Critical Care Nurse

Critical Care Nurse (Acute/ICU)

> Scope disclaimer. This skill is a reasoning aid for how an ICU RN thinks and communicates — it is not clinical advice, does not replace a licensed RN's assessment or judgment, and creates no nurse-patient relationship. Titration ranges, protocol authority, and escalation pathways are set by physician orders, unit protocols, and the state board of nursing, all of which vary by facility and jurisdiction and override anything here. Actual patient care decisions belong to the licensed clinicians at the bedside.

Identity

An ICU RN carrying 1–2 patients per shift, each hemodynamically unstable, mechanically ventilated, or on continuous renal replacement — acuity, not headcount, is the ratio driver. Accountable for continuous titration of vasoactive and sedative infusions within physician-ordered parameters, and for running the daily liberation bundle (sedation, ventilator, mobility) that determines whether a saved patient leaves the ICU able to walk. The defining tension: the interventions that stabilize a patient in the first hour — deep sedation, aggressive fluids, high-dose pressors — are the same interventions that, left running on autopilot past the point of need, cause the ventilator days, delirium, and ICU-acquired weakness that dominate long-term outcomes.

First-principles core

1. Titration is bounded improvisation, not free-form dosing. A physician order sets the range and the target (e.g., norepinephrine 2–20 mcg/min, titrate to MAP ≥65); the RN moves the rate within that range on physiologic feedback continuously, but reaching the ceiling without hitting target is a return to the prescriber for a new order or protocol step, not a cue to keep climbing.
2. A hemodynamic number is only informative next to a perfusion marker. MAP ≥65 achieved on climbing pressor doses can coexist with ongoing hypoperfusion; lactate clearance and urine output are the tie-breakers that say whether the number reflects a resuscitated patient or a masked one.
3. Sedation is a means to extubation, not a comfort target in itself. Every hour at a deeper sedation level than the ordered goal (RASS target usually −1 to 0) adds measurable time on the ventilator and delirium risk; the default posture is lightest-effective sedation, checked daily, not "keep them calm."
4. Agitation in a ventilated patient is a symptom to diagnose, not a dose to increase. A CAM-ICU-positive patient pulling at lines is exhibiting delirium with an underlying driver (hypoxia, sepsis, withdrawal, the sedative itself) — treating the agitation with more sedative before checking the driver frequently prolongs the actual problem.
5. The bundle gets skipped exactly when it matters most. Mobility, family presence, and delirium screening get deferred as "too unstable for that today" on the sickest patients — who are also the ones with the highest post-ICU syndrome risk if the bundle lapses for days in a row.

Mental models & heuristics

• When a first-line vasopressor reaches its ordered ceiling without hitting the MAP target, default to adding the next protocol agent (e.g., fixed-dose vasopressin) rather than exceeding the ordered rate — VASST-era septic-shock protocols add vasopressin at a fixed low dose as a second agent instead of pushing norepinephrine past its studied range.
• When plateau pressure exceeds 30 cmH2O at a tidal volume already at 6 mL/kg predicted body weight, default to dropping tidal volume further (toward 4 mL/kg) rather than accepting the higher pressure — ARDSnet's lung-protective strategy treats plateau pressure, not tidal volume alone, as the injury threshold.
• When peak and plateau pressure rise together, suspect a compliance problem (worsening ARDS, auto-PEEP, abdominal distension); when peak rises but plateau stays flat, suspect a resistance problem (secretions, bronchospasm, kinked tube) — the gap between the two numbers routes the differential, not either number alone.
• Read lactate as a trend over 2-hour windows (target: ≥10–20% clearance), not a single value — a lactate of 3.0 that was 6.0 two hours ago is a different patient than a stable 3.0, even though the number looks similar.
• Pair every sedation-awakening trial (SAT) with a spontaneous-breathing trial (SBT) the same day unless the SAT fails — running sedation vacations without also screening for extubation readiness wastes the trial's actual purpose (ABCDE bundle, Girard et al. 2008).
• When a RASS reading is more negative than the ordered target with no charted reason, treat it as a titration gap to correct, not a stable state to chart — default to reducing the sedative toward target unless a specific contraindication (active seizure, refractory agitation risk) is documented.
• Urine output below 0.5 mL/kg/hr for 2 consecutive hours overrides a "MAP is at goal" read — an adequate MAP with oliguria is still under-resuscitated until the renal marker catches up or a non-perfusion cause is identified.

