Fig. 1.
Consort flow diagram. Of the 70 protocol violations in the RA arm, 10 infants had a full GA with no awake regional attempted, 37 had a full GA after complete block failure, and 23 infants had a partly successful block requiring a short period of GA or sedation. Participants who withdrew consent (n = 1) or were randomized after surgery (n = 2) were excluded from intention-to-treat analyses. GA = general anesthesia; RA = regional anesthesia.