Objective: To determine whether maternal or fetal factors, other than
vacuum-assisted delivery, play a role in neonatal retinal hemorrhage,
and whether correlates are similar in retinal hemorrhage after spontan
eous vaginal delivery. Methods: A cross-section of assisted deliveries
at an urban hospital (n = 156) over 7 months were compared with conte
mporaneous spontaneous vaginal deliveries (n = 122). A subset of assis
ted deliveries (n = 87) was prospectively randomized to forceps or vac
uum-assisted delivery by sealed envelope. Maternal and neonatal biomet
ric data were collected, and Apgar scores, umbilical artery blood gas
analysis, and neonatal ophthalmologic evaluations were performed. Resu
lts: Moderate to severe retinal hemorrhage was found in 18% of spontan
eous, 13% of forceps, 28% of vacuum-assisted, and 50% of sequential va
cuum and forceps-assisted deliveries. Fetal distress (P < .008), vacuu
m-assisted delivery (P < .02), decreased birth weight for gestation (P
< .004), umbilical artery pH less than 7.20 (P < .004), and second st
age of labor less than 30 minutes (P < .05) were most closely associat
ed with increased degrees of retinal hemorrhage. Maternal parity, pree
clampsia, length of labor, and head circumference were not correlated
with retinal hemorrhage. Vacuum-assisted delivery among low birth weig
ht infants (P < .0001), short second stage of labor (P < .006), fetal
acidosis (P < .045), and sequential use of vacuum and forceps for assi
sted delivery (P < .005) formed a logistic model that correctly predic
ted 81% of moderate to severe retinal hemorrhage cases. Logistic analy
sis of the randomized assisted deliveries gave similar results. Conclu
sions: Maternal and fetal factors other than vacuum-assisted delivery
are significant correlates of moderate to severe retinal hemorrhage. V
acuum-assisted delivery among small for gestational age infants is clo
sely correlated with moderate to severe retinal hemorrhage.