General Survey, Pain and Mental Status

Before you place a stethoscope on anyone, you have already collected data. The way a patient sits, whether they finish a sentence without stopping for air, whether they are groomed, whether they meet your eye, whether they shift and wince when they move — all of it is assessment, and all of it is chartable. This chapter teaches the three assessments that frame every other examination in this book: the general survey, the systematic first impression; the pain assessment, which is a symptom only the patient can define and which nurses have historically undertreated; and the mental status assessment, where a change of a single point is often the earliest sign that a patient is deteriorating. You will also learn the four physical examination techniques — inspection, palpation, percussion, auscultation — and the one body region where that order deliberately changes. Get these right and the rest of the head-to-toe becomes a confirmation of what you already suspected.

Section 1The General Survey and the Four Techniques

What the general survey is, and the four domains you scan

The general survey is a deliberate, structured first impression. It is not “eyeballing” the patient. It is a scan that you perform the same way on every patient so that when something is different, you notice it. It begins the moment the patient is in your field of view — in the doorway, in the hallway, in the waiting room — and it continues while you introduce yourself, wash your hands and take vital signs. Experienced nurses do this in under a minute and can tell you, before the monitor is attached, roughly how sick the person is. That skill is not intuition. It is a checklist run so many times that it has become automatic, and you build it by running the checklist consciously until it is.

Scan four domains, always in the same order. Physical appearance: apparent age compared with stated age, level of consciousness, skin color and condition, facial expression and symmetry, hygiene and grooming, dress appropriate to the setting and the weather, and any odor (ketones, alcohol, urine, infected wound, poor hygiene). Body structure: height and build, apparent nutritional state, symmetry of the body, posture, and any obvious deformity, amputation, edema or muscle wasting. Mobility: gait, use of assistive devices, range of motion, how the patient moves from chair to bed, and any involuntary movement such as tremor, tic or restlessness. Behavior: eye contact, facial expression, mood and affect, speech (rate, volume, clarity, articulation), and the relationship between the patient and anyone accompanying them.

Two survey findings deserve special naming because they carry so much weight. Tripod positioning — the patient sitting leaning forward with arms braced on knees or a bedside table — means the patient is recruiting accessory muscles to breathe and should never be laid flat “so you can examine them.” And the number of words a patient can say between breaths is a bedside respiratory measurement you can chart: full sentences, three-to-four-word phrases, or single words. A patient who was speaking in full sentences at 0800 and is speaking in three-word phrases at 1200 has deteriorated, even if the recorded respiratory rate has not changed.

The four techniques, and the order that changes

Every physical examination in this book is built from four techniques, performed in this sequence: inspection, palpation, percussion, auscultation. The logic is that each technique is progressively more disruptive, so the least invasive information is gathered while the body is still undisturbed. There is one deliberate exception. For the abdomen, the sequence is inspect, auscultate, percuss, palpate. Percussing and palpating the abdomen mechanically stimulates the bowel and alters bowel sounds, so listening must happen before you touch. Say that exception out loud until it is reflex, because it is one of the most commonly missed items on both exams and clinical checkoffs — and a nurse who palpates first and then reports “hyperactive bowel sounds” has reported an artifact they created.

Inspection is purposeful looking. Provide adequate light — natural daylight is best, especially for assessing color in darker skin tones — expose only the area being examined, keep the room warm, and compare the side you are examining with its opposite. Inspect before you touch, and give it real time; most nurses under-inspect. Colour findings must be judged where pigment cannot hide them, in every patient and not only in patients with dark skin: look at the oral mucosa and tongue, the palpebral conjunctivae, the lips and nail beds, and the palms and soles. Cyanosis is not reliably “blue” — in richly pigmented skin it shows as grey, ashen or dull darkening, and it is confirmed at the tongue, oral mucosa and conjunctivae rather than at the fingertips. Pallor appears as ash-grey or yellow-brown skin with pale conjunctivae, pale palms and pale mucous membranes. Jaundice is judged at the sclera next to the iris, the hard palate, and the palms and soles, remembering that harmless subconjunctival fat can look yellow in darkly pigmented patients. Erythema and inflammation are frequently invisible in dark skin, so palpate for warmth, induration, tightness and tenderness and compare with the opposite side instead of looking for redness — relying on redness alone is exactly how early pressure injury and cellulitis are missed in patients with dark skin. Palpation uses different parts of the hand for different jobs: finger pads for texture, moisture, pulsation, masses and edema; the dorsum (back) of the hand for temperature, because it is thinner and more sensitive to heat; the ulnar surface and base of the fingers for vibration such as tactile fremitus and thrills. Light palpation depresses the tissue about 1 to 2 cm; deep palpation depresses about 4 to 5 cm and is done with one hand or with two hands stacked. Warm your hands, keep nails short, start light, and palpate tender or painful areas last.

