Abstract 1488

Many VLBW (<1501 g) and Extremely low birth weight (<1001) infants have abnormalities in muscular tone, posture and delays in their gross motor and fine motor development during their first 18 months. early diagnosis of CP is very difficult. Objective: To identify early predictors of CP in VLBW infants. Method: Case control study of infants followed in our Premature Clinic from 1985 to 1995. Cases (CP): 140 VLBW infants with the diagnosis of CP. Controls (CO): 140 VLBW infants who did not have CP, matched for sex, GA +/-2 w, BW +/- 350 g, year of birth, race and multiple gestation. Data: 77 variables of prenatal, intrapartum and infant's perinatal and neonatal course. Statistics: Chi-square, t-test, stepwise logistic regression analysis. Results: CP had a mean GA of 28 wks (SD 0.23), range from (R) 22 - 35 wks and mean BW of 944 g (23.14), R 420 - 1450g, of whom 56% were males. CO had a mean GA of 28 wks (0.19), R 22-34 wks and mean BW of 1008g (19.97), R 520-1480 g. Maternal and neonatal variables showing a moderate (p<0.2) to highly (p<0.001) significant association with CP were: mother's use of alcohol and drugs, urinary tract infection and spotting during pregnancy and PROM as indication for delivery, infant's need for cardiac massage during resuscitation, low 1 min Apgar, initial low PO2, need for higher O2 concentrations, hypotension, PIE, BPD, no Apnea, ICH > than grade 2, periventricular leukomalacia (PVL), need for VP shunt, BW < 10th %tile, and > 50th%tile, HC <10th and >50th %tile, longer days in NICU, in O2 and in respirator. The odds of having CP was 3.8 higher in infants who had ICH > grade 2, 4.8 in infants with PVL, 6.1 in infants with VP shunt, 2.8 if needed O2 concentrations >80, 3.8 if they were in NICU greater than 90 days, 2.7 if they had ROP and 2.4 if mother had PROM. Six risk factors were selected when significant variables were entered for a stepwise regression model: ICH >grade 2 (Max R square 0.11, p=0.001), NICU stay > 90 days (0.18, p<0.001), maternal drug use (0.23, P=0.009), PVL (0.27, p=0.014), no apnea (0.29, p=0.004) and BW < 50th %tile (0.32, p=0.011). This model explained 32% of the variation in the outcome. Conclusion: Early diagnosis of CP is very problematic and it seems to be multifactorial. In controlling for GA and BW we compared similar high risk medically fragile infants. Sicker infants are the ones at higher risk for CP.