After peripartum cardiomyopathy, a recovered ejection fraction does not protect against relapse in a later pregnancy

Design

Systematic review and meta-analysis of 6 cohort studies

Population

266 women with prior peripartum cardiomyopathy in a subsequent pregnancy

Primary outcome

Relapse, by recovered (LVEF 50% or more) vs non-recovered function

Effect

Relapse RR 0.77 (0.50 to 1.19); maternal death 1.7% vs 10.9%, RR 0.27 (0.09 to 0.87)

This meta-analysis pooled six cohort studies of 266 women with prior peripartum cardiomyopathy who became pregnant again, split by LVEF before the next pregnancy: recovered (50% or more, n=174) or not recovered (n=92).

Relapse did not differ significantly between the groups (rate ratio 0.77, 95% CI 0.50 to 1.19; I² = 3%). Maternal death was much lower in the recovered group (1.7% vs 10.9%; RR 0.27, 0.09 to 0.87), and those women were twice as likely to recover ventricular function afterwards (RR 2.07). Symptom worsening and obstetric or neonatal outcomes did not differ.

The common shorthand — the heart has recovered, so a next pregnancy is safe — is not supported. Recovery lowers the risk of dying, not of relapsing. Every woman with prior peripartum cardiomyopathy needs preconception cardio-obstetric counselling and joint surveillance through pregnancy and the months after it, whatever the echo shows.

• Ask every woman at booking or preconception whether she had heart failure in or after a previous pregnancy
• Refer all women with prior peripartum cardiomyopathy for cardio-obstetric review before or early in the next pregnancy
• Arrange echocardiography during pregnancy and postpartum even if LVEF had normalised
• Counsel women with LVEF below 50% that maternal mortality was about 11% in these cohorts
• Plan contraception with women who are advised against another pregnancy

Why it matters

It overturns the reassurance often given to women whose echo has normalised.

Don't overread it

These are six small cohorts with 266 women; the numbers are directional, and the mortality estimate rests on very few deaths.

The statistics, in plain English

A rate ratio of 0.77 with an interval from 0.50 to 1.19 crosses 1, so the study cannot show that recovery lowers relapse. The mortality ratio of 0.27 excludes 1, so the difference in deaths is more secure, though still based on small numbers.

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.