Percussion produces sound that tells you what lies beneath. Hyperextend the middle finger of your non-dominant hand (the pleximeter), press its distal joint firmly against the skin, and keep your other fingers off the surface so they do not damp the vibration. Strike the pleximeter at the distal interphalangeal joint with the tip of your dominant middle finger (the plexor), using a quick, sharp motion from the wrist, two taps, then withdraw immediately. Learn five tones: tympany (drumlike, over the gastric bubble and intestine), resonance (hollow, over normal lung), hyperresonance (booming, over hyperinflated lung as in emphysema or pneumothorax), dullness (thudlike, over liver, spleen, a full bladder or consolidated lung), and flatness (very dull, over muscle and bone). Auscultation uses the diaphragm for high-pitched sounds (breath sounds, bowel sounds, normal S1 and S2) pressed firmly against skin, and the bell for low-pitched sounds (extra heart sounds S3 and S4, the murmur of mitral stenosis, bruits) rested lightly, because pressing the bell hard turns it into a diaphragm. Never auscultate through clothing, and clean the stethoscope with alcohol before and after every patient.

Height, weight, BMI and the numbers that matter

Measure, do not ask. Reported heights and weights are consistently wrong, and in a patient receiving weight-based medications such as heparin, vancomycin or chemotherapy, a guessed weight is a medication error waiting to happen. Weigh adults on the same scale, at the same time of day, in similar clothing, and record in kilograms because that is the unit drug calculations use. For daily weights — the single most sensitive measure of fluid status in heart failure and renal disease — weigh before breakfast, after voiding, in the same gown. A gain of 1 kg corresponds to roughly 1 litre of retained fluid; a gain of more than 1 kg (about 2 lb) in 24 hours or 2.3 kg (about 5 lb) in a week is a reportable finding in a heart failure patient.

Body mass index (BMI) is weight in kilograms divided by height in metres squared. Adult categories: underweight below 18.5; healthy weight 18.5 to 24.9; overweight 25.0 to 29.9; obesity class I 30.0 to 34.9, class II 35.0 to 39.9, class III 40.0 and above. BMI is a screening number, not a diagnosis: it does not distinguish muscle from fat, and it misclassifies very muscular patients, patients with large amounts of ascites or edema, and older adults who have lost lean mass. Pair it with waist circumference, measured at the top of the iliac crest at the end of a normal exhalation; greater than 102 cm (40 in) in men and greater than 88 cm (35 in) in non-pregnant women indicates increased cardiometabolic risk. Unintentional weight loss is significant at greater than 5 percent of body weight in one month or greater than 10 percent in six months, and it should trigger a nutrition referral and a search for cause.

Measurement changes across the lifespan. Infants and children under 24 months are measured as recumbent length on a length board with two people, weighed nude on an infant scale in kilograms, and have head circumference measured at the widest point (just above the eyebrows and around the occipital prominence) at every well visit through 24 to 36 months. All three are plotted on growth charts; a single point means little, but a child who crosses two major percentile lines downward is a concern. In children aged 2 to 20, BMI is interpreted as a percentile for age and sex: underweight below the 5th, healthy 5th to 84th, overweight 85th to 94th, obesity at or above the 95th. In pregnancy, recommended total weight gain depends on pre-pregnancy BMI: 12.5 to 18 kg (28 to 40 lb) if underweight, 11.5 to 16 kg (25 to 35 lb) if healthy weight, 7 to 11.5 kg (15 to 25 lb) if overweight, and 5 to 9 kg (11 to 20 lb) if obese. Older adults lose roughly 1 cm of height per decade after age 40 from disc compression and vertebral change, so a measured height noticeably below the patient’s historic height may reflect osteoporotic vertebral compression fractures rather than measurement error.

