Brown 1998.

Methods	Study design: parallel, randomised controlled trial Number of arms: 2 Experimental arm: fluoxetine 20 mg/day Control arm: placebo Analysis: per protocol; 1 withdrawn (treatment), excluded from analysis
Participants	Georgraphical location: Scotland Setting: Astley Ainslie Hospital Number of participants: 20 Diagnosis: stroke, time from stroke to randomisation not reported Inclusion criteria: 1) following a stroke who had a history of emotionalism at least 4 weeks' duration Exclusion criteria: 1) cognitive impairment, 2) dysphasia, 3) major depressive disorder Age: overall mean age not reported Numbers included in treatment group: 10 participants (55% men, mean age 61.4 years, SD 8.6) Numbers included in control group: 10 participants (60% men, mean age 63.7 years, SD 5.4) Emotionalism criteria: emotionalism of at least 4 weeks duration assessed during semi‐structured interview using a modified Lawson and MacLeod rating scale, in addition to frequency of outbursts
Interventions	Treatment: fluoxetine 20 mg/day Control: matched placebo Duration: treatment continued for 10 days
Outcomes	Outcomes Emotional outburst grade (modified Lawson and Macleod gradings) Depression (HDRS) Cognitive functioning (MMSE) Side effects (24‐item checklist of possible symptoms)
Notes	Able to use in analysis 50% reduction in frequency of emotionalism outbursts Leaving the study early Unable to use in analysis: HDRS, Lawson and Macleod Scale, self‐rating scales (mean and SD not presented) Adverse events (data not presented)
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Quote: "The patients were randomly allocated by an independent statistician …" pp.456 Comments: participants were randomly allocated. Method of allocation not detailed
Allocation concealment (selection bias)	Low risk	Quote: "The medication was repackaged so as to make the active and placebo capsules identical to each other" pp. 456
Blinding of participants and personnel (performance bias) All outcomes	Low risk	Quote: "The patients, nursing staff and rating clinicians were blinded to the allocation of active or placebo medication" pp. 456
Blinding of outcome assessment (detection bias) All outcomes	Low risk	Quote: "The patients, nursing staff and rating clinicians were blinded to the allocation of active or placebo medication" pp. 456
Incomplete outcome data (attrition bias) All outcomes	Low risk	Quote: "One patient had to be withdrawn because he developed a generalised rash on active …" pp. 456 Comments: dropouts or exclusions were detailed. Reason for dropouts or exclusions not related to group allocation
Selective reporting (reporting bias)	Low risk	Comments: all prespecified outcomes reported. No trial protocol available prior to randomisation of first participant
Other bias	Unclear risk	Comments: no other bias detected