Wednesday June 10 2009
Doctors have created a “simple test that can spot Alzheimer’s in five minutes”, according to the Daily Mail. It reported that a new five-minute test doubles the chances of detecting early dementia and is able to detect 93% of Alzheimer’s cases.
This carefully conducted research assessed the potential of a new, brief, self-administered test for dementia by trialling it on people who had clinically confirmed Alzheimer’s, other types of dementia and a large number of people who did not have dementia. This new ‘Test Your Memory’ (TYM) assessment was found to be more accurate than the commonly used testing method, the mini-mental state examination (MMSE). The TYM detected 93% of people who had Alzheimer’s disease compared to the 53% detected by the MMSE.
The test did falsely identify a number of people who did not have Alzheimer’s and, like other cognitive tests, does not remove the need for full clinical assessment and other diagnostic tests for Alzheimer’s disease. Nevertheless, there is no test currently available that is quick, can be performed by non-specialists and is sensitive to Alzheimer’s disease. This new test may potentially fill this gap. It will now require further trials in different population groups and settings before it can be widely used.
Where did the story come from?
Jeremy Brown and colleagues of the Department of Neurology at Addenbrooke’s Hospital, Cambridge, carried out this research. The authors received funding from the Alzheimer’s Research Trust (UK), the Cambridge Commonwealth Trust and the Stroke Association. The study was published in the peer-reviewed British Medical Journal.
What kind of scientific study was this?
This was a cross-sectional study designed to evaluate the ‘Test Your Memory’ (TYM) test as a potential method for identifying Alzheimer’s disease.
Although there are several tests of memory and cognition available, the authors of the study reported that none of them meets three crucial requirements for widespread use by non-specialists, namely taking minimal time to administer, testing a reasonable range of cognitive functions and being able to detect mild Alzheimer’s disease. It was hoped that the TYM could fulfil all these requirements.
The TYM is completed by the patient themselves and involves 10 tasks with different scores for each. These are: orientation (10 points), ability to copy a sentence (two points), “semantic knowledge” on long-established knowledge about facts, objects and meanings of words (three points), calculation (four points), verbal fluency (four points), similarities (four points), naming (five points), two tasks of visuospatial abilities (total seven points) and recall of a copied sentence (six points). The ability to perform the test was assigned a score of up to five points. In total, those taking the test could score up to 50 points, with a higher score indicating better memory and cognition.
The TYM was given to a control group of 540 people aged 18 to 95 who did not have dementia, 31 patients with non-Alzheimer’s forms of dementia and 108 people with amnesic mild cognitive impairment or Alzheimer’s disease. Mild cognitive impairment (MCI) is likely to progress to Alzheimer’s in individuals with certain cognitive examination scores.
Those with MCI were tested and considered to have Alzheimer’s disease or early Alzheimer’s if they scored less than 94 on cognitive tests, while the rest were considered to have MCI that was unlikely to progress.
Diagnoses of dementia were made by a neurologist, who administered the mini-mental state examination (MMSE) and the revised Addenbrooke’s cognitive examination (which includes the MMSE) and assessed the results of brain imaging and blood tests. Many of the patients also had psychiatric and neuropsychological assessment. Diagnoses of Alzheimer’s were made using established criteria, while other published criteria were used to diagnose amnesic MCI.
The control group of people without dementia was recruited by selecting relatives of people attending the neurology clinic where the study was being conducted and also from relatives of people attending neurology and medical outpatient departments at two other hospitals. Three age-matched controls were selected from this larger group for each person with Alzheimer’s.
The researchers compared TYM scores for people with and without Alzheimer’s disease with scores obtained using the MMSE and the revised Addenbrooke’s cognitive examination, to see if there were any differences.
To look at how scores might vary depending on who marked the test, they compared how a sample of 100 tests scored when the test was independently marked by three different people: a consultant experienced in the diagnosis of degenerative dementia, a neurology specialist registrar working in the memory clinic and a registered general nurse who received 10 minutes of tuition on marking the test, but who had no specialised experience of patients with dementia.