Section 2Pain: the Symptom Only the Patient Can Define

The ethical baseline and why pain is measured, not judged

Margo McCaffery’s working definition remains the standard of nursing practice: pain is whatever the experiencing person says it is, existing whenever they say it does. That sentence is not sentiment, it is a rule of practice. Pain is a subjective symptom with no laboratory value and no imaging study that confirms it. Your job is to measure it — consistently, with a validated tool, in the patient’s own words — and to act on the measurement. Your job is not to decide whether the number is deserved. Documented undertreatment of pain in Black and Hispanic patients, in women, in older adults, in patients with substance use disorders and in patients who cannot speak English is one of the best-established disparities in health care. It happens through exactly the mechanism you are guarding against: a clinician substitutes their own judgment for the patient’s report.

Two clinical facts protect patients from that error. First, vital signs do not confirm or refute pain. Acute pain often raises heart rate, blood pressure and respiratory rate and can cause diaphoresis and pupil dilation, but those responses adapt within hours to days. A patient with chronic pain or long-standing severe pain frequently has entirely normal vital signs. Never chart “pain rated 8 out of 10 but vital signs stable, appears comfortable” as if the second half undermines the first. Second, behavior is not a lie detector. Patients smile, joke, watch television and sleep through severe pain; sleep is a coping behavior and an exhausted patient may sleep at a pain score of 8. Chart what you observe alongside what the patient reports, without editorializing.

Use a professional interpreter for any patient with limited English proficiency — never a family member, never a child — and remember that pain expression is culturally patterned. Some patients are raised to bear pain silently and view complaint as weakness; others express pain loudly as a normal signal for help. Neither pattern tells you anything about the intensity of the pain. Ask, in the patient’s language, what word they use for pain: hurt, ache, sore, pressure, discomfort, burning, tight. Patients frequently deny “pain” and then agree readily to “pressure” or “soreness.”

PQRSTU and OLDCARTES: the actual questions

Two mnemonics organize the full symptom analysis, and either is acceptable. PQRSTU: Provocation and palliation — “What brings it on? What makes it better? What makes it worse?” Quality — “Describe the pain in your own words” (ask open-ended first; only offer descriptors such as burning, stabbing, aching, cramping if the patient cannot generate any). Region and radiation — “Point to it with one finger. Does it travel anywhere?” Severity — a number on a validated scale, both now and at its worst and best in the last 24 hours. Timing — onset, duration, frequency, constant or intermittent, and what the patient was doing when it started. Understanding and effect on you — “What do you think is causing it? What can you not do because of it? What is an acceptable level of pain for you?”

OLDCARTES covers the same ground in a different order: Onset, Location, Duration, Character, Aggravating factors, Relieving factors, Treatments tried, Effect on daily life, Severity. The two elements students most often skip are also the two that most change the plan of care: treatments already tried (including what worked at home, over-the-counter drugs, heat, ice, position, cannabis, herbal preparations and prayer) and the patient’s functional goal. “Zero out of ten” is rarely a realistic postoperative goal; “a 3, so I can get up and walk to the bathroom” is a goal you can actually work toward and evaluate against, and it is what belongs in the care plan.

Ask about pain at more than one moment. A patient at rest may report 2 out of 10 and then report 8 out of 10 on standing; incident pain, or pain with movement, is what actually determines whether the patient will ambulate, cough, deep-breathe and avoid pneumonia and venous thromboembolism. Chart both: “3/10 at rest, 8/10 with ambulation.” Also screen for the effect of pain on sleep, appetite, mood, concentration and relationships, because unrelieved pain drives delirium, depression and functional decline, and because those effects justify escalating the plan when a number alone does not persuade a prescriber.

