Blood samples in neonates are generally obtained by lancing the heel. Local anesthetics have not been shown to reduce the pain that accompanies this procedure. Larsson and colleagues conducted a randomized trial to determine if venipuncture of the dorsal aspect of the hand was more efficient and less painful than heel lancing.

A total of 120 healthy, full-term infants were included in the study and were randomly assigned to have a blood sample drawn by heel lancing with a large lancet or a small lancet or by venipuncture of the dorsum of the hand. The test area was warmed before the skin was punctured. For 60 seconds, blood was allowed to flow freely. If the proper sample amount had not been collected, squeezing or further skin puncture was done to ensure an adequate sample. The infants' facial reactions and crying were recorded and analyzed according to the Neonatal Facial Coding System, which uses a scoring range of zero to 600, with higher scores indicating more distress. The time needed to collect an adequate sample was also recorded.

There were 50 infants in the venipuncture group, 50 in the small lancet group and 20 in the large lancet group. An additional 11 neonates were included in a sham heel stick group. Distress was measured 15 seconds after the first skin puncture. Infants in the venipuncture group had a median score of 247, which was significantly lower than scores in the lancet groups. During manipulation to obtain the blood sample, infants in the venipuncture group again showed significantly less discomfort than did infants in the lancet groups. No differences in distress were found between the sham heel stick group and the venipuncture group during manipulation to obtain the blood sample. Most infants (86 percent) in the venipuncture group required only one puncture to obtain an adequate sample, compared with only 40 percent of the large lancet group and 19 percent of the small lancet group. Venipuncture took significantly less time to provide an adequate sample (median: 191 seconds) compared with the large lancet (median: 279 seconds) and the small lancet (median: 419 seconds).

The authors conclude that venipuncture in the dorsum of a neonate's hand is less painful and more effective than lancing the infant's heel with either a small or a large lancet.