Brown 2007.

Study characteristics
Methods	Truly randomised trial (computer‐generated random numbers) Time of randomisation: day of surgery Double‐blind Power calculation: no ITT: yes Location: multi‐centre—16 centres, not stated where Timing and duration: July 2001 to March 2004; 2 years 8 months
Participants	Females 18 years of age and older undergoing laparoscopic gynaecological surgery only (N = 402) Indications for surgery: pelvic pain and infertility (93), infertility investigation (214), endometriosis (243) Surgery performed: laparoscopy and adhesiolysis Pre‐existing adhesions: not all women Number randomly assigned: 449 Number undergoing second‐look laparoscopy: 402 (29 withdrew, 13 declined SLL, 2 excluded intraoperatively, 18 major protocol violations) Timing second‐look laparoscopy: 4 to 8 weeks postoperative Blinding at second‐look laparoscopy: yes Exclusion criteria: < 3 adhesions lysed, systemic steroids, antineoplastic agents, radiation, pregnancy, active pelvic infection, cancer, allergy to Adept, additional non‐O+G procedures performed
Interventions	4% icodextrin versus Ringer's lactate > 100 mL every 30 minutes, no limit on amount used for irrigation. 1000 mL instilled at end
Outcomes	Analysed in review Pelvic pain Adhesions at second‐look laparoscopy Number of women with improvement in adhesion score Other outcomes Adhesions at second‐look laparoscopy Number of adhesion sites Mean AFS score reduction Number of women free of de novo adhesions "Clinical success"—study authors' definition: reduction of 3 or 30% of number of sites lysed Change in AFS score category Adhesion extent, severity Many sub‐analyses based on primary diagnosis
Notes	mAFS score Funded by Innovata Ltd, Vectura Group
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Quote: "Treatment was randomised by computer‐generated randomisation on a 1:1 basis"
Allocation concealment (selection bias)	Unclear risk	no comments on allocation concealment
Blinding of participants and personnel (performance bias) All outcomes	Unclear risk	Initial surgeons were blinded. Not clearly stated whether women were blinded.
Blinding of outcome assessment (detection bias) All outcomes	Unclear risk	Not clearly stated whether the reviewer of the second‐look laparoscopy was blinded.
Incomplete outcome data (attrition bias) All outcomes	High risk	449 women were initially recruited and randomised. 29 withdrew from the studies with reasons given for their withdrawal. However, an additional 18 women had major protocol violations and therefore, were excluded from any efficacy analyses. No clear explanation of these violations was provided. Due to this lack of information and high rate of attrition (47 women, more than 10%), this is considered to be at high risk of attrition bias
Selective reporting (reporting bias)	Unclear risk	Data presented in full for all outcomes specified, however no study protocol or trial registry identified for comparison
Other bias	Low risk