Choosing the right pain scale

Self-report is always the gold standard, so the first question is whether the patient can self-report. If yes, choose a self-report tool matched to development, literacy and language. If no — a preverbal infant, an intubated critically ill adult, a patient with advanced dementia, a patient with a profound intellectual disability — you must use a validated observational tool, and you must not record “denies pain” for a patient who cannot answer. In that situation the sequence is: attempt self-report; look for an obvious cause of pain (surgery, fracture, wound care, positioning, a full bladder); observe behavior with a validated scale; ask a family member or regular caregiver what this person looks like when they hurt; and, when there is a plausible cause, consider an analgesic trial with reassessment of behavior.

Types of pain, reassessment and opioid safety

Classify pain by mechanism, because the words the patient uses point to the treatment. Nociceptive somatic pain arises from skin, muscle, bone and joint; it is well localized and described as aching, throbbing, sharp or gnawing — a fracture, a surgical incision, a pressure injury. Nociceptive visceral pain arises from organs; it is poorly localized, deep, squeezing, cramping or pressure-like, often with nausea and diaphoresis, and it refers to predictable places — diaphragmatic irritation to the shoulder tip, cardiac ischemia to the jaw, left arm, or between the shoulder blades, biliary colic to the right scapula, renal colic from flank to groin. Neuropathic pain arises from damaged nerve tissue; it is burning, shooting, electric, stabbing, or accompanied by numbness, tingling and hypersensitivity to light touch, and it responds poorly to opioids and better to agents such as gabapentin, pregabalin and duloxetine. When a patient says “it feels like electricity” or “the sheet hurts my feet,” those exact words belong in your note because they change the prescription.

Also classify by duration. Acute pain has an identifiable cause, is expected to resolve as tissue heals, and is protective. Chronic (persistent) pain continues beyond expected healing, commonly defined as more than 3 months, is no longer protective, and reshapes sleep, mood and function. Breakthrough pain is a transient flare above otherwise controlled baseline pain and requires a short-acting rescue dose, not a change in the long-acting regimen. Cancer pain and end-of-life pain often combine all of these. And know that pain unrelieved by the ordered regimen is itself a finding to escalate — not a reason to wait for the next scheduled dose.

Reassess after every intervention. The general standard, which your facility policy will specify precisely: reassess roughly 15 to 30 minutes after an intravenous analgesic, 30 to 60 minutes after intramuscular or subcutaneous, and about 60 minutes after an oral dose. An intervention that is not reassessed is not documented care. Reassessment includes the pain score, the patient’s function (did they get up, cough, sleep), and a sedation and respiratory check for any opioid.

The single most important opioid safety concept is this: sedation precedes respiratory depression. By the time the respiratory rate falls, the patient has already been over-sedated for some time. Monitor sedation with a validated tool, most commonly the Pasero Opioid-Induced Sedation Scale (POSS): S — sleep, easy to arouse (acceptable); 1 — awake and alert (acceptable); 2 — slightly drowsy, easily aroused (acceptable, continue to monitor); 3 — frequently drowsy, arousable but drifts off to sleep during conversation (unacceptable: hold or reduce the opioid dose, typically by 25 to 50 percent per order, monitor sedation and respiratory status closely, add supplemental oxygen only per protocol, notify the prescriber); 4 — somnolent, minimal or no response to verbal or physical stimulation (unacceptable and an emergency: stop the opioid, stay with the patient, call for help and rapid response, prepare naloxone, support airway and breathing). Do not confuse POSS with the Richmond Agitation-Sedation Scale (RASS), which measures sedation and agitation for a different purpose in critical care.

Section 3Mental Status and Level of Consciousness

Level of consciousness: AVPU, RASS and the Glasgow Coma Scale

Level of consciousness (LOC) is the most sensitive single indicator of neurologic change, and it is the finding that most often deteriorates first when a patient is becoming hypoxic, septic, hypoglycemic or bleeding into the head. Describe it behaviorally, not with a label. “Lethargic” means different things to different nurses; “opens eyes to loud voice, answers name correctly, falls back asleep within 10 seconds” means the same thing to everyone. The traditional descriptive ladder runs alert, lethargic (drowsy but rousable to voice), obtunded (rousable with difficulty, limited responses), stuporous (rousable only with vigorous or painful stimulus), and comatose (no rousable response). Use those words if your unit does, but always add the behavior.

