Strengths and limitations
The strengths of our study include the use of a large database derived from routine, prospective data entry in a national programme, and the application of a customised standard which adjusts for constitutional variation and thereby better reflects the fetus’ growth potential25 . In a proportion of cases the early pregnancy risk assessment was not recorded, as the data field was not contained in the electronic maternity record. However, the similar stillbirth rate in this ‘missing’ group suggests that this did not introduce a systematic bias.
We had no ultrasound measurements recorded in this database and therefore the trajectory of growth, expressed as velocity or growth rate, could not be calculated. While serial scans can provide useful information about growth rate 26, it has also been argued that it adds little to the effectiveness of the last scan to predict adverse outcome 27–30. The use of birthweight-based centiles could be considered a weakness of the study, as the objective was to predict risk of stillbirth based on fetal rather than neonatal weight. However, it can also be considered a strength, as 1. it allowed comparison between subgroups (the low risk group having had no clinical indication to undergo ultrasound investigations); 2. There was no need to adjust case by case for the respective scan-to-delivery interval; and 3. it is appropriate to infer fetal weight from newborn weight as gold standard, as estimated fetal weight is prone to systematic and random measurement error31. It was possible to apply this method as the customised GROW chart provides the same, contiguous standard for fetal and neonatal weight.