Givens et al1 assessed associations of antibiotics with 2 important outcomes, survival and comfort, in patients with advanced dementia and pneumonia. Attention to end-of-life care in patients with dementia is important and such prospective studies are highly needed, but these results may be open to other interpretations.
Control for confounding of (non)antibiotic treatment with illness severity was inadequate, lacking propensity control, control by validated risk scores, and physician's prognosis. Another issue is who will benefit from antibiotics. Fluid intake was ignored but appears crucial in modifying antibiotic-survival associations.2 Moreover, life or dying may be prolonged by only a few days at least as often as by months or years (as shown in the figures in the study by Givens et al1 and new work2). Of note, we also found that, compared with oral antibiotics, parenteral antibiotics hardly improved survival.3