AVPU is the fast field screen: Alert, responds to Voice, responds to Pain (a central stimulus — trapezius pinch, supraorbital ridge pressure or, where your facility permits it, mandibular pressure — applied only when needed and for no more than about 10 seconds; fingertip pressure on the nail bed is a peripheral stimulus — it is the standard opening stimulus for eye opening and limb response on the GCS, but because a limb can pull away by spinal reflex it cannot establish localization or, by itself, the P of AVPU, so escalate to a central stimulus before you call a patient unresponsive; the sternal rub is no longer recommended, because it bruises and abrades the skin without adding information), Unresponsive. It is quick and coarse, and it belongs in a rapid response or a first look, not in an ongoing neurologic flowsheet. RASS, the Richmond Agitation-Sedation Scale, runs from +4 (combative, violent) through 0 (alert and calm) to −5 (unarousable, no response to voice or physical stimulation); most sedated critical care patients are targeted at 0 to −2. RASS is also the level-of-consciousness input for delirium screening.

The Glasgow Coma Scale scores three independent responses and reports them separately as well as summed — document “E3 V4 M5 = GCS 12,” because the components carry information that the total hides. Always score the best response obtained, and score the best side for motor. Interpretation: 13 to 15 mild, 9 to 12 moderate, 8 or less severe brain injury — and a GCS of 8 or less is the classic threshold at which the airway is considered unprotected and intubation is anticipated (“less than 8, intubate”). The lowest possible score is 3, not 0. If a component cannot be tested — an intubated patient has no testable verbal response, a swollen orbit no testable eye opening — record it as not testable rather than guessing a 1. A drop of 2 or more points is an emergency, and so is any new asymmetry between sides.

Pair LOC with pupils. Normal adult pupils are round, equal and roughly 3 to 5 mm in ambient room light (2 to 6 mm is the usual range — larger in young patients and in a darkened room, smaller in older adults), and briskly reactive to light both directly and consensually — charted as PERRLA. New unequal pupils (anisocoria greater than 1 mm), a sluggish or fixed pupil, or a unilaterally “blown” dilated pupil suggests third cranial nerve compression from rising intracranial pressure and is an emergency. Pinpoint pupils suggest opioid effect or a pontine lesion. The Cushing triad — rising systolic pressure with widening pulse pressure, bradycardia and irregular respirations — is a late sign of increased intracranial pressure; if you wait for it, you have waited far too long.

The components of the mental status examination

Beyond consciousness, the mental status examination has a fixed set of components, and most of them are collected simply by conducting the health history attentively. Appearance and behavior: grooming, hygiene, dress, posture, motor activity, eye contact, cooperation. Speech: rate, volume, fluency, articulation, spontaneity — note dysarthria (slurred, a motor problem) versus aphasia (a language problem: expressive, receptive or global). Mood and affect: mood is the sustained emotional state the patient reports in their own words (“How have your spirits been?”); affect is the emotional expression you observe moment to moment. Describe affect as full, constricted, blunted, flat or labile, and state whether it is congruent with the content of what is being said.

Thought process is how thinking is organized: logical and goal-directed is expected; abnormalities include circumstantial (eventually arrives at the point after excessive detail), tangential (never arrives), flight of ideas, loose associations, perseveration (repeating the same response), neologisms and word salad. Thought content is what the patient is thinking: delusions, obsessions, phobias, and — asked directly of every patient in whom it is a possibility — thoughts of harming self or others. Perception: illusions (a misinterpreted real stimulus) and hallucinations (a perception without a stimulus). Auditory hallucinations are most characteristic of primary psychiatric illness; visual and tactile hallucinations point toward delirium, alcohol or sedative withdrawal, or a drug effect and require a medical work-up, not a psychiatric label. Insight and judgment: does the patient understand they are ill, and can they reason through a realistic problem such as “What would you do if you ran out of your blood pressure medication on a Sunday?”

