Open or Locked Wards? Randomization Tests the Difference

Philippe Pinel releasing [patients] from their chains at the Bicetre asylum in Paris in 1793, Charles Louis Lucien Muller

Letting depressed inpatients come and go modestly outperforms locking them in

STUDY: Ni P et al, Frontiers in Psychiatry 2026

STUDY TYPE: Meta-analysis of randomized controlled trials

FUNDING: Independent

Background

Locked wards keep depressed inpatients safe but can leave them feeling controlled and passive, impairing their engagement in treatment. Open wards let patients come and go during the day, trading some safety oversight for autonomy.

The Study

• Nine trials randomized 797 patients with depression to open or locked wards in China, meta-analyzed here.

Results

Adherence was the clearest winner. It was 23% higher with open wards (RR 1.23), with no meaningful variation across studies.

Depression scores also favored open wards on both scales, but the difference was small: about 4 points lower on the self-rated SDS (an 80-point scale) and less than 1 point lower on observer-rated HAM-D (approx 50-point scale). Statistically real, but not enough to notice at the bedside. Response rate showed no difference.

Limitations

Safety outcomes like suicide and self-harm weren’t tracked. All nine trials came from China, none was registered, and blinding patients or staff to ward type was impossible.

Practice Implications

1. Open wards make a meaningful difference in adherence and a small benefit for depression. Now let’s look for international studies and better tracking of safety.