A response to Zhou et al., regarding thiamine supplementation in altered mental status.

We read with interest the article by Zhou et al. entitled, ‘Thiamine supplementation in hospitalized patients with altered mental status: does it help?’ We applaud the authors for tackling the questions surrounding hospital use of thiamine and clinical outcomes, however, we have some concerns regarding their methodology and conclusions. This retrospective study examined patients who had altered their mental status and received thiamine. Because of the nonrandomized design, numerous potential confounders such as the varying etiologies of altered mental status, varying indications for treatment, polypharmacy, and so on, make comparisons questionable. Furthermore, there was a thiamine first group and a glucose first group. Again, because of design, we have no way of knowing if the patients who received glucose first were more sick, as acute hypoglycemia often presents more dramatically and emergently than thiamine deficiency. We respectfully disagree with the claim that the authors make about their patient population, saying that they believe it ‘will in general reflect those at all tertiary care centers’ – because we simply have no way of knowing this. Also, there is no insight as to how much thiamine patients received outside of ‘at least one dose.’ This is problematic because although there are no definitive-dose finding studies, respected resources recommend initial thiamine regimens for patients with Wernicke encephalopathy to be 500 mg IV (three times a day for up to 5 days) [1,2]. Furthermore, there is the issue of how thiamine was administered in this study. We caution the authors about drawing conclusions from the data in which patients received enteral thiamine. This is because the oral administration of thiamine is unpredictably absorbed and is not recommended in patients who have altered mental status due to thiamine deficiency [3]. Lastly, we take issue with the final line of the abstract. Saying that thiamine must be administered prior to glucose in altered mental status patients perpetuates a myth in medicine [4,5]. Thiamine uptake into cells and subsequent enzyme activation is slower than the uptake of glucose, so one could argue that it makes no difference whether you pretreat or post-treat with thiamine [6]. We again applaud the authors for undertaking this study and agree with their conclusion that there is a need to recognize the risks of Wernicke encephalopathy and provide supplemental thiamine.