This pack consists of a book including the questions in the Cambridge Examination for Mental Disorders of the Elderly; a computer disk onto which answers can be entered and from which questionnaires can be printed; and a smaller book with pictorial materials for cognitive examination. Within the main book there is a structured clinical interview; a brief neuropsychological battery; a structured interview with a relative; the diagnostic criteria from DSM-IV and ICD-10 for dementia and other categories including differential with depression. The CAMDEX-R also gives operational criteria which it suggests are used for clinical diagnosis and guidelines for classifying dementia according to clinical severity.

The first aim is to enable a differential diagnosis of dementia to be made according to the most recent criteria with the materials needed (apart from for physical examination and biochemical examination) included. The book gives the range of information required for differential diagnosis of the varying forms of dementia available in a single standardised interview and examination pack. However, I found it surprising that the criteria for Alzheimer's disease and vascular dementia are not given, although I agree they are fairly well known, but you could argue that about the rest of the material as well. Most mental health professionals know how to elicit the history or mental state.

The pack is designed so that different mental health professionals can use it. However, a physical examination and blood tests are needed to fill in the checklists. As a result only medically trained professionals can use the pack to make a differential diagnosis.

The computer pack has no installation instructions in the handbook. Once installed I was pleased to see that it was year 2000 compliant, but it would accept ridiculous dates for the year the interview was done, for example, 1957. The package is not as professionally laid out as the handbook and is DOS based. I was disappointed that the diagnosis had to be entered into the computer package by the interviewer, as I was hoping that the diagnostic criteria would be matched up with the answers to give an indication as to how they were fulfilled even if the programme did not come to a diagnosis. The GMS—AGECAT (a similar package designed by Professor Copeland and his team in Liverpool) comes to a standardised diagnosis from the information given with which the interviewers are free to disagree clinically. It would be helpful if CAMDEX—R did this as well. The authors state that they are currently developing a computer programme for examining individual scores obtained versus expected scores on both the total and the sub-scales based on demographic characteristics.

In summary, the CAMDEX—R is a well-organised and generally comprehensive research instrument for the differential diagnosis of dementia. The materials are beautifully laid out and a pleasure to handle. It would be a helpful research tool in providing standardised assessments. The computer package is, however, disappointing.