Preterm infants undergoing heel blood sampling were randomly allocated
to specimen collection by heel puncture (Autolet II Clinisafe) or inc
ision (Tenderfoot 'preemie'). A total of 187 procedures was observed i
n 47 infants. No significant difference was found in the infants' beha
vioural response, increase in heart rate, or in the the frequency of s
pecimen haemolysis. Collection times for small to medium sized samples
were similar, but for large samples (>1 ml) the Tenderfoot method was
superior and fewer repeat procedures were necessary. Tenderfoot has t
he added advantage of improved safety, but it is too costly for routin
e use.