Nursing care
Pain in the Cognitively Impaired: the method, the errors, and the exam
Written and reviewed by Dana Whitfield, RN, MSN · 5 min read · Updated September 2026
Short answer
Pain in the cognitively impaired is assessed with a behaviour observation scale, not a self-report tool. You watch breathing, vocalisation, facial expression, body language and consolability, because the patient cannot tell you where it hurts or how much. A rising score across these five categories means treat the pain, then reassess.
Why this skill decides answers
A patient with advanced dementia does not fail a 0-10 scale because the pain is absent. They fail it because the words are gone. If you chart "unable to assess pain" and move on, the exam will mark that wrong every time, and so will your unit's audit. Untreated pain in this population shows up as agitation, resistance to care, and a fall risk that gets labelled "behavioural" when it is actually orthopaedic.
This is why a behaviour scale exists as a distinct skill rather than a footnote to pain assessment generally. It converts what you can observe into a number you can act on and trend. The five categories are fixed: breathing, negative vocalisation, facial expression, body language, and consolability. Miss one and you underscore the patient. Score them all and a hip fracture stops looking like sundowning.
How to do it reliably
Watch before you touch. Observe breathing pattern and vocalisation for at least two minutes at rest, ideally before repositioning or wound care, because movement will spike every category and give you a false ceiling. Note facial expression next: a frown, grimace, or rapid blinking counts even without a sound. Then body language, rigid, tense, fists clenched, knees drawn up, and finally whether the patient can be consoled by voice, touch, or distraction.
Score each category 0, 1, or 2 against the defined descriptors, sum them, and treat a total of 2 or more as pain requiring intervention on most validated tools. Reassess 30 to 60 minutes after an analgesic to confirm the score has fallen, not just that time has passed. Do the same assessment during a known painful stimulus, such as a dressing change, to catch pain that rest conceals.
The common errors
The most frequent mistake is scoring only the face. A calm expression with laboured, noisy breathing and a clenched fist is not a zero, it is pain the nurse has half-read. The second error is scoring during activity and treating that number as baseline, which inflates every subsequent comparison and can lead to over-sedation instead of correct dosing.
The third error is skipping reassessment. A single score at the start of a shift tells you nothing about whether the intervention worked. The fourth is assuming a non-verbal patient feels less pain because they show it differently; the evidence does not support that, and grading pain as "probably not that bad" because the patient is quiet is a documented source of undertreatment in this group.
Drills that build it
Practise scoring from video rather than only from written vignettes. Real facial expression and breathing pattern are harder to categorise on a screen than on paper, and that friction is the point, it trains the same discrimination you will need at the bedside. Pause the clip at 30 seconds and force yourself to assign a number to each of the five categories before watching further.
Pair drills with a peer: one of you narrates observed behaviour without naming a diagnosis, the other scores blind. Then compare and argue the discrepancy category by category. This catches the habit of anchoring on one obvious cue, usually the face, and ignoring the other four.
Exam application
NCLEX questions on this topic usually present a patient who cannot self-report, most often severe dementia, sedation, or intubation, alongside a cluster of behaviours. The correct answer is nearly always the option that names or applies a structured behaviour assessment before medicating, rather than either withholding treatment for lack of a number or medicating on a hunch.
Watch for distractors that offer a numeric self-report scale as the answer; if the stem states the patient is non-verbal or severely cognitively impaired, a 0-10 scale is the wrong tool regardless of how familiar it looks. Also watch for options that treat agitation alone as a behavioural diagnosis when the stem includes grimacing or guarding, since those are pain cues being mislabelled.
Quick reference
Five categories, every time: breathing, vocalisation, facial expression, body language, consolability. Observe at rest first, then during a known painful activity, and reassess after treatment to confirm effect.
If the patient cannot self-report, do not ask them to. Use the behaviour scale, treat a rising score, and document the specific behaviours you saw, not just the total number, so the next nurse can trend it too.
The next step on this is the same as on everything else here: answer questions and read the rationales. Our basic care and comfort practice questions are the closest set to what this page covers.
One question from the basic care and comfort set
A nurse is preparing to feed a client who had a stroke and has mild dysphagia. Which action best reduces the risk of aspiration?
Rationale
Upright at 90 degrees with a chin tuck narrows the airway entrance and directs the bolus toward the esophagus — it is the single highest-yield positioning intervention for dysphagia. Thin liquids and straws move fastest and aspirate most easily, side-lying removes the gravity you want, and large bites overwhelm a swallow that is already impaired.
Answer: B
Common questions
What is the PAINAD scale used for?
PAINAD (Pain Assessment in Advanced Dementia) is a behaviour observation tool for patients who cannot self-report pain reliably. It scores breathing, negative vocalisation, facial expression, body language, and consolability from 0 to 2 each, giving a total from 0 to 10. It is used most often in advanced dementia, but the same five-category logic applies to other non-verbal patients.
Can you use a numeric pain scale on a cognitively impaired patient?
Not reliably once the impairment is moderate to severe, because a numeric self-report scale assumes the patient can understand the question and communicate a consistent answer. Mild impairment may still tolerate a simple scale with extra time and repetition. When self-report is inconsistent or absent, switch to a behaviour scale instead.
How often should pain be reassessed after giving analgesia?
Reassess within 30 to 60 minutes for oral or subcutaneous medication, and sooner, often 15 to 30 minutes, for IV administration, depending on the drug and route. The reassessment should repeat the same behaviour observation used to identify the pain, not just ask if the patient looks more settled.
Is agitation always a sign of pain in dementia patients?
No. Agitation can come from delirium, unmet needs such as hunger or a full bladder, environmental overstimulation, or pain. The behaviour scale helps separate these by looking specifically at facial expression, guarding, and vocalisation alongside the agitation, rather than treating restlessness alone as diagnostic.