What is the 4AT?
The 4AT—often called the 4 A’s Test because its domains begin with “A”—is a short, structured assessment designed to identify possible delirium and cognitive impairment at the point of care. It was developed for busy general hospital environments where staff need a tool that is quick to administer, easy to remember, and usable without lengthy certification courses. The acronym highlights four components: Alertness, AMT4 (a four-item orientation screen), Attention (tested with months of the year backwards), and Acute change or fluctuating course of mental state.
Unlike some delirium instruments that yield only a yes/no label, the 4AT produces a graded score from 0 to 12. That scale supports three broad clinical messages: a low score suggesting delirium is unlikely (with the usual caveat that no screen is perfect), an intermediate band suggesting cognitive impairment that merits fuller evaluation, and a higher band that should trigger structured assessment for possible delirium. The score is intended as a screening aid; it does not replace history, examination, investigation, or specialist judgment.
Why screen for delirium?
Delirium is an acute disorder of attention and awareness, typically with fluctuating course, caused by medical illness, drugs, substance withdrawal, or multifactorial stressors in a vulnerable brain. It is common in older hospitalized adults, postoperative patients, and those with infection, dehydration, pain, immobility, or sensory deprivation. When missed, delirium associates with longer stays, higher complication rates, worse functional outcomes, and increased care needs. When recognized early, teams can search for precipitants, adjust medications, optimize hydration and nutrition, treat infection, mobilize patients, restore orientation with glasses and hearing aids, and involve families—interventions that map onto multicomponent delirium prevention and management bundles.
Operational barriers to detection include time pressure, overlap between delirium and dementia, hypoactive presentations that look like “sleepiness,” and tools that cannot be scored when patients are too drowsy to answer questions. The 4AT was designed partly to reduce those barriers by combining observation, brief cognitive tasks, and collateral history, and by allowing explicit “untestable” scoring on cognitive items so that very ill patients still receive a meaningful total score rather than an “unable to assess” dead end.
Structure of the assessment
Administration typically takes under two minutes in routine practice. The examiner observes the patient, asks a small number of orientation questions, performs one attention task, and synthesizes information about recent change or fluctuation from the patient, bedside staff, carers, or the record. Items are scored as specified in the official instrument; points from all four items are summed to produce the total. The maximum contribution from any single item reflects how strongly that feature discriminates in validation work—for example, markedly altered alertness and clear acute fluctuation each carry a larger weight than a single error on orientation.
Item 1: Alertness
This item captures the patient’s level of consciousness and arousal at the time of testing. The clinician observes whether the patient is fully alert, whether sleepiness is trivial and brief after waking, or whether alertness is clearly abnormal. Abnormal alertness includes being markedly drowsy (for example, difficult to rouse or obviously sleepy throughout the encounter) or agitated or hyperactive in a way that represents a disturbed level of arousal rather than normal anxiety alone.
Practical tips include observing the patient while introducing yourself, attempting gentle verbal and tactile stimulation if they appear asleep, and noting whether the patient can sustain attention to simple questions. Because hypoactive delirium can masquerade as fatigue, this item rewards careful observation rather than a single glance. In hospital cohorts, grossly altered alertness is a strong signal that should lower the threshold for full delirium evaluation even before cognitive scores are tallied.
Item 2: AMT4 (Abbreviated Mental Test — four items)
The AMT4 is a condensed orientation battery. The patient is asked for age, date of birth, place (usually the name of the hospital or building), and current year. It descends from the classic ten-item Abbreviated Mental Test but keeps only four questions to preserve speed. Scoring reflects the number of errors: no errors, one error, or two or more errors. If the patient cannot engage with the task because they are too drowsy, unwell, or otherwise untestable, the item is scored at the highest cognitive-impairment level for that domain, consistent with the instrument’s logic that patients who cannot be examined cognitively remain at risk.
Clinicians should ensure adequate hearing and vision, allow time for answers, and use interpreters when language barriers exist. A single mistake in an otherwise sharp older adult may reflect benign slip or stress; repeated errors or inability to attempt the task carries more weight. This item helps separate global disorientation from isolated attention failure tested in the next section.
