Extending the hips at crowning did not reduce perineal tears

Design

Pragmatic randomised trial, non-blinded

Population

1,206 nulliparous women with singleton term pregnancies delivering vaginally

Primary outcome

Second-, third- or fourth-degree perineal laceration

Effect

66.6% vs 66.2%, P = .89 (intention-to-treat)

A pragmatic, non-blinded randomised trial allocated nulliparous women at term either to be directed to lower their legs and extend their hips immediately before crowning, or to usual care, which was mostly the flexed position. 1,206 women delivered vaginally after randomisation.

On intention-to-treat, second-degree or worse tears were no different: 66.6% with the intervention and 66.2% with usual care (P = .89). A quarter of women allocated to extension did not adopt it. In as-treated analyses, extension was associated with fewer significant tears (62.7% vs 68.7%) and fewer anal sphincter injuries (5.2% vs 9.9%), but these differences did not survive adjustment.

The primary answer is no. The as-treated signal is not randomised — women who could extend their hips may have been having easier births — so it is a reason for another trial, not a reason to change position routinely.

• Keep to proven measures: warm perineal compresses, controlled delivery of the head and manual perineal protection.
• Do not insist on hip extension if the woman prefers flexion.
• Examine carefully for anal sphincter injury after every second-degree tear.
• Record birth position in the notes so local audit can follow this question.

Why it matters

A cheap, plausible positioning change has been tested properly and did not work on its primary outcome.

Don't overread it

The fewer sphincter injuries seen with extension come from an as-treated comparison that lost significance after adjustment.

The statistics, in plain English

The intention-to-treat analysis compares women as randomised and is the fair test; it showed nothing. The as-treated analysis compares women by what they actually did, which reintroduces the bias randomisation removes. When a benefit appears only there, and disappears on adjustment, it is best read as hypothesis-generating.

Read the rest in the app

You have read your two free briefings this month. The app carries all 27 specialties, every morning, free — and this finding is waiting in it.

Get Daily Doctor — free

Scan to keep reading on your phone. No account needed to start.