Reply letter to “Nutritional risk in hospitalized elderly patients”

Sir,

In reply to the letter sent by Sánchez-Muñoz et al. in relation to our article “Nutritional risk in hospitalized elderly patients”, we would like to clarify some points.

As regards the methodological limitations pointed out: 1) Sample selection among patients on nutritional support is a known limitation inherent to study design and is recognised as such in the text (page 110, sixth paragraph). 2) The reason for the lack of correlation between risk indices and mean stay is reflected in our article in the same terms as in the Sánchez-Muñoz et al. letter (page 110, third paragraph), but this was not the final objective of the study, which was not designed for this purpose, as is also stated in the text. 3) We could not know whether patient weights were actual or estimated as this was a retrospective study. This was one of the main study limitations, as is also stated in the text. 4) Finally, ideal weight calculated using the Lorentz formula is the weight validated for GNRI, according to Bouillanne et al.¹ While it is true that the Lorentz formula is not adapted for the Spanish population, the study was designed to compare the original formula to NRI.

With regard to the value of the Mini-Nutritional Assessment (MNA) or nutritional risk indices (NRI/GNRI) to assess hospitalized elderly patients, it should be noted that: 1) The study was designed to assess whether or not GNRI is superior to NRI for predicting complications, rather than to assess whether MNA is superior or inferior to GNRI and NRI. 2) MNA was not fully evaluable in this study because it had only been performed in patients admitted to non-surgical floors (39.8% of patients). 3) In addition, among nutritional tools, MNA is a recognised nutritional status assessment method, NRS-2002 a recognised malnutrition screening system, and nutritional risk indices (NRI/GNRI) are used to predict the development of complications-related malnutrition (rather than to assess nutritional patient status itself). These measures are therefore different but complementary rather than mutually exclusive.

To sum up, because of the study design and the characteristics of these indices, at no time was it intended to state that NRI or GNRI are better than MNA for the detection of malnutrition or for nutritional status assessment. We only stated in our study that, based on the results found and taking its limitations into account, NRI may complement other measures, and so help to predict complications related to malnutrition. These tools require little time when the parameters needed to calculate them (which would have to be measured in any inpatient) are available, and have therefore no negative influence on the burden of care.