The researchers then assessed the sensitivity and specificity of the TYM test:
- Sensitivity is the ability to accurately detect that someone does have Alzheimer’s disease, and
- Specificity is the ability to accurately detect that someone does not have Alzheimer’s.
The researchers tested these factors by applying a range different cutoff scores on the TYM results to see how well these cutoff scores could distinguish between those with Alzheimer’s disease or MCI that was likely to progress to Alzheimer's (total 92 patients) and those without the disease (282 randomly selected age-matched controls), based on diagnosis by the neurologist.
This method was used to identify which cutoff score gave the best balance of sensitivity and specificity. For the TYM, the optimal cutoff score was calculated to be 42 or less.
The researchers also looked at how the sensitivity and specificity of the TYM compared to that of the commonly used MMSE test. They did this by calculating the percentage of patients with Alzheimer’s disease who were detected using the cutoff score of 42 or less for the TYM and a score of 23 or less for the MMSE (the accepted cutoff for dementia with this test).
The value of a test result in ruling in or out a disease (its positive or negative predictive value) is affected by how common the disease is in the group tested (its prevalence). The researchers therefore looked at what the positive or negative predictive value of the TYM would be in populations with varying prevalence of Alzheimer’s disease.
What were the results of the study?
The average score on the TYM among control participants without dementia was 47/50. Patients with Alzheimer’s disease scored an average of 33/50. Patients with non-Alzheimer’s dementias (for example, Parkinson’s or frontotemporal dementia) scored an average of 39/50. Control participants took on average five minutes to complete the TYM test. Scores were very similar regardless of who marked the tests (an experienced consultant, neurology specialist registrar or registered general nurse).
Results from the TYM test were shown to have good correlation with the widely used MMSE and Addenbrooke’s cognitive examinations. This means that people who scored well on the TYM tended to also score well on the other scales.
The researchers found that applying a cutoff score of 42/50 or less to indicate a diagnosis of Alzheimer’s gave a sensitivity of 93%, meaning that it correctly positively identified 93% of cases. It also produced a slightly lower specificity of 86%, meaning that it correctly confirmed that 86% of those without Alzheimer’s did not have the condition. The sensitivity of 93% means that the TYM was more sensitive in detecting Alzheimer’s than the commonly used MMSE test, which has sensitivity of 52%.
Using a cutoff score of 42 or less on the TYM in a group of people where 10% had Alzheimer’s gave a negative predictive value of 99% and a positive predictive value of 42%.
What interpretations did the researchers draw from these results?
The researchers concluded that the TYM can be completed quickly and accurately by normal controls without dementia. They say that it is a powerful and valid screening test for the detection of Alzheimer’s disease.
What does the NHS Knowledge Service make of this study?
This carefully conducted research assessed the validity of a new, brief, self-administered test for dementia by trialling it in people who had clinically confirmed Alzheimer’s or other types of dementia as well as a large number of people who did not have dementia. Compared to the widely used mini-mental state examination (MMSE), the TYM test had higher accuracy, detecting 93% of people who had Alzheimer’s disease. A few points to note are:
- A negative result on the TYM would mean that you could be fairly confident that the person tested did not have Alzheimer’s due to the test’s high sensitivity for detecting cases of Alzheimer’s disease. However, if the test was positive, you could be less confident that the person definitely had Alzheimer’s as the TYM had lower specificity. In other words, it had a reasonable chance of indicating that a person had dementia when they did not.
- The TYM test would not negate the need for full specialist assessment with examination, clinical tests and brain imaging. Alzheimer’s disease is a diagnosis of exclusion, which means it can only be confirmed when other causes of dementia have been excluded.
- This is a self-administered test, which means that people with more severe, advanced dementia may not be able to complete the test. Other factors, such as educational level, reading ability, first language and comprehension, are likely to influence an individual’s ability to complete the test. These factors were not tested in the study and further research will be needed to assess their influence.
- The test will need further national testing in other population groups and settings before it can be considered for wider use in clinical practice.
Currently, there is a need for a cognitive test that can be used quickly and administered by non-professionals, that examines various skills and is sensitive at detecting Alzheimer’s disease. This new test may have the potential to fulfill this need.