guide to the 4at: INTRODUCTION

The 4AT has 4 items, each of which begins with an ‘A’:

1. Alertness

2. AMT4: Abbreviated Mental Test - 4.

3. Attention: Months of the Year Backwards

4. Acute change or fluctuating course

Note that the 4AT can be scored in patients too drowsy to engage in testing or conversation. No patients are Unable to Assess (UTA) with the 4AT: see below.

The 4AT does not require special training. Yet some knowledge of delirium is essential when using the 4AT or other delirium tools: see here for a brief overview.

4AT case examples: 6 cases to demonstrate how the 4AT is scored in different situations.

OVerview of scoring

The 4AT is scored from 0-12.

A score of 4 or more suggests delirium but is not diagnostic. In every case the diagnosis is reached by clinical judgement.

A score of 1-3 suggests cognitive impairment and more detailed cognitive testing and informant history-taking may be indicated.

A score of 0 does not definitively exclude delirium or cognitive impairment: more detailed testing may be needed depending on the clinical context.

The tester should take account of communication difficulties (hearing impairment, dysphasia, lack of common language) when carrying out the test and interpreting the score.

Note: 4AT scoring of patients unable to engage in conversation

Many patients with delirium are unable to produce meaningful speech because of drowsiness or severe inattention.

The design of the 4AT allows these patients to have a score on the test. No patients with delirium are ‘Unable to Assess’ (UTA) with the 4AT.

How does this work? If the patient cannot engage with the tester and attempt the AMT4 or the Attention test, then they are rated ‘untestable’ and given a score for this. Untestable status on both of these items yields a score of 4, which suggests possible delirium.

So if a patient is unable to speak because of drowsiness the tester does not record that the patient is unable to be assessed (or UTA) on the 4AT. Instead, we encourage that the patient is scored as having abnormal alertness (Item 1) and also an untestable result on the cognitive testing items (Items 2 & 3). This will allow a 4AT score to be given, rather than the delirium assessment not being completed and potentially no diagnosis being made.

scoring 4AT ITEMS 

Guide to scoring Item 1: altered level of alertness is >95% likely to be delirium in general hospital settings. If the patient shows significant altered alertness during the bedside assessment, score 4 for this item.

Guide to scoring Item 2: the Abbreviated Mental Test 4 or AMT4 is a brief test of orientation in which the patient is asked: age, date of birth, place (name of the hospital or building), and the current year. 1 mistake scores 1 point on the item, and 2 or more mistakes scores 2 points.

If the patient cannot provide meaningful answers because of altered arousal, inability to produce speech, etc., then the patient is given a score of 2 (given for patients who are ‘untestable’ on simple cognitive tests).

Guide to scoring Item 3: Months of the Year Backwards is a simple, widely-used test of attention which is sensitive to both delirium and general cognitive impairment. The patient is asked to recite the months of the year in backwards order from December.

If the patient verbally declines to start the test or is not able to correctly recite to June, score 1. If the patient cannot start the test for example through being drowsy or too inattentive they are in the ‘untestable’ category for this item and receive a score of 2.

Guide to scoring Item 4: rapid (hours, days) deterioration in mental functioning is highly specific to delirium. If there is evidence of change or fluctuation then this item scores 4. This gives an overall 4AT score of at least 4, indicating likely delirium.

Item 4 requires information from one or more source(s), eg. your own knowledge of the patient, other staff who know the patient (eg. ward nurses), GP letter, case notes, or carers.

As part of the process of determining change from baseline in non-cognitive areas it can be helpful to elicit any hallucinations and/or paranoid thoughts by asking the questions such as, “Are you concerned about anything going on here?”; “Do you feel frightened by anything or anyone?”; “Have you been seeing or hearing anything unusual?”

Fluctuation can occur without delirium in some cases of dementia, but marked fluctuation usually indicates delirium.

Note that Item 4 incorporates the ‘Single Question in Delirium’ or SQiD.

What if there is no informant history, or the informant history may not be reliable?

