Objective To review the practice regarding trial of instrumental delivery (ID) in theatre.

Study design All trials of ID in theatre over a 4 month period were retrospectively reviewed. Maternal demographics, labour and delivery data and maternal and foetal outcome measures were assessed.

Results Of the 58 trials of ID, six failed (delivery by caesarean section) (10%). Factors increasing the likelihood of failure included primiparity (p=0.002, CI 0.067 to 0.28), non-occipito-anterior position (p<0.0001, CI −0.52 to −0.25) and manual rotation, with 4 (67%) having had manual rotation attempted (p=0.001, CI −0.85 to −0.22). Five out of the 29 deliveries (17%) by a registrar were not supervised. Mal-position was reported in 38 (66%) cases, of which Ventouse was the initial management in 4 (10%), manual rotation in 11 (29%) and Keillands in 23 (61%). When compared to other instruments, Ventouse was more likely to require a second instrument or CS (p=0.67) and neonates had significantly reduced arterial cord pH's (p=0.034, CI 0.31 to 0.01). Keilland delivery was associated with trauma to the neonate, especially significant trauma (p>0.2).

Conclusions Four out of the five Royal College of Obstetricians and Gynaecologists pre- requisites for trial of ID were met consistently, yet the trust guidelines were less stringently adhered to. Further knowledge and guidance is required for rotational deliveries as weaknesses in manual rotation, Keilland's and Ventouse are identified. Documentation needs improving to accurately assess the peri-natal care and outcomes of IDs in theatre.