Data collection from very low birthweight infants in a geographical region: methods, costs, and trends in mortality, admission rates, and resource utilisation over a five-year period

Early Hum Dev. 2006 Feb;82(2):117-24. doi: 10.1016/j.earlhumdev.2005.10.019. Epub 2006 Jan 18.


Aims: 1. To determine the survival and morbidity of infants at discharge with a birthweight of less than 1500 g in the geographically defined population of East Anglia. 2. To demonstrate a cost-effective method of regional data collection. 3. To determine whether there were any changes in the demand for neonatal care.

Study design and subjects: A prospective cohort analysis using a single database to collect data on 1244 very low birthweight infants from eight neonatal units in one Region from 1993 to 1997.

Results: Estimated ascertainment of VLBW infants to the study was 96%. Over the 5 years survival rates were stable (75-79%). 52% of deaths in infants admitted for neonatal care occurred on day 1, with just 15% of deaths occurring after 28 days of life. Mortality risk significantly decreased with increasing gestational age at birth. Compared to 22-25-week old infants, the mortality risk decreased by 65% for 26-27-week old infants (OR 0.35 95% CI (0.21, 0.59)) and by 92% for 32-39-week old infants (OR 0.08 95% CI (0.03, 0.21)) with intermediate odds ratios of 0.22 (0.12, 0.42) and 0.13 (0.06, 0.28) for the 28-29 and 30-39 weeks gestation, respectively. Higher birthweight, after adjusting for gestation also decreased the mortality risk (OR 0.78 per 100 g difference, 95% CI (0.71, 0.86)). No change was seen in the number of extremely preterm infants admitted for intensive care or resource utilisation, although a significant increase was seen in the number of infants dying in delivery rooms. There was a reduction in the reported incidence of pulmonary interstitial emphysema (10-4%) but no change in the number of ventilation days or the rate of chronic lung disease. The mean maternal age increased from 27.7 years to 28.9 years during the study. Maternal steroid administration increased (30% to 59%) and was associated with a decreased risk of mortality (OR 0.44, 95% CI: 0.31-0.62).

Conclusions: It is possible to collect useful data from the neonatal period at a reasonable cost from a geographically defined population. This information was used for informing clinicians, counselling parents and in the era of managed clinical networks will be useful in guiding the provision of effective health care resources.

Publication types

  • Research Support, Non-U.S. Gov't

MeSH terms

  • Cohort Studies
  • Data Collection
  • Gestational Age
  • Humans
  • Infant Mortality / trends*
  • Infant, Newborn
  • Infant, Very Low Birth Weight*
  • Patient Admission / statistics & numerical data*
  • Utilization Review*