Good Practice in Assessing Risk by Hazel Kemshall Bernadette Wilkinson

Author:Hazel Kemshall, Bernadette Wilkinson [Hazel Kemshall, Bernadette Wilkinson]
Language: eng
Format: epub
ISBN: 9781849050593
Goodreads: 9119286
Publisher: Jessica Kingsley Publishers
Published: 2011-01-15T00:00:00+00:00

Actuarial risk assessment

The term derives from the insurance industry. A small number of facts are combined according to a strict formula to estimate the level of risk. Actuarial assessment is therefore at the opposite end of the spectrum from the unstructured clinical method. There are fixed rules about the data to be used and how they must be interpreted.

The strengths are that it is systematic and eliminates variation between individual practitioners as well as excluding bias. Statistical evaluation shows it to be reliably better than chance in predicting violence.

The weakness is that statistical evaluations are done on populations whereas clinical risk management deals with individuals. Individuals cannot be treated as identical members of a group to which they may belong. So a group of a thousand people may be estimated accurately to have a 30 per cent chance of violence within a given period, but it would be quite wrong to assume any individual from that group has a 30 per cent chance of behaving violently.

The inflexibility of the actuarial method is also a problem because an individual may present a high risk of violence because of a single problem such as untreated psychosis. If there are no other generic risk factors, an actuarial assessment would be seriously misleading.

The Violence Risk Appraisal Guide (or VRAG, Quinsey et al. 1998) is a widely used actuarial scale whose authors have said it could and should supplant clinical estimation of risk (Quinsey et al. 1998, p.171).

The VRAG was developed on adult male patients followed up after release from a high-security hospital with violent re-offending as the outcome measure. Twelve items were found to predict re-offending. Total scores on the instrument are divided into nine ‘bins’, each with an attached probability of violence expressed in percentage terms and relating to a 7-year time frame.

The VRAG amounts to a detailed description of which factors were associated with violent re-offending in the population used to develop it. In that sense it is a history lesson. According to our first principle of risk assessment, it is a reasonable basis for prediction so long as it is used on patients similar to those used in the design. In population terms it fares as well as any other instrument in suitable settings.

For clinical use it has too many associated problems. One of its oddities is the negative effect on risk of a diagnosis of schizophrenia. The anomaly arises because the psychopathic patients in the original group presented even higher risks. The research literature would not support downgrading of risk because the diagnosis is schizophrenia.

Incidentally, none of these objections matters in large groups where the oddities cancel each other out. Actuarial measures such as the VRAG are useful and valid for commissioners, planners and researchers who wish to compare or evaluate services. The clinician should use them with caution and only as part of a comprehensive assessment of the individual.

Download

Copyright Disclaimer:
This site does not store any files on its server. We only index and link to content provided by other sites. Please contact the content providers to delete copyright contents if any and email us, we'll remove relevant links or contents immediately.