Cognition is tested actively. Orientation ×4 is person, place, time and situation; ask it conversationally rather than as an interrogation (“Just to make sure I have things right — can you tell me your full name, where we are, what month it is, and what brought you in?”). Orientation is lost in a predictable order: time first, then place, then situation, and person last. Attention is the hallmark of delirium and is tested with digit span (5 to 7 digits forward is normal for an adult), reciting the months of the year backward, spelling “WORLD” backward, or serial sevens. Memory: immediate (register three unrelated words and repeat them at once), recent (recall the same three words after 3 to 5 minutes of distraction), and remote (verifiable historical detail). Language: naming, repetition, comprehension, reading, writing. Higher function: similarities (“how are an apple and an orange alike?”) and simple calculation. Avoid proverb interpretation as a stand-alone test; it is heavily influenced by culture, language and education.

Cognitive screening tools and how to use them honestly

Screening tools give you a number to compare over time, but they are not diagnoses and they perform badly when misapplied. The Mini-Cog is the most practical bedside screen: register three unrelated words, have the patient draw a clock face with all the numbers and set the hands to a specified time such as ten past eleven, then ask for the three words. Recall scores 0 to 3 (one point per word); a normal clock scores 2 and an abnormal clock scores 0, for a total of 0 to 5. A total of less than 3 — that is, 0, 1 or 2 — is a positive screen and warrants full evaluation; 3, 4 or 5 is a negative screen but does not rule out milder impairment. It takes about 3 minutes and is relatively resistant to education and language effects.

The Mini-Mental State Examination (MMSE) scores 0 to 30 across orientation, registration, attention, recall, language and construction; a score below 24 suggests cognitive impairment, though the cut point must be adjusted for education. The Montreal Cognitive Assessment (MoCA) also scores 0 to 30, is more sensitive to mild cognitive impairment and executive dysfunction, defines 26 or above as normal, and adds one point for patients with 12 or fewer years of formal education. Both require training and, in the case of the MMSE and MoCA, permission or certification depending on the version and setting.

Use these tools honestly. A patient who cannot hear you, cannot see the paper, has no glasses, is in severe pain, has just received an opioid, is being tested at 0300, or is being tested in their second language will score poorly for reasons that have nothing to do with cognition. Correct the sensory and comfort barriers first — glasses on, hearing aids in and switched on, lights up, pain treated, an interpreter present — then screen. And never chart a screening score as a diagnosis: “Mini-Cog 2 of 5” is a finding; “patient has dementia” is a medical diagnosis you do not make.

Section 4Delirium, Dementia and Depression

Telling the three apart

Confusion is not a diagnosis; it is a symptom with three common explanations that are treated completely differently. Delirium is an acute, fluctuating disturbance of attention and awareness caused by a medical problem — it is a medical emergency and it is usually reversible. Dementia (major neurocognitive disorder) is a slow, progressive loss of cognitive function that is not reversible. Depression can mimic dementia so closely that it was historically called pseudodementia, and it is treatable. The single question that separates them is: what was this person like two weeks ago? Delirium is defined by acute change from baseline, which means you must find the baseline — from family, from the group home staff, from the long-term care nurse, from the prior admission’s notes. A patient who has been oriented only to person for six years is at their baseline. A patient who was balancing a checkbook last Tuesday and cannot say the month today is delirious until proven otherwise.

Screening for delirium and preventing it

Delirium affects a very large share of hospitalized older adults, is independently associated with prolonged stay, functional decline, new nursing home placement and death, and is missed most of the time. It is missed because the hypoactive subtype — quiet, withdrawn, drowsy, slow, not making trouble — is the most common and carries the worst prognosis, while nurses are trained by experience to notice the hyperactive patient who is pulling at lines. A quiet patient who is not their usual self is the one to worry about.

