Ref (type)	Population	Outcome, Interventions	Results and statistical analysis	Effect size	Favours
Adverse effects
Systematic review	People (>90% aged >16 years) with idiopathic chronic constipation (modified Rome II criteria and negative investigations) 3 RCTs in this analysis	Total number of adverse events with lubiprostone with placebo Absolute results not reported	RR 1.79 95% CI 1.21 to 2.65 NNH = 4 95% CI 3 to 6	Small effect size	placebo
Systematic review	People (>90% aged >16 years) with idiopathic chronic constipation (modified Rome II criteria and negative investigations) 3 RCTs in this analysis	Diarrhoea with lubiprostone with placebo Absolute results not reported	RR 4.46 95% CI 1.28 to 15.48	Moderate effect size	placebo
Systematic review	People (>90% aged >16 years) with idiopathic chronic constipation (modified Rome II criteria and negative investigations) 3 RCTs in this analysis	Nausea with lubiprostone with placebo Absolute results not reported	RR 7.27 95% CI 3.76 to 14.06	Large effect size	placebo
RCT 4-armed trial	People, average age about 40 years, with idiopathic chronic constipation	At least 1 adverse event 17/44 (39%) with highest-dose lubiprostone 13/43 (30%) with middle-dose lubiprostone 1/41 (2%) with lowest-dose lubiprostone 2/42 (5%) with placebo	P value among groups not reported
RCT 4-armed trial	People, average age about 40 years, with idiopathic chronic constipation	Diarrhoea 8/44 (18%) with highest-dose lubiprostone 4/43 (9%) with middle-dose lubiprostone 0/41 (0%) with lowest-dose lubiprostone 0/42 (0%) with placebo	P value among groups not reported Incidence of diarrhoea significantly higher in highest-dose arm v placebo (P = 0.0037) and middle dose arm v placebo (P = 0.0429) No other pairwise analysis reported	Effect size not calculated	placebo compared with highest-dose and middle-dose lubiprostone
RCT 4-armed trial	People, average age about 40 years, with idiopathic chronic constipation	Nausea 7/44 (16%) with highest-dose lubiprostone 3/43 (7%) with middle-dose lubiprostone 0/41 (0%) with lowest-dose lubiprostone 0/42 (0%) with placebo	P value among groups not reported Incidence of nausea significantly higher in highest-dose arm v placebo (P = 0.007) No other pairwise analysis reported	Effect size not calculated	placebo compared with highest-dose lubiprostone