Item 3: Attention (months of the year backwards)
Attention is central to delirium in DSM-informed definitions. The 4AT uses the months-backwards task: the patient is asked to recite the months in reverse order starting from December. The official instructions allow one initial prompt (for example, “What is the month before December?”) to help the patient start. Performance is scored by how many correct months are produced in sequence before failure.
A strong performance—reciting at least seven months correctly, reaching back through June—is scored as intact attention for this item. Weaker performance—starting but achieving fewer than seven months, or refusing to begin—is scored in an intermediate category. Patients who are too unwell, drowsy, or inattentive to start at all are scored in the highest impairment category for this item, again preserving sensitivity in patients who cannot complete formal psychometry.
This task is brief but demanding; it stresses working memory and sustained concentration. It can be affected by education and language, so results must be interpreted beside baseline function and Item 4 (acute change). Still, in general medical populations it has proven a practical proxy for inattention when embedded in the full 4AT context.
Item 4: Acute change or fluctuating course
This item operationalizes the temporal onset and variability that distinguish delirium from chronic cognitive disorders in many cases. The examiner asks whether, over roughly the last two weeks, there has been evidence of acute change or fluctuation in alertness, cognition (such as new confusion or disorientation), or other mental functions (such as perceptual disturbances or paranoid thoughts), and whether that issue is still relevant in the last twenty-four hours. Information may come from the patient, nursing documentation, therapists, family, or prior admissions.
Fluctuation—often described as the patient being “clear at times and muddled at others”—is a hallmark feature. This domain also embeds the spirit of very short screens such as the single informant question about whether the patient is more confused or drowsy than usual. When collateral history is positive, the item contributes substantially to the total score, reflecting how strongly temporal change supports a delirium hypothesis even when bedside cognitive testing is limited.
How the total score is used
After each item is scored according to the official rules, points are added. The resulting 0–12 total is interpreted in three bands that guide next steps rather than dictating diagnosis. A score of zero suggests that delirium or severe cognitive impairment is unlikely, though clinicians should remain alert if data are incomplete or the patient’s course changes. Scores from one through three suggest possible cognitive impairment and should prompt more detailed cognitive testing, informant history, and review of medications and comorbidities. A score of four or higher raises concern for possible delirium (often alongside underlying cognitive impairment) and should lead to a full delirium assessment, investigation for precipitants, and implementation of supportive and preventive measures appropriate to the setting.
These bands align the screen with common pathway logic: triage at admission, before and after high-risk procedures, at handoffs, and whenever staff notice altered mental status. The 4AT is generally used as an episodic instrument rather than for continuous high-frequency monitoring, to avoid patient burden and desensitization.
Integration with dementia and baseline cognition
The 4AT is applicable when pre-existing dementia is known or suspected, because delirium superimposed on dementia is common and serious. In such patients, orientation and attention items may already be abnormal at baseline, which can reduce specificity for new delirium. In that situation, Item 4 and careful comparison to the person’s usual function become especially important: delirium should be suspected when there is a clear acute deviation from baseline, new fluctuation, or new sleep–wake disruption even if chronic impairment is present.
Settings and practical caveats
The tool is intended for adults across wards, emergency departments, rehabilitation units, and many specialty services. In intensive care, particularly for mechanically ventilated patients, many guidelines prioritize ICU-specific instruments (such as CAM-ICU or ICDSC) because sedation, ventilation, and critical illness alter examination. Hearing impairment, delirium-mimicking aphasia, profound frailty, and language barriers all require thoughtful adaptation—optimizing sensory aids, using interpreters, and weighting collateral history.
Quality implementation depends on consistent scoring, clear documentation, and linking positive screens to a defined clinical response (who is notified, what investigations follow, how often to reassess). Hospitals often embed the 4AT in admission clerking, nursing observation charts, and perioperative checklists so that delirium risk is visible early in the stay.