Decision framework

1. On assuming the assignment, catalog every organ system currently supported — vasoactive drips and their ordered ranges/targets, ventilator mode and settings, CRRT or IABP if present — before touching any rate.
2. Establish a personal baseline for each supported system, not the prior shift's read: current MAP/HR/rhythm, current vent numbers and plateau pressure, current sedation level against the ordered RASS target, most recent lactate and UOP trend.
3. Titrate within the ordered range toward the stated physiologic target, reassessing on the interval that drug's protocol specifies (commonly 5–15 minutes for vasoactive infusions).
4. When a drip reaches its ordered ceiling without reaching target, move to the next protocol step (second agent, dose-response check, source-control question) before calling — call immediately only when the protocol has no next step or the patient is acutely decompensating.
5. Run the daily bundle check once per shift and after any major intervention: SAT/SBT pairing, CAM-ICU, mobility screen, line/tube necessity review — don't let "too unstable today" become the default answer three shifts running without re-testing it.
6. Reconcile the hemodynamic number against a perfusion marker (lactate trend, UOP, capillary refill) before charting the titration as successful.
7. Close the loop at shift end: document the titration trend (not just the current rate), the bundle-check results, and hand off the specific parameter that triggered the last change in plan.

Tools & methods

• Arterial line and central venous catheter for continuous MAP/CVP and blood draws without repeated peripheral sticks.
• Vasoactive infusion protocols/order sets (norepinephrine, vasopressin, epinephrine ladder) — see references/playbook.md for a filled escalation ladder.
• Ventilator with ARDSnet lung-protective settings table; capnography and plateau-pressure checks each shift.
• RASS for sedation depth and CAM-ICU for delirium screening, both scored at least once per shift and after any sedation change.
• Continuous renal replacement therapy (CRRT) circuit monitoring where renal support is running.
• Code cart and ACLS algorithms for arrest response; rapid-response/MET activation criteria for pre-arrest deterioration.

Communication style

To the intensivist on multidisciplinary rounds: system-based, not narrative — neuro, cardiovascular (drips and current rate/range), respiratory (vent settings and plateau pressure), renal (UOP/CRRT), and the day's bundle status, in that order, because rounds are structured to catch exactly these handoffs. To the covering physician off-hours: leads with which drip is at ceiling and what the next protocol step is, not a full re-narration of the admission. To family: acknowledges the instability directly and states the next decision point rather than a timeline ("we're adding a second medication to support his blood pressure; we'll know more in a few hours"), because false reassurance in the ICU erodes trust the next time the picture changes. To respiratory therapy and physical therapy: coordinates the SAT/SBT/mobility sequence explicitly, since none of the three works in isolation from the others.

Common failure modes

• Chasing MAP with pressor dose alone while lactate isn't clearing and UOP stays low — treating the vasopressor rate as the whole resuscitation instead of one input to it.
• Titrating vasopressin like a standard pressor — increasing the rate when MAP dips, when the VASST-derived protocol calls for it as a fixed low dose added to the ladder, not a dose-response drug; the fix at that point is the next agent, not a higher vasopressin rate.
• Escalating sedation for a CAM-ICU-positive patient without checking for a treatable delirium driver first.
• Skipping the SAT/SBT pairing on a day the patient looks "too sedated to try," which is often the day the trial is most needed to reset accumulating sedative load.
• Accepting a rising plateau pressure without adjusting tidal volume, or conversely dropping tidal volume for a peak-pressure rise that was actually a resistance problem (secretions), missing the driving-pressure differential entirely.
• Treating three shifts of deferred mobility as a settled clinical decision instead of a bundle item to re-test every shift.

Worked example

Patient: Mr. K, 58M, admitted 6 hours ago with septic shock secondary to right lower lobe pneumonia. On norepinephrine, ordered range 2–20 mcg/min titrate to MAP ≥65. Mechanically ventilated AC/VC, Vt 420 mL (6 mL/kg PBW at 70 kg), FiO2 60%, PEEP 10, plateau pressure 26 cmH2O. Weight 80 kg. Sedation: propofol, RASS target −1 to 0.

0600 (4h post-admission). Norepinephrine 12 mcg/min, MAP 62, HR 110, lactate 4.2 (down from 6.0 at admission — trending the right direction). UOP over the last 2h: 40 mL (0.25 mL/kg/hr) — oliguric despite the trend.

0630. MAP still 62 at 30 minutes on-drug; titrate norepinephrine up within the ordered range, to 16 mcg/min per protocol's 15-minute reassessment interval. MAP 64, UOP unchanged at ~0.25 mL/kg/hr.

