Category A studies: 7 RCTs (576 patients).
Category B studies: 4 pre-test, post-test studies (385 patients).
Category C studies (mixed population): 8 RCTs (757 patients).
Most studies demonstrated that knowledge and compliance could be improved by educational interventions. Compliance seems to be most readily influenced by interventions including behavioural components. A few studies indicated that relapse and symptomatology could be influenced by educational interventions as well. No influence of the duration of interventions was found. Where educability was studied, age, medication and level of symptoms were potential predictors.
Only results from category A studies are reported below.
Knowledge (6 RCTs): 5 out of 6 RCTs reported a statistically-significant benefit from the intervention (P<0.05).
Compliance (5 RCTs): 2 out of 5 RCTs reported a statistically-significant benefit from the intervention (P<0.05).
Relapse (3 RCTs): 1 out of 3 RCTs reported a statistically-significant benefit from the intervention (P<0.05).
Symptoms (5 RCTs): 2 out of 5 RCTs reported a statistically-significant benefit from the intervention (P<0.05).
Social function (4 RCTs): results were inconsistent with 2 out of 4 RCTs reporting a statistically-significant benefit from the intervention (P<0.05).
Insight (2 RCTs): results were inconsistent with 1 out of 2 RCTs reporting a statistically-significant benefit from the intervention (P<0.05).
Quality of life (1 RCT, 114 patients): a statistically-significant benefit from the intervention was reported.
Satisfaction (1 RCT, 46 patients): a statistically-significant benefit from the intervention was reported.
Methodological problems.
The main methodological problems in the category A studies related to acceptance to participate (in 34 to 100% of studies), lack of blind rating, attrition (drop-out rate ranged from 4.5 to 35.5%), intention to treat analysis, inter-rater reliability measures, and inclusion of changes in medication as a potential confounder for the impact on outcome measures.
Educational method.
Category A studies: all didactic studies where knowledge was measured reported improved level of knowledge. Only 1 out of 3 RCTs measuring compliance reported benefit. Only 1 out of 2 RCTs of didactic behavioural interventions measuring knowledge reported benefit, and only 1 out of 2 RCTs measuring compliance reported benefit.
All studies: overall, all didactic studies measuring knowledge reported benefit, compared to 2 out of 4 RCTs of didactic and behavioural interventions.
Duration of intervention: 7 out of 9 short programmes had an impact on knowledge, compared to 4 out of 4 studies with long interventions.
Educability: there was no apparent difference in knowledge change between patients with illness of 15 years duration or less (4 RCTs), compared to patients with duration of illness greater than 15 years (2 RCTs).
Gender (2 RCTs): No apparent influence of gender on knowledge gain was found.