According to Wright et al, taken in isolation, a mean cell haemoglobin (MCH) of <25 pg is more likely to predict a significant haematological response to a trial of iron replacement therapy than a mean cell volume (MCV) of <75 fl, on its own.¹ My own approach to identifying which of the two red blood cell indices, namely MCH and MCV, was the stronger predictor of iron deficiency was to evaluate the cut-off levels which yielded the optimum combination of sensitivity, specificity, and positive predictive value for unequivocal iron deficiency, the latter being defined as a serum ferritin of <10 μg/l. In a study comprising 365 adults characterised by an MCH of <26 pg and/or an MCV <80 fl,² 145 of whom proved to be unequivocally iron deficient,³ an MCH of <24 pg was identified as being the one associated with the optimum combination of sensitivity (74%), specificity (59%), and positive predictive value (80%) for this diagnosis. Correspondingly, an MCV <76 fl was the one associated with the optimum combination of sensitivity (65%), specificity (66%), and positive predictive value (55%).³ Fortuitously, in the ABC of clinical haematology, it is also an MCV of <76 fl which is utilised in what I would describe as “automated” screening for iron deficiency.⁴

However, what has not been addressed until very recently, is the issue, not only of the superiority of MCH in predicting a favourable response to a therapeutic trial of iron replacement therapy,¹ but also its robustness, relative to the MCV, under laboratory conditions of automated screening. According to one of the leading authorities on the subject, different counting systems yield “clinically significant different” estimates of the MCV, as shown by the monthly reports of the UK General Haematology NEQAS Scheme. In contrast, MCH yielded a “consistent equality of results reported by the different technologies within the UK NEQAS schemes”.⁵

These observations tend to support the suggestion made by the authors of the present study that, as opposed to the MCV, the MCH should be the preferred screening test for predicting a satisfactory haematological response to iron replacement therapy.¹