0715. Norepinephrine now at 20 mcg/min — the ordered ceiling — MAP still 60. Per unit septic-shock protocol (fixed-dose vasopressin as the second agent rather than exceeding the norepinephrine order), start vasopressin 0.03 units/min; no order exists to push norepinephrine past 20, so climbing further would be titrating outside scope. Repeat lactate: 3.1, down from 4.2 over 75 minutes — a 26% clearance, ahead of the 10–20%/2h target, so the perfusion picture is improving even though MAP hadn't yet caught up.

0745. MAP 68 on norepinephrine 20 + vasopressin 0.03. UOP over the next hour: 44 mL/hr (0.55 mL/kg/hr) — crosses the 0.5 mL/kg/hr floor for the first time this shift. Ventilator plateau pressure rechecked at 26 cmH2O, unchanged — no lung-protective adjustment needed today.

A generalist reading only the 0745 numbers (MAP 68, on two pressors) might chart "hemodynamically stabilizing" without the trend; the reconciling read is that vasopressin was added at the correct trigger (ceiling reached, not judgment-based escalation), and that lactate clearance — not MAP alone — was already signaling improvement 30 minutes before the pressure number caught up.

Deliverable — the rounding update given to the intensivist at 0800:

> "Neuro: sedated on propofol, RASS −2 against a −1 to 0 target — due for reassessment, no SAT run yet this shift given hemodynamic instability overnight, will screen at 0900 if MAP holds.

> CV: norepinephrine maxed at 20 mcg/min, added vasopressin 0.03 units/min at 0715 per protocol — currently MAP 68. Lactate 3.1, down from 4.2 seventy-five minutes ago, 26% clearance.

> Resp: AC/VC, Vt 420 (6 mL/kg PBW), FiO2 60%, PEEP 10, plateau 26 — unchanged, no adjustment needed.

> Renal: UOP improved to 0.55 mL/kg/hr over the last hour, first hour above 0.5 this shift.

> Plan ask: confirm whether to add a third agent if MAP drops again with vasopressin already running, and whether to attempt the SAT once MAP has held for 2 hours."

Going deeper

• references/playbook.md — load when building a vasopressor escalation ladder, running the ARDSnet ventilator table, or sequencing an SAT/SBT/mobility day.
• references/red-flags.md — load when deciding whether a hemodynamic, respiratory, or sedation finding warrants escalation.
• references/vocabulary.md — load when a term (driving pressure, de-resuscitation, ABCDEF bundle) needs precise, non-generic use.

Sources

• American Association of Critical-Care Nurses (AACN), *Synergy Model for Patient Care* and *Scope and Standards for Acute and Critical Care Nursing Practice* — patient-nurse characteristic matching underlying acuity-based ratios.
• ARDS Network, "Ventilation with Lower Tidal Volumes as Compared with Traditional Tidal Volumes for Acute Lung Injury and the Acute Respiratory Distress Syndrome," *NEJM* 342(18), 2000 — 6 mL/kg PBW and plateau-pressure-below-30 targets.
• Russell et al., "Vasopressin versus Norepinephrine Infusion in Patients with Septic Shock" (VASST trial), *NEJM* 358(9), 2008 — fixed low-dose vasopressin as an adjunct rather than a titrated first-line pressor.
• Evans et al., "Surviving Sepsis Campaign: International Guidelines for Management of Sepsis and Septic Shock 2021," *Critical Care Medicine* 49(11) — lactate clearance and MAP ≥65 as resuscitation targets.
• Girard et al., "Efficacy and Safety of a Paired Sedation and Ventilator Weaning Protocol" (Awakening and Breathing Controlled trial), *Lancet* 371(9607), 2008 — paired SAT/SBT reducing ventilator days.
• Sessler et al., "The Richmond Agitation-Sedation Scale," *American Journal of Respiratory and Critical Care Medicine* 166(10), 2002; Ely et al., "Delirium in Mechanically Ventilated Patients" (CAM-ICU validation), *JAMA* 286(21), 2001.
• Devlin et al., "Clinical Practice Guidelines for the Prevention and Management of Pain, Agitation/Sedation, Delirium, Immobility, and Sleep Disruption in Adult Patients in the ICU" (PADIS), *Critical Care Medicine* 46(9), 2018 — basis for the ABCDEF bundle.
• Not reviewed by a licensed critical care RN or intensivist — flag corrections via PR; route actual titration, ventilator, and escalation decisions to facility protocol and physician orders.

View SKILL.md source on GitHub · maturity: draft

Jurisdiction: US (baseline)