antenatal period. Adequately, powered subgroup analyses taking
account of whether forceps or vacuum or indeed both
were used would require a larger sample size again. When
we surveyed obstetricians, there was concern raised about
the validity of a RCT in evaluating a surgical approach that
is not dichotomised into two types of practice but one that is
based on clinical judgement.2 There is some justification for
this criticism in that obstetricians vary their approach in
response to subtle clinical findings, and we therefore conducted
a prospective cohort study alongside the trial to take
account of the full spectrum of clinical practice.
Comparison with existing literature
A number of population-based studies have provided conflicting
evidence on the relationship between episiotomy and anal
sphincter tears at operative vaginal delivery.4,5 Themost striking
comparison, however, relates to the findings of the recentDutch
study where episiotomy was associated with a dramatic reduction
in anal sphincter tears for both vacuum and forceps delivery.
6 Themagnitude of association, if confirmedwithin our pilot
RCT, would easily have produced statistically significant results,
butwe found only small differences. Thismay reflect differences
in the conduct of operative vaginal delivery, the technique and
timing of episiotomy or the marked difference in reported rates
of anal sphincter tears between the different settings. Equally, it
highlights the shortcomings of observational studies when evaluating
a complex clinical intervention.
Implications for practice
In a previous national survey, two-thirds of obstetricians
held the view that routine use of episiotomy decreases the
likelihood of anal sphincter tears at forceps delivery with
a divided view for vacuum delivery (45% decreases risk
and 42% no difference).2 Few obstetricians (