Abbreviated mental test score

The Abbreviated Mental Test Score (AMTS) is a 10-point test for rapidly assessing elderly patients for the possibility of dementia. It is recommended as the go-to screening tool in emergency and hospital settings for patients over 65. It was first used in 1972, and is now sometimes also used to assess for mental confusion (including delirium) and other cognitive impairments. It takes approximately 3–4 minutes to administer and requires no specialist training or licensing.

The AMTS was developed to address the need for a rapid, practical method of assessing cognitive impairment in geriatric patients. It was adapted by Hodkinson (1972) from the 26-item Blessed Dementia Scale (BDS) by removing 16 items: 13 for repetitiveness, 2 for being too easy, 1 for being too difficult. Validation studies revealed a near-linear correlation between AMTS and BDS scores, indicating strong convergent validity and accuracy of measurement.

The standard AMTS consists of 10 questions that assess orientation, memory, and attention. The following questions are put to the patient. Each question correctly answered scores one point. A score of 7–8 or less suggests cognitive impairment at the time of testing, although further and more formal tests are necessary to confirm a diagnosis of dementia, delirium or other causes of cognitive impairment. Culturally-specific questions may vary based on region.

The AMTS has been criticised for containing culturally and temporally outdated questions. For example, the World War I question was gradually revised to ask for the start of World War II as fewer elderly patients had direct experience of the earlier conflict post-1970s. However, even World War II is now beyond the lived experience of many older adults. As a result, these questions no longer assess time-orientation but semantic memory. Many patients struggle to answer correctly, not due to cognitive impairment but because of limited personal relevance, reducing the diagnostic accuracy of the test. Experts have suggested that recalling distant historical dates is an unreliable measure of cognitive impairment as answers are often confounded by retroactive interference from recent memories. It has been recommended that these items be replaced with questions referencing recent and culturally relevant events to improve the test’s validity.

As no formal administration training is required, many clinicians administer and score the AMTS incorrectly. Score cut-off thresholds for cognitive impairment vary widely from 6–10, undermining the test's diagnostic reliability.

The AMTS poorly distinguishes between dementia and delirium, and lacks sensitivity to detect mild cognitive impairment, making it a poor tool for differentiation and early-stage diagnostics.

Shorter Versions

The AMT4 uses 4 items from the AMTS: (i) What is your age? (ii) What is your date of birth? (iii) What is the name of this place? (iv) What is the year? A cut off score of 3/4 performs comparably to an AMTS cut-off score of 8/9. The AMT4 is part of the 4AT scale for delirium.

AMT5: Includes 5 items - address recall, recognition of two people, date of birth, current monarch, and counting backwards from 20. Despite cut-off score being adjusted to 4, it is still highly prone to false-positives.

AMT7: Includes 7 items - current time, address recall, recognition of two people, date of birth, current monarch, and counting backwards from 20. When the cut-off score was adjusted to 5, the same sensitivity and specificity levels were observed as in the original AMTS, making it the most reliable short-form version without compromising diagnostic accuracy.

The original AMTS has limited cultural validity as it relies on UK-specific knowledge such as naming the current Monarch. This limits its generalisability to non-UK settings requiring countries to adapt questions to avoid misdiagnosis. It also has limited construct validity as items like the start year of WWI no longer test time-orientation but factual crystal intelligence, requiring questions to be adapted to more recent events.

Despite these limitations, the AMTS demonstrates strong convergent validity between different diagnostic tools and versions of the AMTS. High test-retest reliability makes the AMTS more applicable by producing consistent results over time, allowing clinicians to reliably track changes in cognitive function.

Comparison to Other Tools

The AMTS has been shown to outperform tests like the Digit Span Backwards Test, Time and Change Test, IQCODE, SPMSQ, and the frequently used MMSE in general hospital settings by exhibiting high diagnostic accuracy, ease of use, and brevity. AMTS show strong convergent validity with MMSE scores while taking 3–4 minutes to administer rather than 10–15. Unlike many other cognitive tests, the AMTS is not significantly influenced by the patient's education level, making it suitable for diagnosing dementia in individuals with limited literacy.

The AMTS is less effective at detecting mild cognitive impairment, missing over half of cases compared to the MoCA which provides a more comprehensive cognitive profile. AMTS's narrow focus on memory and orientation leads to a ceiling effect, reducing its usefulness for early cognitive impairment detection.

International Adaptations

The AMTS demonstrates strong reliability and validity across different languages and cultures, with most versions using similar cut-off scores to identify cognitive impairment.

The AMTS was adapted by replacing the World War I question with the Mid-Autumn Festival and the Monarch’s name with the current Chinese leader.

The adapted version had a cut-off score of 7, with high sensitivity (92%) and specificity (87%), indicating strong validity. It also demonstrated high internal consistency and test-retest reliability, making it a highly effective cognitive screening tool.

The AMTS was adapted by replacing the World War I question with the Iraqi-Iranian War and the Monarch’s name with the Iranian leader.

The Persian version was confirmed to have high statistical validity, specificity, and sensitivity. It proved more applicable than the MMSE in Iran due to its brevity and lack of licensing requirement, making it suitable for the over-crowded and under-funded hospital settings. Additionally, unlike other cognitive tests, results are not affected by education or literacy levels which is critical for testing in Iran where many elderly adults lack formal education.

The AMTS was adapted by replacing the World War I question with World War II and the Monarch’s name with the Polish President.

The Polish version showed a strong correlation with the original AMTS, with no significant differences in sensitivity or specificity, confirming its effectiveness as a culturally adapted screening tool.

The AMTS was adapted by replacing the World War I question with the date of the Great Sorrow, the Monarch’s name with the Thai King, and the address recall task with asking the patient’s current address. The last change aimed to reflect cultural norms, as most rural elderly individuals are unfamiliar with the task of memorising arbitrary information like made-up addresses. This alteration has been criticised for shifting the task from testing short-term memory to semantic memory.

Despite these adaptations, the Thai AMTS still shows high rates of false-negative diagnoses. Many older adults were unable to provide their birth date or recall the current year due to Thailand’s mixed lunar/solar calendar system. These issues highlight the cultural limitations of the AMTS and the need for further adaptation in Thailand.

• General Practitioner Assessment Of Cognition – a brief screening tool for cognitive impairment designed for primary care
• GERRI
• Mini-mental state examination