Membrane sweep: what does it really involve?

During this pelvic examination, the practitioner may "sweep the membranes" — that is, use their fingers to separate the amniotic sac from the uterine wall. This can trigger the release of chemical mediators called prostaglandins, which soften the cervix and prepare the body for labour, and may also irritate the cervix, causing it to contract [1]. 
 

The plasma concentration of prostaglandins following membrane sweeping was found to be equivalent to 10% of the level observed during labour [2].

There is no guarantee that membrane sweeping will lead to induction of labour or contractions. Membrane sweeping is often offered as a first option to help start labour and encourage a natural birth before setting an induction date.

Studies are not necessarily reliable

In a review of 44 studies, they examined the effect of membrane sweeping on the induction of labour [3]. 14 studies indicated that if the cervix was closed, they performed a cervical massage instead of a membrane sweep. Thus, in these 14 studies, "membrane sweeping" could in fact have been either a cervical massage or a sweep. 

For the most part, the authors judged that the studies they included had a low or unclear risk of bias.

However, in general, the risk of performance bias was high, as the 44 studies did not carry out a study blinding, otherwise known as blinding. 

Blinding means that clinicians and researchers do not know who is receiving which treatment — whether the actual treatment, no treatment, or a placebo. When blinding is not used, as is the case in all these studies, it can introduce bias, meaning clinicians may be inclined to provide better care to the treatment group in the hope that it proves effective. 

For example, if a provider knew that a person was in the treatment group for membrane sweeping, they might delay scheduling a formal induction in the hope that the person in the treatment group would go into labour spontaneously on their own.

The results are mixed

Cervical massage: a better alternative?

There is less data available on whether membrane sweeping or cervical massage can contribute to cervical ripening or softening. In a randomised trial, 165 participants with a low Bishop score — that is, an unripe cervix at 41 weeks and four days of pregnancy — were randomly assigned to membrane sweeping, cervical massage only, or no treatment [6].
 

Both the cervical massage and membrane sweeping groups were associated with a significant increase in mean Bishop score 48 hours after treatment compared with the control group. This is evidence that cervical massage may be an effective alternative to membrane sweeping and that it could be offered to pregnant women whose cervix is closed. 

When the reviewers of the study looked at other outcomes, they found no difference between the groups regarding the rate of caesarean sections, the use of forceps or ventouse to assist delivery, or serious illness or death in mothers or babies. 

Risk of premature rupture of membranes
 

It is possible that membrane sweeping, but not cervical massage, increases the risk of membrane rupture before labour. The case where the waters break BEFORE the onset of labour is referred to as "premature rupture of membranes", regardless of the stage of pregnancy (in other words, the term "premature" refers solely to the timing relative to the onset of labour).

Labour often begins within the hours following membrane rupture. It should be noted that the risk of foeto-maternal infection increases significantly after 12 hours of ruptured membranes.

In a randomised trial of 300 women, the risk of membrane rupture was 9% in the group who underwent a membrane sweep, compared to 0% in the group who received no treatment, for women whose cervix was dilated by more than 1 cm at the time of the membrane sweep [7]. 

Another similar randomised study of nearly 300 women showed that membrane sweeping significantly increased the risk of premature rupture of membranes (38% rupture rate following membrane sweep compared to 26% with no intervention) [8].

But the consequences of this risk should be put into perspective and weighed against the fact that:

- This situation is most commonly found in women whose cervix is dilated by more than 1 finger (i.e. women who would in any case be offered membrane rupture as the method of induction).

- If the membrane sweep causes a rupture, the protocol is to wait 48 hours in expectant management before inducing, which amounts to much the same as the planned induction. That said, it is true that any prolonged rupture carries a risk of foeto-maternal infection.

Pain and bleeding

Pain during vaginal examination and other side effects are more frequently reported by women who have had a membrane sweep [9]. 

A randomised study of more than 700 women [10], who either had a membrane sweep or no intervention, showed that the sweep multiplied vaginal bleeding by 6. The reason: the cervix is highly vascularised. The more it is touched, the more it bleeds.

They also studied the pain associated with this practice, and showed that 31% of women said it was not painful, 51% said it was somewhat painful, and 17% said it was painful or very painful. 

Sometimes this procedure is carried out routinely during a vaginal examination without any discussion of informed consent. It should only be performed if you agree to it, and must never be done without informing you first.

Its effectiveness in inducing labour is debated — some studies show a slight effect whilst others show none. The risk is that it may trigger painful, regular contractions without causing any cervical changes (false labour). Many midwives prefer to use it on women who have already given birth.

Some people have reported feeling pain or discomfort during the procedure. You may also experience some bleeding afterwards. 

Membrane sweeping, but not cervical massage, may increase the risk of premature rupture of membranes before labour.

The raspberry leaf tea may be a safer natural alternative.