Talking to parents in intensive care works better than the soft options

Design

systematic review and meta-analysis of quantitative studies with control groups, 2010 to December 2024

Population

16,639 parents of NICU infants across 160 studies; mean age 30.2 years, 93.5% mothers

Primary outcome

parent anxiety, depression, stress and trauma symptoms on validated self-report measures

Effect

anxiety: psychotherapy/emotional support Hedges' g −1.66 (95% CI −1.97 to −1.34); family-centred care −1.49 (−2.80 to −0.19)

This review covered 160 studies and 16,639 parents of infants in neonatal intensive care, with 85 estimates pooled in meta-analysis. Mean parent age was 30.2 years and 93.5% were mothers, which is itself worth noting when planning who a service actually reaches.

For anxiety, three categories of intervention showed significant effects: bonding interventions (Hedges' g −0.90, 95% CI −1.50 to −0.30), family-centred care and education (−1.49, −2.80 to −0.19) and psychotherapy and emotional support (−1.66, −1.97 to −1.34). For parent stress, family-centred care and education (−0.75, −1.30 to −0.21), psychotherapy (−1.54, −2.17 to −0.19) and meditation or holistic interventions (−0.65, −1.20 to −0.10) all helped. For depression, only expressive and artistic therapies (−0.36, −0.68 to −0.04) and psychotherapy (−0.95, −1.80 to −0.10) reached significance.

Psychotherapy and emotional support was the only category with a consistent effect across all three outcomes, and the only one whose anxiety estimate is precise. Family-centred care and education is the one the authors flag as requiring few extra resources — it is largely about how the unit already communicates rather than about adding a service. Both are more useful conclusions than the usual advice to be supportive.

• Ask about parental anxiety and low mood directly during admission, using the same words each time
• Build family-centred care and education into ward routine — it needs organisation, not new staff
• Refer to psychotherapy or structured emotional support where symptoms are established
• Note that 93.5% of participants were mothers; plan deliberately for how fathers and partners are reached

Why it matters

It separates the interventions with consistent evidence from the ones that simply feel kind.

Don't overread it

These are unblinded interventions with self-reported outcomes — the effect sizes are almost certainly inflated.

The statistics, in plain English

Hedges' g of around −1.5 is a large standardised effect, but several of these intervals are enormous — family-centred care for anxiety runs from −2.80 to −0.19, which spans 'transformative' to 'barely detectable'. The psychotherapy anxiety estimate, −1.97 to −1.34, is the tight one and is what makes that category the safest recommendation. Self-report questionnaires in unblinded interventions reliably overstate benefit, because parents who received attention report feeling better.

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.