INTRODUCTION

Clinical cognitive assessment involves the informal and formal testing of memory and other intellectual abilities in the context of the overall clinical assessment of the older person. It needs to be undertaken with most, if not all, people being assessed by an older persons’ mental health service (OPMHS). Clinical cognitive assessment is not the same thing as formal neuropsychological testing, which should be undertaken by specialised personnel, preferably by clinical neuropsychologists. Most people being assessed by an OPMHS will not require formal neuropsychological testing. However, clinical cognitive assessment is a basic skill that all health workers within an OPMHS team should master.

CONTEXT OF COGNITIVE ASSESSMENT

The context of the interview and the response of the older person to initial history taking will determine the priority given to cognitive testing. However, most, if not all, older people will need to undergo some type of cognitive assessment.

Experienced mental health workers learn to integrate history taking, mental state examination (MSE) and some components of the cognitive assessment during their time with the older person. As a consequence, there will usually be no clear demarcation between history taking and MSE. In addition, the initial components of cognitive testing involve observing the older person responding to questions during history taking.

Cognitive testing during home visits sometimes poses challenges for the mental health worker. With any individual person seen in the community, the approach to clinical cognitive testing is likely to be influenced by the purpose of the interview and the history obtained from the person and at least one informant. It might also be informed by what the worker already knows about the person. For instance, the type of cognitive examination undertaken annually on a well-known, case-managed person might be different from the type of cognitive examination undertaken during the initial assessment of a newly referred person. Thus, it is appropriate to have a flexible approach to clinical cognitive assessment with a test selection strategy that is modified according to circumstances.

Although clinical cognitive assessment can be undertaken by experienced mental health workers without the assistance of standardised rating scales, in most cases the worker will use standardised measures. The advantage of using such scales is that it is easier to communicate the findings with other health workers and to make longitudinal comparisons by using the same scales each time. It is also easier to interpret scores in the light of normative data for potential confounders, like education and age, if one has used a standardised test.

Cognitive testing never takes place in a vacuum. Both the person and the clinician come to testing with preconceived notions of what will take place and both are often surprised. The person’s level of cooperativeness and motivation is likely to be an important ingredient in the successful completion of valid and reliable cognitive testing. It is very difficult to test an uncooperative person, and in such circumstances the worker’s time is often better spent working on the therapeutic alliance to prepare the way for cognitive testing at a later date. Meanwhile, the worker can make general observations of the person’s mental state and behaviour that are likely to inform a broad judgment of the person’s possible cognitive state.

When assessing an older person’s performance on cognitive testing, it is important to consider the effort they put into it. Poorly motivated test taking generates falsely low scores and might lead to erroneous conclusions. Poor motivation might be due to the presence of a mental health problem such as depression or a contextual factor such as a compensation claim following a head injury. Sometimes, testing is conducted in less than optimal circumstances (e.g. in a very noisy environment), which does not allow the older person to do their best.

In Chapter 13, the distinction between the MSE and the Mini-Mental State Examination (MMSE) (Folstein et al 1975) was emphasised. This needs to be borne in mind by inexperienced mental health workers.

COMPONENTS OF CLINICAL COGNITIVE TESTING

When conducted under ideal circumstances with cooperative people, clinical cognitive testing should consist of the following components:

COMPONENTS OF MEMORY

The terminology for the various components of memory sometimes causes confusion. The ability to hold information in the mind for brief periods, either in the visuospatial sketchpad (imagine a jotting pad) or in the phonological loop (imagine a loop of audio tape), is now generally referred to as ‘working memory’. Working memory has been likened to RAM in a computer. Material in working memory disappears when conscious effort is turned off, just like material stored in RAM, which disappears when the power is turned off. Neuropsychologists previously referred to working memory as ‘short-term memory’. However, mental health workers often use the term ‘short-term memory’ to refer to ‘recent memory’ and the term ‘long-term memory’ to refer to ‘remote memory’.

Neuropsychologists consider both recent and remote memory to be types of long-term memory. They further divide long-term memory into episodic memory (memory for events in the person’s life) and semantic memory (memory for things that have been deliberately learned). Both of these are types of declarative memory (i.e. memory that the person knows about). There is another type of memory that is important in people with dementia: ‘procedural memory’. This is a form of non-declarative memory that involves subcortical brain circuits. Procedural memory encompasses the acquisition of motor skills or routines, including automatic behaviours like riding a bicycle. It is also critical for learning in people with dementias that affect cortical circuits (e.g. Alzheimer’s disease).

SCREENING INSTRUMENTS

It is important that screening is distinguished from diagnosis. Screening is designed to identify a subgroup of older people in which further diagnostic work-up is required. OPMHS often use a standardised approach to clinical cognitive assessment. Many services use long-established screening instruments such as the MMSE (Folstein et al 1975), although, in the United States, the Department of Veterans Affairs has supported the development of the Saint Louis University Mental Status (SLUMS) examination to replace the MMSE (Tariq et al 2006). In the United Kingdom, the cognitive section (CAMCOG–R) of the revised Cambridge Mental Disorders of the Elderly Examination (CAMDEX–R) (Roth et al 1999) has been used for many years. More recently, the revised version of the Addenbrooke’s Cognitive Examination (ACE–R) (Mioshi et al 2006) has been adopted by some centres. The ACE–R includes the MMSE and executive function tests. In Canada, the Montreal Cognitive Assessment scale (MoCA) (Nasreddine et al 2005) has been developed to screen for mild cognitive impairment. In Australia, access to subsidised cholinesterase medication (donepezil, rivastigmine and galantamine) and to subsidised memantine under the Pharmaceutical Benefits Scheme (PBS) generally requires testing the person on the MMSE or on a more extended battery, the cognitive subscale of the Alzheimer’s Disease Assessment Scale (ADAS–Cog) (Mohs et al 1983).

