Morris et al (2000) describe the varieties of current approaches to more detailed neuropsychological assessment in older people. How does this fit in with the reality of daily work for psychiatrists in general and old age psychiatrists in particular?

For some, perhaps many, simply the idea of having access to clinical psychology services may seem like an aspirational fantasy: the chance of even getting a referral seen would be a step forward. However, it is reasonable rather than aspirational to expect that any half-decent mental health service for older adults should certainly have an establishment for, and usually have recruited to, clinical psychology posts. They are of course unlikely to be plentiful and therefore need to be used wisely. How do you decide what to ask for and when?

Approaches to the assessment of cognitive impairment vary from observation (I once heard a junior doctor presenting a case in geriatric medicine include the observation that the patient “looked demented”) to highly detailed neuropsychological examination with multiple forms of brain imaging.

I can recall the early days of computed tomography (CT) scanning, which was just beginning when I was a trainee clinical psychologist. Just for a brief while it seemed as though all that I had learned as an undergraduate psychologist would become redundant as the new imaging revealed all. Of course this did not happen then and has not happened yet. As Morris et al point out, neuropsychological assessment is about function rather than structure. However, the successive introduction of increasingly sensitive brain imaging techniques does give a useful model for how we might approach developing and using clinical psychology assessment skills.

Initially CT was rare, expensive and had to be justified at length for each patient, especially older ones. It is now seen as part of the standard work up for Alzheimer's disease. Perhaps functional magnetic resonance imaging (fMRI) will be like this in the future. For the moment you would ask for, say, a positron emission tomography (PET) or single photon emission computed tomography (SPECT) scan only if you thought it would add significantly to the picture.

The Mini Mental State Examination (MMSE) or something similar is routine. It works quite well for most cases, although it should be supplemented by systematic evaluation of activities of daily living, including those based on carer information and carer well-being. Requests for detailed neuropsychological assessment are for those cases where the picture does not quite add up. Typically, these include unusual presentations or cases where the screening assessment is at odds with other information, such as carers' or relatives' reports. One common example is when the patient has a premorbid IQ outside the normal range. More often the uncertainties arise for those with an IQ well above average. There are data that quite clearly show how the MMSE, for example, correlates with IQ (Reference Christensen and JormChristensen & Jorm, 1992). I well remember a retired professional who presented at a memory clinic with a MMSE score of 29/30 but with relatives' reports suggesting cognitive problems. The Object Learning Test of the Kendrick Battery has no correlation with IQ (Reference Kendrick and WattsKendrick & Watts, 1999) and it, like the relatives' reports, more accurately predicted the subsequent clinical course.