Coleman 2005.

Methods	Aim of study: To examine the effects of age and exposure to educational material on schizophrenia on stigma in student nurses Study design: Cluster RCT Size and description of clusters: Participants were randomly assigned, by class section, to either treatment or control. Size of clusters not specified Recruitment: From college course Inclusion/exclusion criteria: None stated Informed consent: Yes Ethical approval: Not stated, although 'they were treated in accordance with the ethical principles of the American Psychological Association' Funding: None stated Consumer involvement: None stated
Participants	Description: Second year nursing students in a Community College School of Nursing Geographic location: Suffolk County, New York, USA Income level of country: High Setting: Community College Number: Total randomised 240 (intervention 119, control 121), included in analysis 240 (prejudice 240) Age: mean 35.4, (SD for whole group not stated, 9.2 and 8.7 for intervention and control groups) Gender: 15.43 male Ethnicity: not stated Other social/demographic details: For intervention and control groups respectively (no overall data available): Mean (SD) years of education 15 (1.4), 15.3 (1.4); previous work or experience with the mentally ill 3.36%, 0%; planned future work with the mentally ill 48.73%, 41.32%
Interventions	Aim of intervention:  To educate about schizophrenia Content of intervention: Video entitled ‘Fires of the mind, dark voices: Schizophrenia’‐ an educational documentary. Covers the causes, treatments and outcomes for persons with schizophrenia. Contains interviews with individuals diagnosed with schizophrenia and their families, provides first‐hand accounts of the effects of the disorder on their family dynamics, personal relationships, and overall level of functioning. Provides education on the biological causes and treatments of schizophrenia, social consequences of the symptomatology, and various treatment options including psychopharmacology, social skills training, and outpatient cognitive‐behaviour therapy, giving a comprehensive look at the causes, treatments and outcomes for individuals with chronic schizophrenia. Aimed to provide information in an objective, straightforward manner and by design avoids emotional content and tone Content of control:  Educational video entitled ‘What is diabetes' from the American Association of Diabetes Educators patient education video series, covering the cause, treatments and maintenance of health, diet, exercise and effective ways to cope with the effects diabetes can have on an individual’s health. It did not discuss any psychological components of diabetes nor did it offer any advice on coping or treatment of any psychological effect of diabetes Details of co‐interventions in all groups: None Delivery of intervention: Duration of intervention video one hour, not stated for control video Providers: Intervention: Produced by the US television Discovery Channel.  Control produced by American Association of Diabetes Educators Type(s) of mass media: Audiovisual (video, television documentary) Number of mass media components: One Combined with non‐mass media component: No Contains personal narrative:  Yes Celebrities involved:  No Fictional portrayals: No Primary message: Biomedical Mental health condition: Schizophrenia
Outcomes	Primary outcome measures: Discrimination: none. Prejudice: Custodial Mental Illness Ideology Scale (CMI); three subscales of the Opinions about Mental Illness Scale (OMI): Authoritarianism (OMI‐A), Benevolence (OMI‐B) and Social Restrictiveness (OMI‐SR). Two of the OMI subscales Mental Hygiene Etiology (OMI‐C) and Interpersonal Etiology (OMI‐E) focus on aetiology beliefs rather than stigma and so were not considered stigma outcomes Secondary outcome measures: None Measure(s) selected for analysis if multiple measures per outcome: CMI (as authors based hypothesis on this) Methods of assessing outcome measures: Self‐complete questionnaire Validity and reliability of primary outcome measures: CMI: Spearman‐Brown corrected reliability estimate of 0.85 and test re‐test reliability was in the mid 0.80s. Two of the OMI subscales have low internal consistency (C and E) but neither were considered to be stigma outcomes (see above). The other OMI scales had internal consistency as follows: Authoritarianism 0.77 to 0.80, Benevolence 0.70 to 0.72; Social Restrictiveness 0.71 to 0.76 Methods of follow‐up for non‐respondents:  N/A Timing of outcome assessment: Immediate
Notes	Contact with author: Attempted, unsuccessful Power calculation: None stated Unpublished: Dissertation
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Quote: ‘Participants were randomly assigned, by class section, to either treatment or control’
Allocation concealment (selection bias)	Unclear risk	Quote: ‘Participants were randomly assigned, by class section, to either treatment or control’
Incomplete outcome data (attrition bias) All outcomes	Unclear risk	Comment: Follow‐up was immediate and no statement was given about any missing data
Selective reporting (reporting bias)	Unclear risk	Comment:  No protocol mentioned.  Study reports that one scale and five subscales were assessed and these are reported.
Other bias	High risk	Comment: Unclear risk for lack of evidence for reliability or validity of primary outcomes. CMI: Spearman‐Brown corrected reliability estimate of 0.85 and test re‐test reliability was in the mid 0.80s. Two of the OMI subscales have low internal consistency (C and E) but neither were considered to be stigma outcomes (see above). The other OMI scales had internal consistency as follows: Authoritarianism 0.77 to 0.80, Benevolence 0.70 to 0.72; Social Restrictiveness 0.71 to 0.76, however these data were from old citations and internal consistency for the current study population was not reported. High risk due to it being a cluster trial with the number of clusters being unstated and with some baseline differences apparent e.g. 17.64 vs 13.22% male and 48.73 vs 41.32 planned future work with mentally ill. Also 3.36% vs 0% of treatment and control groups respectively had previous work experience and experience generally with people with mental illness and in discussion it is stated that this is significant. N/A for evidence of counter‐discourse in follow‐up period N/A for outcomes between audience members and non‐audience not compared Unclear risk as the analysis did not take clustering into account, or did not state that this was done or how Low risk for other items listed as potential other sources of bias in Higgins 2011 (section 8.15.1)
Blinding of participants and personnel (performance bias) All outcomes	High risk	Comment: Blinding of participants and personnel not possible
Blinding of outcome assessment (detection bias) All outcomes	High risk	Comment: Outcome assessment was by self‐complete questionnaire