No informant history: it is common for there to be no carer available to give information on acute change or fluctuation. Some studies show that at the front door 25% of patients may have no informant. Often an informant can be contacted later, but this can delay a potential diagnosis if the assessment process requires an informant at the time of assessment.

Unreliable informant history: not all informants are able to provide a reliable history of change. For example, some relatives have very limited contact with patients, or some relatives may have cognitive impairment themselves. Therefore, all informant history needs to be used in the diagnostic process in the light of all sources of information and clinical judgement.

Clinical course strongly suggests mental status change: it can sometimes be determined by the practitioner that the observed mental status deterioration must be acute. For example, if the practitioner is seeing a patient who is drowsy but is known to be living independently before hospital admission, it is obvious that this is an acute change. In this situation, the practitioner should score Item 4 as a 4.

In summary, it is clear that in some situations in which there is no informant or no reliable informant it is clinically very likely that a patient has delirium. If a tool requires an informant history to give a final score, this creates a patient safety problem because requiring that an informant history is always required to make this diagnosis can lead to delayed diagnosis and treatment, a serious problem given that delirium is a medical emergency.

The 4AT was purposely designed to allow for these real-world clinical situations, in that a score of 4/12 or more can be reached in several ways, even with no informant history:

• Item 1 scoring 4

• Items 2 & 3 both scoring 2

• Item 4 scoring 4 using evidence from clinical history

This design feature of the 4AT reflects clinical practice, because practitioners do sometimes make a provisional diagnosis of delirium based on bedside features alone if no informant history is available. This reflects safe care: while informant history should always be sought, a diagnosis of delirium should not be delayed if it is difficult or impossible to get this history immediately.

4AT SCORING IN PATIENTS WITH SEVERE DEMENTIA

The 4AT is designed to balance sensitivity and specificity for delirium. It is also designed to yield a score even in situations where the patient is unable to speak or communicate meaningfully. Patients with severe dementia may have a level of impairment that means that they score 2 on both Items 2 and Item 3, giving a score of 4. This means specificity of the 4AT for delirium in patients with severe dementia is less high than the overall figure. See this review for a meta-analysis of 4AT studies which includes discussion relating to the relevant papers.

Note that deciding if a person has severe dementia it is known to be challenging if there is superimposed delirium. This is because the cognitive deficits in severe dementia are often indistinguishable from those seen in delirium; additionally many people with severe dementia are unable to produce meaningful speech and they may also show altered level of arousal. In such cases the diagnosis relies particularly on informant history and/or observation of the person over time. The diagnosis, as with all delirium, is clinical. No delirium assessment test is sufficient on its own to diagnose delirium.

use of the 4AT in practice

The 4AT is used in the following situations:

• ED or other acute settings as a screening tool for delirium in older people

• At home or in care homes when delirium is a concern

• Transitions of care

• Pre-op delirium & cognitive assessment

• Post-op, repeated daily for an appropriate period (usually 3-7 days)

• Repeated daily during an episode of delirium to assess for recovery

• At any time when delirium is suspected (e.g. with family concern, or if clinical observation is suggestive)

Multiple times per day monitoring for new-onset delirium in inpatients can be performed with the 4AT for limited periods. Some centres have modified the 4AT to use Days of the Week Backwards rather than Months Backwards to reduce test time and burden on patients and staff. Longer-term monitoring is best done with observational tools.

The reason for using observational tools for monitoring is that asking patients to do repeated cognitive tests several times per day for periods of several days or more is generally too burdensome for both patients and staff. Compliance with tools that require this is usually poor. Even use of ‘ultra-brief’ cognitive tests multiple times per day is clinically inappropriate, not just because of patient and staff burden but because of practice effects.

In inpatients without delirium or in patients or residents in longer-term facilities, ongoing monitoring through observational tools is more appropriate than using cognitive tests multiple times per day. Several options for observational monitoring tools are available.

SQiD / NEWS2 as a monitoring tool with the 4AT as the follow-on delirium assessment tool is the recommended process in the UK National Health Service. See here for more details.