Abstract
Studies show that counsellors in primary care see many clients with difficulties of a severity similar to those found in secondary care services. Evidence from both RCTs and studies of routinely collected data indicates that many of these ‘clinical’ clients benefit from brief counselling intervention. However, little is known about why some benefit while others fail to do so despite completing their counselling contract. This paper considers client characteristics recorded at assessment and aims, using logistic regression analysis, to identify those characteristics predictive of a poor outcome. Results indicate that a number of characteristics are associated with poor outcome; the most important predictors are economic inactivity and aspects of the patient's condition, particularly continuous/recurrent depression, with some differences between genders. However, the models produced were not acceptable in their predictive power. This may be partly due to data quality issues or important characteristics not being available in the data. The paper concludes that being unemployed or on state benefits may be a proxy measure of severity that has an important impact on outcome for all patients, and particularly for males. Some reasons are suggested and areas of future research are identified.