Prapas 2012.

Study characteristics
Methods	Study design: RCT Setting: Iakentro IVF centre Period: patients enrolled from June 2009 through November 2010 A total of 400 cycles, of which 200 IVF/ICSI and 200 oocyte donor (OD), were randomly assigned to have their uterine cavity injected (group I) or not (group II)
Participants	Group I (injection of the uterine cavity with embryo culture supernatant) Group I consisted of two subgroups, IA (ET on day 3) and IB (ET on day 5) Group II (no injection) Group II consisted of IIA (ET on day 3) and IIB (ET on day 5) After the exclusion of OD cycles (based on data from direct communication), the remaining IVF/ICSI 190 cases (200 cases allocated, 10 excluded because of difficult ET) were allocated into the intervention and control groups as follows; flushing group (study group) 93 cases ‐ non‐flushing group (control) 97 cases Inclusion criteria: All women had a history of at least one previous unsuccessful IVF/ET. age ≤ 38 years for the IVF women and ≤ 50 years for the oocyte receivers, without known endometriosis, hydrosalpinx or uterine anomalies, including small submucosal myomas or polyps. Baseline characteristics of the participants: flushing group i (n = 188) BMI 24.3±4.9 Age (mean ±SD) 37.2±5.2 Number of oocytes (mean ±SD) 13.4±5.0 Endometrial thickness (mean ±SD) 10.9±2.1 Mean embryo transferred (mean ±SD) 2.20±0.70 Embryo quality (mean ±SD) 1.84±0.94 Fertilisation rate non‐flushing group ii (n = 196) BMI 24.4±4.7 Age (mean ±SD) 35.8±5.8 Number of oocytes (mean ±SD) 13.3±4.8 Endometrial thickness (mean ±SD) 10.8±2.6 Mean embryo transferred (mean ±SD) 2.24±0.69 Embryo quality (mean ±SD) 1.82±0.93 Fertilisation rate
Interventions	Injection of embryo culture supernatant into the uterine cavity, 30 min before the embryo transfer on either day 3 or 5, in IVF/ICSI versus simple air insertion (as a placebo intervention used in the control group)
Outcomes	Main outcome measure(s): Clinical pregnancy rate per transfer Implantation rate Miscarriage rate Ectopic pregnancy rate A pregnancy test was performed 15 days after the ET, and, if positive, an ultrasound scan was scheduled after two weeks to determine the number and status of implanted embryos. The concurrency of a positive beta‐ HCG test and a foetal heartbeat (seen by ultrasound) was defined as a clinical pregnancy.
Notes	The study was approved by the Institutional Review Board. Four hundred women were allocated into group I or II when they were called to be informed about the day of their embryo transfer. All cases had a mock transfer in a cycle previous to IVF and if difficulty was encountered a cervical dilatation was performed. All cases included in the statistical analysis had at least one good quality embryo. Separate details on LBR were obtained after communication with the authors for the 200 non‐donor cycles. These cycles were eligible
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Allocation sequence generated from a computerised random number table
Allocation concealment (selection bias)	Low risk	Computerized allocation performed by a third party (midwife not further involved in the procedures) before initiation of the intervention (day of hcg injection)
Blinding of participants and personnel (performance bias) All outcomes	Low risk	Data from personal communication with author:"The doctor who performed the ET was blinded about the necessity of injection or not until the moment of ET. In all cases, there were two insertions, the first was embryo culture or simple air (flushing/non‐flushing group) and the second was ET"
Blinding of outcome assessment (detection bias) All outcomes	Low risk	According to information from personal communication with author, the clinician who performed the outcome assessment was blinded
Incomplete outcome data (attrition bias) All outcomes	Low risk	Authors report no drop‐ outs / losses to follow‐up In addition, due to difficult transfer 12 cases from group I and 4 cases from group II were not included in the statistical analysis
Selective reporting (reporting bias)	Unclear risk	Adverse events and clinical outcomes were provided by authors after communication‐ no registered protocol
Other bias	Low risk	There is insufficient rationale or evidence that any problem would introduce bias