Regardless of which screening instrument is used, the work of diagnosis requires the knowledgeable integration of information from history taking, general MSE, cognitive examination, physical examination, laboratory tests and neuroimaging studies. In some older people, detailed neuropsychological assessment will also be required.

MINI-MENTAL STATE EXAMINATION

When the Mini-Mental State Examination (MMSE) was first introduced in 1975, it represented a significant advance because it allowed doctors and other clinicians without formal training in neuropsychology to administer a brief screening test to identify people who might be suffering from dementia. Although the MMSE is in widespread use around the world, and has been translated into many languages, it has some major deficiencies that mean that it should not be used as the sole clinical cognitive assessment tool in OPMHS. It suffers from so-called ‘ceiling’ and ‘floor’ effects, and is difficult to administer via an interpreter. Ceiling effects mean that some people score maximally despite having obvious cognitive impairment, whereas floor effects mean that some people, particularly those living in residential aged care facilities (RACFs), score zero, making it difficult to track their progress with this instrument.

MMSE items do not cover frontal executive function very well, whereas assessment of this is often essential in people presenting to an OPMHS, as frontal executive dysfunction often accompanies changes in mood, personality and behaviour. It is also commonly associated with impaired insight and poor judgment. Thus, if the MMSE is to be used as a routine cognitive screening test, it needs to be supplemented with tests of frontal executive function. In well-educated people, and those with a history of high occupational function, more difficult memory testing is prudent, as such individuals sometimes score maximally on the MMSE despite undoubted cognitive impairment.

Because the MMSE is so commonly used, it is worthwhile considering its items in more detail here. However, it is worth emphasising that the individual MMSE items do not measure single cognitive domains. The MMSE begins with five items that test orientation to time and five items that test orientation to place. Following the original publication of the MMSE in 1975, there has developed considerable variation in the precise wording of these items and in the scoring approach used. The nature of the items testing orientation to place is such that people examined at home are likely to score higher than those examined in less familiar environments, such as hospitals and clinics. This variability reduces the generalisability of scores obtained in different environments. The next item involves the registration of three words for later recall. The original three words, ‘apple, table, penny’, have now been over-learned by some people, including some with dementia, such that they may no longer be a valid test of new learning ability. The Standardised MMSE (SMMSE) (Molloy et al 1991) contains different words as well as alternative sets of equivalent words to allow valid retesting.

The MMSE then tests attention and calculation through the serial subtraction of 7 from 100 (usually referred to as ‘serial sevens’) or attention through spelling the word ‘world’ backwards. Confusingly, different versions of the MMSE impose different rules for interpreting the spelling of ‘world’ backwards. The MMSE then requires the person to recall the three words that they were asked to register before attempting the serial sevens and ‘world’ backwards items. The person is then asked to do a series of five language tasks. The first task is to name two objects, a watch and a pencil. The second task is to repeat a phrase: ‘No ifs, ands or buts.’ The third task is to follow a written instruction: ‘Close your eyes.’ The fourth task is to follow a three-part instruction: ‘Take this piece of paper in your right hand, fold it in two with both hands and then place it on the floor.’ In the fifth language task, the person is asked to write a sentence. Finally, to test constructional praxis, the person is asked to copy a diagram of intersecting pentagons.

When using the MMSE to monitor progress over time, it is important that a consistent method of administration is used and that consistent scoring rules are applied. For instance, one needs to decide whether the person will be scored as correct if they are only one or two days out in stating the date. The SMMSE and the ACE–R both have explicit scoring rules for the MMSE and for this reason might be preferred to the original MMSE by some teams. The other important issue is the use of alternative memory tasks. Although it is common practice for people to be given the same memory task (e.g. ‘apple, table, penny’) each time they are tested on the MMSE, this does not make good sense from a neuropsychological perspective unless the interval between testing occasions is at least 6 months, and preferably longer.

In older people, particularly those with some degree of cognitive impairment, performance on the MMSE often varies by time of day. Many older people do better on this test in the morning. As a consequence, it is important to standardise the time of day that testing takes place, if one is to make valid comparisons between testing occasions. Alternatively, it would be important not to over-interpret small changes in the MMSE score when testing has been undertaken at differing times of day.

When the MMSE was originally published in 1975, scores of 23 or less were found to be predictive of dementia in general medical inpatients in the United States. However, educational achievement has improved since then and scores higher than 23 are now commonly found in people with dementia. Under the Australian PBS, subsidised cholinesterase inhibitor therapy for mild to moderate Alzheimer’s disease is available to people with MMSE scores of 10–24 inclusive. Normative data for the MMSE have been published for many countries, including Australia and New Zealand. Normal community-residing older people score 26 or over, even in advanced old age.