The Confusion Assessment Method (CAM) is the standard bedside screen and requires four features. Delirium is present when you have Feature 1 AND Feature 2, plus either Feature 3 OR Feature 4. Feature 1: acute onset or fluctuating course — is there evidence of an acute change from baseline, and does the behavior come and go? Feature 2: inattention — does the patient have difficulty focusing, become easily distracted, lose track of what is being said? Feature 3: disorganized thinking — rambling, illogical, unpredictable switching of subject. Feature 4: altered level of consciousness — anything other than alert (vigilant, lethargic, stuporous, comatose). The ICU version (CAM-ICU) uses RASS and a picture and letter task so that intubated patients can be screened without speaking. Screen at least once a shift in any at-risk patient and any time the family says “that is not how she normally is.”

When a screen is positive, look for the cause with a memory aid such as PINCH ME: Pain, Infection (urinary tract infection and pneumonia are the classics), Nutrition, Constipation and urinary retention, Hydration, Medication (benzodiazepines, anticholinergics including diphenhydramine, opioids, steroids, and abrupt withdrawal of alcohol or benzodiazepines), Environment (sensory deprivation, sleep disruption, ICU, restraints). Check a glucose and an oxygen saturation immediately — hypoglycemia and hypoxia are fast, common and correctable. Prevention is largely nursing work: get glasses and hearing aids on the patient, reorient with a visible clock and calendar, keep familiar objects and family present, mobilize early, protect sleep by clustering care and dimming lights, maintain hydration and bowel function, treat pain adequately, avoid deliriogenic drugs, and use physical restraints only as a genuine last resort because they increase agitation, injury and delirium duration.

Depression, mood and suicide risk

Screen for depression routinely, because it is common, treatable and frequently masked by physical complaints. The PHQ-2 asks about the last two weeks: little interest or pleasure in doing things, and feeling down, depressed or hopeless, each scored 0 (not at all) to 3 (nearly every day). A total of 3 or more is a positive screen and should be followed by the PHQ-9, where a score of 10 or more indicates at least moderate depression. In older adults the Geriatric Depression Scale (GDS-15) is better tolerated, and a score of 5 or more suggests depression. Note that in older adults depression more often presents as somatic complaints, memory complaints, irritability or withdrawal than as reported sadness.

Question 9 of the PHQ-9 asks about thoughts of being better off dead or of self-harm, and any positive answer requires immediate action, not a note for the next shift. Ask about suicide directly and plainly: “Are you having thoughts of killing yourself?” Asking does not plant the idea and does not increase risk — that is a myth, and the evidence contradicts it. If the answer is yes, use a structured tool such as the Columbia Suicide Severity Rating Scale (C-SSRS) to assess ideation, intent, plan, means and prior attempts. Do not leave the patient alone, remove means from the environment, initiate your facility’s suicide precautions, and notify the provider and rapid response or behavioral health team immediately. Never promise confidentiality about safety information, and never make a “no-harm contract” a substitute for supervision and a real safety plan. In the United States, patients and families can reach the Suicide and Crisis Lifeline by calling or texting 988.

Warning signs that raise concern include hopelessness, giving away possessions, putting affairs in order, a sudden unexplained calm after a period of severe depression, increased substance use, withdrawal from others, and access to firearms or a stockpile of medication. Risk is elevated in older men, in patients with a prior attempt, in those with serious or painful medical illness, in those recently discharged from psychiatric care and in those who are socially isolated. Document exactly what you asked and exactly what the patient answered, in quotation marks.

Section 5Documenting, Escalating and Delegating

How to chart what you found

Charting rules that apply to everything in this chapter: write objectively, quantify with numbers and named scales, quote the patient when the exact words matter, timestamp every entry, and chart what you did about the finding and what happened next. Avoid conclusions you did not test (“drug-seeking,” “non-compliant,” “appears fine”), avoid words that mean different things to different readers (“lethargic,” “good,” “tolerated well”), and never chart an assessment you did not personally perform. If a finding is abnormal and you notified someone, document the name, the time, what you reported and what was ordered.

