Sir,The article by Tripathy and Mishra.[1] on comparison of two screening tools for nutritional assessment in the critically ill elderly patient made for a very interesting read. Nutritional support in critically ill remains a topic of debate, and therefore, it is extremely important to stratify your patient for nutritional need. Nutritional screening in Intensive Care Unit (ICU) requires an understanding of two important points that are nutritional status at admission and severity of illness, and likely duration of organ support in the form of mechanical ventilation, vasopressor therapy. The severity of illness defines the catabolic stress and therefore, can be used to anticipate the expected loss of lean body mass.Nutritional scoring used in hospitalized patients cannot be generalized to ICU patients due to various reasons:Inappropriate assessment of weight due to fluid shiftsUsing body mass index as an indicator may mask the patients with sarcopenic obesity[2]Inclusion of catabolic stress in screening tool is absentGeriatric nutritional risk index (GNRI) uses albumin for assessment that is an acute phase reactant and should not be used in ICU patients.The methodology does not specify the method by which the ulnar length and knee length were measured. Who was the person who did the measurements? There is a possibility of bias if the person involved had a part in the data analysis and interpretation. The gold standard against which the two screening tools were compared is not described clearly. About 52.3% of patients were found to be “at risk” of malnutrition by this preset definition. In how many patients was the history of weight loss over last 6 months recorded. The gold standard should be able to detect 100% cases of at risk to malnutrition only then the comparison can be done.[34] The demographic of the patients is not described in details. It appears 70% of the patients had weight recorded, which is not reflected by the severity of illness, which was having a mean of 19. While comparing the two screening methods, the mortality is not described. How many patients died during the study period is not known. How was the patient classification according to GNRI? Did this classification have relation with mortality? The survival curve shows four patients who died. It is described for only Malnutrition Universal Screening Tool (MUST) no data for GNRI is given. We do not think conclusions should be drawn from the significant log rank when a number of events are so less. MUST/GNRI is for hospitalized patients using such tool for critically ill patients may oversimplify the assessment of nutritional status and we may miss the “high risk” or may include “low risk” patients, and over treat.