession by 2022.

  Effect on the work

  Travel nurse weekly rates peaked during Delta and Omicron waves. The labor crisis accelerated automation investment in AI documentation tools.

• Ambient AI nursing scribes2023–present

  Ambient AI nursing scribes: Dragon Copilot, Abridge, DAX for nurses

  The physician-facing ambient AI scribes (Nuance DAX, Abridge, DeepScribe) that emerged 2021-2023 began adapting their models for nursing documentation by 2024-2025. Microsoft's Dragon Copilot for Nurses, announced late 2025, listens to nurse-patient conversations in the background and produces structured flowsheet entries and nursing notes without manual typing. A 2025 JAMA Network Open randomized trial of DAX Copilot found significant reductions in documentation burden and improved patient engagement. Unlike the Epic Sepsis Model controversy, the ambient scribe use case has a clear, measurable benefit: nurses spend less time typing and more time at the bedside.

  Effect on the work

  Early health system deployments (Mercy health system in Chesterfield, MO was a cited pilot) report measurable increases in direct patient-care time after ambient nursing documentation.