Red flags: escalate now

Escalate immediately — call the provider, activate rapid response, or both — for any of the following. Any acute change in level of consciousness or mental status from the patient’s baseline, including new inattention. A GCS drop of 2 or more points, or GCS 8 or less. New pupil asymmetry, a fixed or sluggish pupil, or new unilateral weakness or facial droop. Stroke signs screened with BE-FAST — Balance loss, Eyes (vision change), Face droop, Arm drift, Speech change, Time to call — and always establish and report the last known well time, because thrombolysis and thrombectomy windows depend on it. Sudden severe headache described as the worst of the patient’s life. New chest pain, or pain described as crushing, tearing or ripping. Pain that is severe, unrelieved by the ordered regimen, or disproportionate to the physical findings — this is the classic pattern of compartment syndrome and of mesenteric ischemia. A rigid, board-like abdomen with rebound tenderness. POSS 3 or 4 in a patient receiving opioids, or a respiratory rate below 8 to 10 with sedation. Any suicidal or homicidal ideation. Cushing triad. And any patient whose family says clearly that this is not how the patient normally is.

One more escalation trigger deserves emphasis because it is entirely nursing-driven: the patient just does not look right. Deterioration frequently announces itself as a change in the general survey — a patient who was chatty is now quiet, a patient who was sitting up is now slumped, a patient whose color was good is now gray. That impression is legitimate clinical data. Convert it into specifics (respiratory rate, SpO2, work of breathing, mental status, capillary refill, urine output), then escalate.

Escalating with SBAR

SBAR gives your report a shape the receiving clinician expects, and it forces you to end with a request instead of a description. S — Situation: who you are, who the patient is, and the one-sentence problem. B — Background: the relevant history in three or four sentences, not the whole chart. A — Assessment: your vital signs, your findings, and your clinical impression — nurses are permitted and expected to say what they think is happening. R — Recommendation: what you want, stated explicitly, with a timeframe. Write the numbers down before you call and have the chart, the medication record and the last set of vitals in front of you.

A worked example. S: “Dr. Nunez, this is Amara Boyd, RN, on 4 West. I am calling about Mrs. Ellen Rowe in room 412. She has had an acute change in mental status in the last four hours.” B: “She is 81, admitted two days ago with a right hip fracture, surgery yesterday morning, history of hypertension and hypothyroidism. She has a urinary catheter placed preoperatively and she is on scheduled oxycodone. At 0800 she was alert and oriented ×4 and joking with her daughter.” A: “Right now her CAM is positive — acute onset with fluctuation, inattention, and disorganized thinking. She is oriented to person and place only, thinks the year is 1998, and reports seeing children at the window. Her temperature is 38.1 orally, heart rate 110, blood pressure 102 over 58, respirations 24, oxygen saturation 92 percent on room air, glucose 96. Her urine is cloudy and strong smelling. I think she has a urinary tract infection with delirium, and I am concerned about early sepsis.” R: “I would like you to come assess her within the next 15 minutes. I am requesting orders for blood cultures before antibiotics, a urinalysis with culture, a lactate and CBC, and an order for intravenous fluids. I would also like to discuss holding her oxycodone and switching to a scheduled non-opioid, and removing the catheter if she no longer needs it. Do you want me to activate rapid response while you are on your way?”

Delegation, supervision and scope

The rule that governs everything in this section: the registered nurse may delegate a task, but never the nursing process. Assessment, analysis and nursing diagnosis, planning, patient and family teaching, evaluation of outcomes, and any care of an unstable or unpredictable patient remain with the RN. Unlicensed assistive personnel (UAP) may collect and report data; only the RN interprets it. Apply the five rights of delegation every time: the right task, under the right circumstances, to the right person, with the right direction and communication, and with the right supervision and evaluation. And remember that delegating does not transfer accountability — you remain accountable for the outcome and for verifying that the task was done.

Give direction that is specific and includes the report-back threshold, because “let me know if anything changes” is not a delegation instruction. Say instead: “Please take Mr. Alvarez’s vital signs at 1000 and 1400 and tell me right away if his heart rate is over 110, his respiratory rate is over 24, his systolic is under 100, or his oxygen saturation is under 92 percent. Also tell me immediately if he seems more sleepy than he was this morning, even if the numbers look fine.” Then verify. The UAP who tells you a patient “seems more sleepy” has just handed you the earliest sign of opioid-induced sedation, and your job is to go look.

Words worth knowing