Background The safety of playgrounds is important to protect children
from injury, hut studies are mostly done mainly under laboratory condi
tions without epidemiological data. We investigated the safety of diff
erent playground surfaces, and types and heights of equipment in publi
c playgrounds in the City of Cardiff, UK. Methods We did a correlation
al study of 330 children aged between 0 and 14 years. All children wer
e hurt when playing in playgrounds in Cardiff and presented to the Acc
ident and Emergency Department in Cardiff Royal Infirmary during summe
r (April to September) 1992 and 1993, and the whole of 1994. We studie
d the children's hospital records to establish the type of injury and
interviewed their parents to find out the playground and type of equip
ment involved. The main outcome measures were the number of children i
njured whilst playing, and injury rates per observed number of childre
n on different surfaces, types, and heights of equipment. Findings Chi
ldren sustained significantly more injuries in playgrounds with concre
te surfaces than in those with bark or rubberised surfaces (p<0.001).
Playgrounds with rubber surfaces had the lowest rare of injury, with a
risk half that of bark and a fifth of that of concrete. Bark surfaces
were not significantly more protective against arm fractures than con
crete. Most injuries were equipment related. Injury risk due to falls
from monkey bars (suspended parallel bars or rings between which child
ren swing) was twice that far climbing-frames and seven times that for
swings or slides. The height of the equipment correlated significantl
y with the number of fractures (p=0.005) from falls. Interpretation Ru
bber or bath surfacing is associated with a low rate of injuries and w
e support their use in all public playgrounds. Bark alone is insuffici
ent, however, to prevent all injuries, particularly arm fractures. Rub
berised impact-absorbing surfaces are safer than bark. We believe that
playing on monkey bars increases the risk of injury in playgrounds an
d that they should generally not he installed. Safety standards should
be based on physical and epidemiological data. Our data suggest that
the proposed raising of the maximum fall height from 2.5 m to 3.0 m in
Europe is worrying.