Reply

Abstract

We thank Drs Salomon and Bernard for their interest in our paper. In fetuses with a short femur, the outcome was primarily dependent on the coexistence of other abnormalities. So, while the risk of an underlying skeletal dysplasia, genetic syndrome or chromosomal abnormality was high in fetuses with non-isolated short femur, no such outcomes were found in fetuses in which the short femur was isolated (defined as bilateral and symmetrical shortening of the femur in the absence of associated fetal abnormalities). We would therefore like to qualify their statement that ‘One should not fear such a high incidence of abnormalities when finding an isolated short femur in a first-trimester screened woman.’ In such cases with isolated short femur we showed that in about half the cases this will remain unexplained (defined as cases in which intrauterine growth restriction (IUGR) did not develop, or no postnatal diagnosis of skeletal dysplasia or genetic abnormality was made), in about 10% of the cases subsequent re-measurement will find the femur length to be within the normal range, while in the remainder (about 40%) severe IUGR will develop (defined in our study as fetal abdominal circumference below the 5th centile with abnormal fetal umbilical artery Doppler requiring delivery before 37 weeks). Uterine artery Doppler was a useful tool in predicting which pregnancies were at particular risk of subsequent severe IUGR and associated pre-eclampsia and abruption. We agree that the referral population was likely to be biased to some degree, and in the discussion we highlight this: ‘One of the limitations of this study is that there may have been a degree of referral bias. This, in combination with the policy of first-trimester screening for chromosomal abnormalities, may change the relative proportions of pregnancy outcomes. However, the outcomes in each group will be similar…’. In other words, the proportion of isolated to non-isolated short femur may be higher, but the relative proportions of outcomes within these two groups should be similar to those that we report; this is of particular relevance in the isolated group. On the issue of normal ranges, we regret that Salomon and Bernard felt that these were discussed inadequately, and the reference for these is given (reference 7). Some light research would have shown that the population from which these normal ranges were constructed was from the same geographical area as that of our hospital (the two main hospitals are fewer than 10 km apart and have overlapping catchment areas2). We therefore believe that the reference charts used do fit our practice. We accept the importance of providing locally relevant normal ranges. However, one could question from how narrow a geographical area such ranges need to come to be sufficiently accurate, and whether using ranges derived from a very narrow segment of the population can be generalized sufficiently for use. Regarding identification of cases, as we explain in the methods (Page 508): ‘we performed a search of our computerized database in order to identify all cases of short femur diagnosed between 18 and 24 weeks of gestation during the study period’, and this is what was done. Given the likely appropriate charts, it is difficult for us to hypothesize why the number of pregnancies with a femur length < 5th centile was lower than expected. One possible explanation is that sonographers with ‘real time’ charts (i.e. when measurements are plotted on a chart at the time of the scan, as is the case in the referring units) may be more likely to re-measure and ‘normalize’ values that are borderline, when the finding is isolated. As our population was screened previously by first-trimester nuchal translucency measurement it is also likely that the most severe cases associated with chromosomal abnormalities or multiple abnormalities had undergone pregnancy termination before the time of the anomaly scan. Finally, we were impressed with the ability of the authors to back-calculate the number of scans needed, and were amused to find that we may be doing fractions of scans! A. T. Papageorghiou*, N. Fratelli*, A. Bhide*, B. Thilaganathan*, * St George's Hospital, Fetal Medicine Unit, Lanesborough Wing, Blackshaw Road, London SW17 0QT, UK