Shen 2006

Methods	Study design: parallel group randomised trial. Study dates: June 2002 to December 2005. Setting: outpatient setting. Country: China.
Participants	Inclusion criteria: participants who meet the diagnostic criteria for chronic non‐bacterial prostatitis. Exclusion criteria: infection of urinary tract, EPS culture positive for bacteria, EPS routine test negative findings, benign prostatic hyperplasia, serious psychoneurosis, narrowing of urinary tract or prostate tumour. Participants with poor compliance and missing data were excluded. Sample size: 72. Age (years): overall 20˜46, mean 31. Baseline NIH‐CPSI score: not available. Sex: men.
Interventions	Group 1 (n = 40): modified BiXieFenQing drink in morning and evening, 200 mL each + prostate massage once weekly. Group 2 (n = 32): modified BiXieFenQing drink in morning and evening, 200 mL each. Cointerventions: for participants with EPS mycoplasma culture positive, doxycycline given 0.1 g orally twice daily 2 weeks. Mycoplasma culture performed at weeks 1 and 2 after doxycycline treatment; if positive, another 2‐week doxycycline treatment was given (abnormal liver function was a contradiction for this treatment).
Outcomes	Prostatitis symptoms How measured: NIH‐CPSI subscores and 'clinical cure.' Time points measured: before and after treatment. Time points reported: before and after treatment.
Funding sources	Not mentioned.
Declarations of interest	Not mentioned.
Notes	In the 72 participants included in this trial, 1 participant was complicated by epididymitis, 7 participants with mycoplasma culture positive.
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Randomly assigned to Group 1 (40 participants) and Group 2 (32 participants) (in Chinese), but method for randomisation not mentioned.
Allocation concealment (selection bias)	Unclear risk	Allocation concealment not described.
Blinding of participants and personnel (performance bias) Subjective outcomes	High risk	Blinding of participants and personnel not described. Considering the visibly different interventions, blinding was unlikely.
Blinding of outcome assessment (detection bias) All outcomes	High risk	Blinding of outcome assessment not described. Self‐reported outcomes, see comment above.
Incomplete outcome data (attrition bias) All outcomes	Low risk	All outcomes: outcome data available for all participants.
Selective reporting (reporting bias)	Unclear risk	Unclear whether there was selective outcome reporting (no protocol available).
Other bias	Low risk	No other sources of bias identified.