Design: Systematic review and meta-analysis of RCTs and non-RCTs, with meta-regression
Population: Adult hospital inpatients (11 RCTs in primary analysis)
Primary outcome: Rate of falls, injurious falls, proportion of fallers
Effect: Falls IRR 0.93 (0.83-1.04); injurious falls IRR 1.00 (0.87-1.16)

A systematic review and meta-analyses examined mobilisation alarms, which sound when an at-risk patient gets up, for preventing falls among adult hospital inpatients.

In randomised trials, alarms did not reduce the rate of falls (IRR 0.93, 95% CI 0.83-1.04; 13 comparisons), injurious falls (IRR 1.00, 0.87-1.16) or the proportion of patients who fell (OR 1.11, 0.91-1.35). Meta-regression pointed even more strongly to no effect, and prospective trial registration was the factor most protective against over-optimistic results.

Alarms are widely used and create noise, staff interruption and restriction of mobility. The best-quality evidence shows they do not deliver the benefit they are used for.

The resources spent on routine alarms would be better directed at measures that address why patients fall.

Do not rely on routine, ward-wide mobilisation alarms to prevent falls.
Prioritise individual falls risk assessment, medication review and delirium prevention.
Encourage supervised mobility rather than restricting movement.
Keep alarms, if used at all, for specific patients with a clear reason.

Why it matters

A common, disruptive intervention does not work, which frees staff time for measures that might.

The statistics, in plain English

An IRR of 0.93 with an interval crossing 1.0 means the data are compatible with anything from a 17% reduction to a 4% increase in falls. That studies with prospective registration showed less benefit suggests earlier positive results were inflated by bias.