Aims To present results from the national neonatal audit, and to review issues with data collection, interpretation of results, and the need for change.

Methods A national web based audit of 10 specific measures of neonatal care.

Results Since 2006 the number of neonatal units contributing data to National Neonatal Audit Programme (NNAP) has risen to 166, 93% of the 178 units in England. For those babies upon whom data were received, the rates (ie, percentage positive answers) have remained relatively constant at approximately 85% for temperature and 75% for BP measurement in the first hour of babies <28 weeks, whereas surfactant administration has fallen from a quarterly peak of 88% to 76% (difference=12%, 95% CI 6.4 to 17.6). The recorded rate for antenatal steroid administration at 24–34 weeks gestation is 70%. 50% of parents are recorded as being seen by a senior staff member within 24 h of admission. Data recording remains incomplete for all questions, varying from >95% completion for data on antenatal steroids to <60% for that on senior staff seeing the parents. There is a huge variation in the percentage of babies >37 weeks being admitted to NNUs: in the last quarter of 2008 level 1 units admitted 2.5 to 10%, level 2 units 1.9 to 13.3% and level 3 units 1.2 to 11.5% of such babies. The collection of data on breast milk feeding has changed since inception of the audit: 78% of all babies on NNUs get their mother's own milk. A pilot study on the 2-year health status questionnaire yielded a low return. Some units have reported changes in practice as a result of these data.

Conclusion The NNAP is progressing and data collection is improving, but it needs time, commitment and funding to become more complete and accurate. Some questions need reviewing. Data-based quality improvement has begun as shown by the descriptions of change received from participating units.