Evaluate impulse-control treatments and Grant et al. (2008)

Grant et al. (2008): study what predicts response, not a universal prescription

The 2028–2030 Cambridge reference list identifies Grant, Kim, Hollander and Potenza (2008), ‘Predicting response to opiate antagonists and placebo in the treatment of pathological gambling’, as the Clinical 1.3 key study. The researchers analysed 284 adults with DSM-IV pathological gambling who had taken part in one of two double-blind, placebo-controlled trials: a 16-week nalmefene trial or an 18-week naltrexone trial.

The paper asked which clinical characteristics were associated with treatment response. Response was defined as a reduction of at least 35% on the gambling-adapted Yale–Brown scale that lasted at least one month by the trial endpoint. A positive family history of alcoholism was the strongest predictor of response among people receiving an opioid antagonist. Gambling-urge scores showed a trend-level association with response at higher doses; younger age showed a trend-level association with placebo response.

This was a pooled predictor analysis of existing trial groups, not a new experiment that randomly assigned all 284 people in one study. The result suggests that responses differed between individuals; it does not prove the family history caused response, identify a treatment for an individual, or establish effectiveness for everyone with gambling disorder. The two component trials used different drugs and durations, and the placebo group was smaller than the active-medication group. Several authors disclosed pharmaceutical research grants or consultancy work, which readers can consider when evaluating the report.

The source uses the earlier DSM-IV term ‘pathological gambling’; current ICD-11 classification uses gambling disorder. Naltrexone and nalmefene are medicines requiring qualified clinical assessment. This lesson gives no dose, recommendation or substitute for professional care.

Covert sensitisation: Glover’s single-case example

Covert sensitisation is a behavioural technique in which a person imagines an unwanted behaviour together with an unpleasant imagined consequence, with the aim of making the behaviour less appealing. It is a treatment example from the syllabus, not a self-help exercise for students to carry out.

Glover (1985) described one adult with a 14-year history of daily shoplifting. The report says that, after covert sensitisation chosen with the client, stealing had stopped apart from one lapse at a 19-month follow-up. The report also described improvements in self-esteem and social relationships.

A long follow-up in one individual gives detailed, idiographic evidence about that case. Without a comparison group, independent replication or an objective measure of all behaviour, it cannot show that covert sensitisation alone caused the change or that the outcome will generalise to other people. A case report can support a hypothesis and guide questions for further study, not estimate an average treatment effect.

Imaginal desensitisation: relaxation and imagined choice

Blaszczynski and Nower (2003) describe imaginal desensitisation as a relaxation-based technique for impulse-control problems. In broad terms, a clinician teaches relaxation and a client imagines encountering a relevant cue or urge while remaining relaxed and choosing not to act. The article explains the approach and discusses reported clinical outcomes; it should not be mistaken for a large, independent, randomised trial of this method.

This differs from covert sensitisation: imaginal desensitisation pairs the imagined cue with relaxation and a non-acting response, while covert sensitisation pairs an imagined behaviour with an aversive consequence. Both use imagery, but their proposed learning processes differ.

The 2003 publication is an example for learning the technique, not a basis for learners to practise therapy on themselves or someone else. Claims about clinical benefit require appropriately controlled, replicated research and should be interpreted in light of samples, outcome measures and follow-up. Only qualified practitioners can assess whether a treatment is suitable.

Compare the evidence and state what it cannot decide

Grant et al. use standardised clinical measures and a placebo comparison inherited from the two trials; the pooled sample supports analysis of treatment-response predictors. It does not make the predictor associations causal, and the report cannot settle which treatment is right for a particular person. Blinding can reduce expectancy differences, but placebo response and clinical context still matter.

Glover’s case provides detailed individual history and a long follow-up, but no comparison group and a sample of one. Blaszczynski and Nower give a structured account of an imagery-based approach, but their technique paper is not equivalent to a large controlled trial. Compare these methods without pretending that a case report, a clinical technique article and a pooled drug-trial analysis have the same evidential strength.

For the syllabus debates, idiographic evidence describes one person in depth; nomothetic evidence looks for patterns across a group. Generalisation asks whether a result applies beyond its sample. Interviews can provide context but may rely on recall, while standardised measures support comparisons but do not capture every aspect of a person's life. These are educational evaluations, not clinical decisions or treatment recommendations.