Two-year neurological outcomes of very low birth weight infants at Kenyatta National Hospital

Abstract

Background: Determination of the long-term outcomes for high-risk newborns such as the very low birth weight should be standard practice for facilities from which such infants graduate. Centres in the developed countries have been consistently making such reviews in a systematic manner since the early 1980s. The outcome measures employed in these evaluations include quantification of neurological disabilities, post-natal growth monitoring and estimation of postdischarge morbidity and post-discharge mortality rates. Such information can be used as tools for assessing the performance of individual newborn units as well as comparison of more than one facility. The results of medical audits also provide an opportunity for evaluating the long-term effects of specific interventions in the newborn period. This study was the first of the kind to be undertaken at the Kenyatta National Hospital, a big center with more than 7000 deliveries/year. Objectives: To estimate the prevalence of neurological disabilities, postdischarge morbidity and mortality at the age of two years for a cohort of very low birth weight infants discharged from Kenyatta National Hospital and identify some of the factors associated with these outcomes. Design: Longitudinal cohort study with nested case control components. 15 Methods: One hundred and seventy five very low birth weight infants were recruited during the calendar year 2002 and scheduled for follow-up until the completed age of two years. The follow-up was terminated at the age of two years or death whichever occurred earlier. The initial documentation included recording of intra-uterine growth status, gender, neonatal morbidity and neonatal growth. The measures of neonatal growth were as follows; weight in grams/kilogram of birth weight/day and length and head circumference in centimeters/week. The type of milk consumed during the first month of life was also recorded. During follow-up, the infants' growth was monitored, rehospitalization and mortality records obtained and detailed neurological assessments performed at the age of two years corrected for gestation. Cognitive function was determined using Egan FD's developmental examination of infants and preschool children method while functional disability employed the tool designed by Saigal, Rosenbaum, Stoskopf and Milner. Cerebr also assessed and specific lesions described. Results: Out of the 175 infants recruited, 10 (5.6%) were lost to follow-up while 45 (25.7%) died leaving 120 (68.6%) survivors available for the neuroloqical assessment. The male: female ratio for the whole cohort of 175 was 2:3, 64 (36.6%) had been intra-uterine growth retarded while neonatal illnesses were reported in 109 (62.3%). During the first month of life 78 (44.6%), 54(30.8%) and 33 (18.9%) of the infants were fed on exclusive breast milk, breast milk supplemented with preterm formula and exclusive preterm formula respectively. 16 The mean neonatal weight gain for the 175 neonatal survivors was 13.5.:!:.3.9 g/kg/day, length 0.34.:!:.0.11cm/week and head circumference 0.32.:!:.0.71 cm/week. By the time the infants reached the age of 40weeks from conception, . term 33 (18.9%), 37 (21.1 %) and 48 (28%) had attained or passed the 3rd percentile of the expected weight, length and head circumference respectively. The factors associated with better neonatal growth and growth attained at term included feeding on milk with increasing preterm formula content, P<0.001 and absence of neonatal illness, P<0.001. Infants who were appropriate for gestational age at birth showed better catch-up growth at term compared to those born intra-uterine growth retarded, P<0.001. Growth faultering in weight, head circumference and length at two years were found in 57.5, 62.5 and 60.8% of the infants respectively. Eighty-nine (53.8%),95% CI; 25.6-73.2 of the 165 who completed the study were re-hospitalized during the follow-up period. Rehospitalization was associated with; neonatal illness P=0.015, exclusive use of breast milk P=0.001, discharge weight less than two kilos P=0.041, neonatal weight gain <15grams/kilo/day P=0.001, growth faultering at the age of two years P=0.005 and the presence of functional disability P=0.001. The post-di charge mortality for the 165 infants with complete information was 27.3%,· 95% (;1 9.8- 43.6. The factors associated with post-discharge death were; neonatal IIness P=0.001, neonatal weight gain less than 15g/kg/d P=0.004, discharge weight less than two kilograms P=0.001 and history of re-hospitalization P=0,002. 17 Among the 120 infants who survived to the age of two years 14 (11.7%) 95% CI 6.7 - 17.1 had Cerebral Palsy, 11 (9.2%) 95% CI 4.8 - 16.9 were delayed in Cognitive Scores while 32 (26.7%) 95% CI, 9.3-381 had Functional Disability. The factors associated with functional disability included; neonatal illnesses P=0.005, exclusive use of breast milk during the newborn period P=0.020, neonatal weight gain less than 15g/kg/d P=0.014, growth faultering at two years P=0.019 and history of re-hospitalization P=0.001. The 88 survivors who did not have any functional disability were evaluated to determine the factors associated with low cognitive scores. There were significantly lower cognitive scores for those with neonatal weight gain less than 15g/kg/day P=0.001, neonatal illness P=0.005, increasing content of breast milk in feeds during the newborn period P=0.001 and weight less than the 3rd percentile at term P=0.001 and at two years P=0.005. The 88 infants who were alive and free of neurological disability at two years constituted 25.6% of the birth cohort of 344 and 50.3% of the 175 neonatal survivors. These are hence the estimates of intact neurological survi al rates for the birth cohort and the neonatal survivors of the institution respectively. Conclusions: The mean neonatal growth indices were all less than levels traditionally considered adequate, 15g/kg/d for weight and 0.5cm/week for both length and head circumference. Post-natal growth was poor with 80.6, 69.6 and 57.5% of the infants weighing less than the 3rd percentile of the expected at term, one and two years of age respectively. 18 The prevalence of cerebral palsy (11.7%), cognitive delay (9.2%) and functional disability (26.7%) were higher than expected. The predictors of neurological abnormalities at two years of age were; presence of neonatal illness, poor neonatal and post-discharge growth in weight. Post-discharge re-hospitalization (53.9%) and mortality (27.3%) were also higher than expected. Neonatal weight gain less than 15g/kg/d, history of neonatal illness and discharge weight below two kilograms were the consistent predictors of both re-hospitalization and postdischarge mortality. Recommendations: Neonatal growth faultering was consistently associated with increased re-hospitalization, mortality and neurological disabilities. Since feeding with nutrient enriched preterm formula was also consistently associated with better neonatal growth, the preferential use of such milk or fortification of mothers' milk may be beneficial in improving these outcomes. Very L.ow Birth Weight infants should be discharged after attaining the weight of two kilograms, as this was associated with lower rates of re-hospitalization and post-discharge mortality.