Abstract
To the Editor,
We thank Dr Ismail for his interest in our paper. Indeed we agree with his central point, namely that the finding of a ‘normal’ serum magnesium concentration does not exclude magnesium deficiency. 1 His warning not to assume that it does is apposite. As the title of our paper indicated, we wished to characterize the process of adding on laboratory tests, in terms of efficiency and effectiveness 2 ; we did not examine the diagnoses that were or were not made as a result of adding tests. This is a very important area for further examination, but was beyond the scope of our study.
Dr Ismail's letter highlights just one example of the need to examine the veracity and significance of the diagnoses made by reflex and reflective testing. Figure 1 in our paper highlights another example. 2 Part (b) of that figure plots the efficiency and effectiveness of reflex diagnosis of hypothyroidism across the thyroid-stimulating hormone (TSH) range. In the unequivocally hypothyroid part of the TSH range, the likelihood is that the biochemical diagnosis of primary hypothyroidism made by reflex addition of free thyroxine (i.e. the finding of a low free thyroxine [fT4] triggered by a high TSH) will be correct. However, this assumption does not hold for TSH results that are in the euthyroid range – the finding of a low fT4 in conjunction with a normal TSH does not permit the diagnosis of hypothyroidism to be made with confidence; the results may be explained in other ways (e.g. non-thyroidal illness). Similar comments, modified by degree, apply to TSH results that are close to the euthyroid range. In other words, the plot of effectiveness (number of diagnoses made) against triggering analyte needs to be interpreted with caution, particularly in those parts of the range where reflex testing would not normally be applied. The rise in numbers needed to diagnose (fall in the efficiency of reflex addition of fT4) seen in the same parts of the range is unlikely to be coincidental and highlights the need for caution in interpretation.
DECLARATIONS