Langemark 1993.

Methods	Multicentre, double‐blind, randomised, response‐conditional cross‐over study: paroxetine vs sulpiride No control for symptomatic/analgesic medications use
Participants	Country: Denmark N = 50; Sex: 20 male, 30 female Mean age: 42 (range 20 ‐ 70) Diagnosis: Chronic tension‐type headache (not defined) of at least 6 months duration and no more than 14 headache‐free days per months Exclusion criteria: patients suffering migraine more than one day per month; moderate or severe cardiovascular disease or other chronic disease Recruitment: mailed questionnaire among patients from a private neurology clinic and a neurological hospital department
Interventions	N = 25 first treated with paroxetine to a maximum of 30 mg/day N = 25 first treated with sulpiride to a maximum of 100 mg/day Active treatment: 16 weeks (8 weeks + 8 weeks, no in‐between washout period if patients crossed over)
Outcomes	1. Severity Index (5 ordinal scale: a. 'no headache'; b. 'slight'; c. 'moderate'; d. 'very troublesome'; e. 'worst possible') 2. Patient's global Evaluation drug performance (5 ordinal scale: a. 'worthless'; b. 'poor'; c. 'fair'; d. 'good'; e. 'excellent') 3. Observer's Global Evaluation drug performance (5 ordinal scale: a. 'worthless'; b. 'poor'; c. 'fair'; d. 'good'; e. 'excellent') 4. Symptomatic/analgesic drug consumption 5. Side effects (4 ordinal scale: a. 'none'; b. 'slight'; c. 'moderate'; d. 'severe = discontinues drug')
Notes	First 8 weeks: 10 dropouts (20%): paroxetine: 7 (6 for side effects, 1 for other reason) sulpiride: 3 (2 for side effects, 1 for other reason) 13 patients crossed over to sulpiride after paroxetine for no response or intolerable side effects, while 10 crossed over to paroxetine after sulpiride Per protocol analyses
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	No information
Allocation concealment (selection bias)	Unclear risk	No information
Blinding (performance bias and detection bias) All outcomes	High risk	Drugs identical in appearance (but the two groups received different numbers of tablets)
Incomplete outcome data (attrition bias) All outcomes	High risk	Moderate dropout rate, unbalanced, reasons partially reported
Selective reporting (reporting bias)	Unclear risk	No information
Other bias	Unclear risk	Financial support not reported