Abstract:
A procedure called endometrial scratching has been suggested to improve the success rate of in vitro fertilisation (IVF). The procedure can be performed by a pipelle biopsy, the action of which results in a ‘scratch’ to the endometrium which is proposed to facilitate embryo implantation (Chapter 1). Meta-analysis of randomised trials suggested an improved probability of live birth from endometrial scratching performed prior to IVF, intrauterine insemination (IUI) or intercourse (Chapter 2). Sensitivity analyses reduced the effect estimates, and the magnitude of any benefit from endometrial scratching remained unclear. Despite an uncertain evidence base, an international survey found that most clinicians already offer endometrial scratching to their IVF patients, usually at a cost to the patient (Chapter 3). A randomised trial (the Pipelle for Pregnancy (PIP) study) was undertaken comparing endometrial scratching to no procedure, in women undergoing IVF (Chapter 4). A total of 1364 women were randomised across 13 sites in five countries. There was no evidence of an improved probability of clinical pregnancy from endometrial scratching: RR 1.03 (95% CI 0.88 to 1.21). The median pain score from endometrial scratching was 3.5/10 and 1% of women reported adverse events such as fainting. A separate randomised survey was conducted to assess the utility of a placebo procedure for endometrial scratching; observing that women undergoing the placebo procedure were blind to their allocation, but women undergoing endometrial scratching were not (Chapter 5). Despite the large size of the PIP study, simulation modelling suggests the trial only had a 42% probability of being able to shift the existing meta-analysis definitively, due to large between-study heterogeneity (Chapter 6). The combined results from the PIP study and two other large ongoing studies are anticipated to have sufficient power to conclusively impact the sensitivity analysis such that the pooled results demonstrate no effect from endometrial scratching. Based on the results of the PIP study, endometrial scratching should not be offered to women undergoing IVF; the procedure is often painful with a small risk of adverse events, a cost to the patient, and no association with any increased probability of conception.