- Design
- umbrella review of nine systematic reviews and meta-analyses, 92 comparisons, AMSTAR-2 and GRADE applied
- Population
- postpartum and postabortion women receiving immediate or delayed long-acting reversible contraception
- Primary outcome
- six-month pregnancy rate and device expulsion
- Effect
- pregnancy RR 0.16 (95% CI 0.04-0.71); expulsion after vaginal delivery RR 5.09 (2.04-12.73)
An umbrella review screened 1,998 records and drew nine systematic reviews yielding 92 comparisons of immediate against delayed long-acting reversible contraception after delivery or abortion. Methodological quality was rated with AMSTAR-2 and certainty with GRADE.
Immediate insertion raised initiation, raised six-month use, and lowered six-month pregnancy rates in postpartum women (RR 0.16, 95% CI 0.04-0.71, moderate certainty), with the same pattern after abortion at low certainty. The cost is expulsion. Immediate intrauterine device insertion raised expulsion sharply, most of all after vaginal delivery (RR 5.09, 95% CI 2.04-12.73, high certainty) and in first-trimester medical abortion (RR 2.02, 1.37-2.98, moderate certainty).
That combination is not a contradiction — it is a choice between two failure modes. A woman who leaves without a device may never return for one; a woman who leaves with one may lose it and not notice. Which risk dominates depends on whether follow-up in your setting is realistic, and in much of Indian practice the postpartum visit is the one that is missed.
So the change is in how the conversation is framed rather than in a blanket policy. Offer immediate insertion, say plainly that the device is more likely to come out than one placed later, teach the woman to check her strings and to present if she cannot feel them, and arrange a check rather than assuming one will happen. Implants are not subject to expulsion and are the cleaner immediate option where available.
- Offer immediate postpartum or postabortion LARC rather than deferring the discussion to a visit that may not happen
- State the expulsion risk explicitly — after vaginal delivery it was about five times that of delayed insertion
- Teach string self-checks before discharge and give a clear instruction to present if the strings cannot be felt
- Prefer an implant for immediate placement where one is available, since expulsion does not apply
- Arrange and document a device check rather than relying on the routine postpartum visit
Why it matters
The routine advice to wait six weeks assumes the woman comes back, and the evidence now prices that assumption.
Don't overread it
The postabortion pregnancy findings were low certainty, and an umbrella review inherits the limitations of the reviews inside it rather than correcting them.
The statistics, in plain English
The pregnancy risk ratio of 0.16 with an interval from 0.04 to 0.71 is a large effect but a very wide interval, which means few pregnancies were observed — the direction is reliable, the size is not. The expulsion interval, 2.04 to 12.73, is wide for the same reason yet stays well above 1.0, and was graded high certainty because the reviews agreed. These are pooled estimates from reviews of reviews, so the underlying studies feed into more than one comparison and the numbers are not independent of each other.
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.
Scan to keep reading on your phone. No account needed to start.
Tomorrow morning, before your first patient
One edition a day for obstetrics & gynaecology, written by the desk, every claim tied to its paper. Six minutes.
Get the app — free