Wadhwa 2015.

Study characteristics
Methods	Randomised controlled trial, 3 groups, set in the Department of Obstetrics and Gynaecology at a tertiary care centre, India August 2012 to March 2014 Number of participants randomised: 225 (26 not randomised), total of 251 Number of participants analysed: 251
Participants	Inclusion criteria: women aged between 18 and 38 years with primary or secondary infertility who were attending the clinic planning stimulated intrauterine insemination (IUI), with either both or 1 patent (functioning) fallopian tube (demonstrated by "laparohysteroscopy" or hysterosalpingography) Exclusion criteria: known pelvic inflammatory disease with bilateral tubal blockage; severe male factor infertility with intrauterine pathology (submucosal fibroid, endometrial polyp, adhesions); acute vaginal or cervical infection Cause of infertility: unexplained, mild male factor, tubal factor (unilateral)
Interventions	Intervention group A: endometrial scratching on Day 19 to 24 of the spontaneous menstrual cycle that precedes the fertility treatment and IUI Intervention group B: endometrial scratching between Day 1 and Day 6 of the same spontaneous menstrual cycle in which ovarian stimulation and IUI are done Control group C: no endometrial scratching All groups: each participant underwent single IUI 36 hours after human chorionic gonadotropin trigger, or 24 hours later if luteinising hormone surge was positive Degree of endometrial injury: endometrial aspiration cannula Timing of endometrial injury: in group A, injury was during the luteal phase between Day 19 and 24 of the preceding spontaneous menstrual cycle; in group B, injury was during the follicular phase before Day 6 of the same spontaneous menstrual cycle. Endometrial scratching was performed in the first cycle only Study length: 1 cycle (the paper reports pregnancy rates over 3 cycles, but as the numbers of participants attending for the second and third cycles are unbalanced, study authors provided data for the first cycle only) Types of conception: IUI and intercourse; women who failed to commence stimulated IUI tried to conceive spontaneously and were followed up and included in the analysis
Outcomes	Reported in the paper: Clinical pregnancy rate: confirmed by the presence of a gestational sac on ultrasonography Miscarriage rate (by author correspondence: number of clinical pregnancy losses before 12 completed weeks' gestation) Multiple pregnancy rate Pain/bleeding not actively recorded but noted "no complaints of severe pain"
Notes	Funding source: no financial support or sponsorship Conflicts of interest: none declared Trial registration: CTRI/2012/12/004356 (retrospectively registered) Author correspondence was undertaken
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Quote: "the random allocation was generated using a random number table" From author correspondence, it was discovered that 11 participants in group A and 15 in group B were not randomised but were allocated to the intervention group to replace participants who dropped out. Therefore 26 participants were not randomly allocated However study authors provided data for randomised participants only
Allocation concealment (selection bias)	High risk	Quote: "sealed envelope system was used...allocation was done by the doctor posted in infertility outpatient department" Study authors confirmed that the envelopes were not numbered
Blinding of participants (performance bias)	High risk	Quote: "this study was not blinded" Lack of participant blinding is anticipated to introduce performance bias
Blinding of personnel (performance bias)	High risk	Study authors did not report blinding of personnel
Blinding of outcome assessment (detection bias) All outcomes	Low risk	Blinding of outcome assessors was not reported and is unlikely; however outcomes are unlikely to be influenced by lack of blinding
Incomplete outcome data (attrition bias) All outcomes	Low risk	Eleven participants from group A, 15 from group B, and zero from group C failed to commence their allocated procedure (reasons not reported). Although it was intended for all participants to complete 3 IUI cycles (unless they fell pregnant), only 93 cycles took place in group A, 156 in group B, and 113 in group C (number of cycles in group C provided by author correspondence). Additionally this gave group B more opportunities to conceive, and it is possible that this could account for the higher pregnancy rate in group B. However, intention‐to‐treat analysis was performed, and data were available for those who did not attend for IUI
Selective reporting (reporting bias)	Unclear risk	Retrospectively registered. Study authors confirmed that they did not record any live birth or pain
Other bias	Low risk	Groups B and C were not advised abstinence prior to their IUI cycle, but no pregnancies were reported during this period