Rozzini et al provide important outcome data, which adds to the debate on appropriate treatment of intercurrent disease in patients with severe dementia. Their data are consistent with data from the United States; Morrison and Siu1 also found that despite hospitalization and aggressive treatment, pneumonia is often, but not always, fatal in patients with end-stage dementia (6-month mortality: 53%, compared with 13% for cognitively intact patients). Our Italian colleagues raise 2 issues we would like to address. First, could Dutch physicians have saved additional patients with antibiotic treatment? This issue is raised by the 3 (20%) of 15 survivors with "end-stage dementia" found by Rozzini et al. There could be debate about the definition of end-stage dementia; however, the bigger issue is how much difference antibiotic treatment makes. In a 1990 report, Fabiszewski and colleagues2 found that nursing home residents with the most severe dementia had a similar probability of long-term survival from a febrile illness whether treated with antibiotics or palliatively. However, in our larger study on pneumonia, we found much higher mortality among residents with dementia in whom antibiotic treatment was withheld.3 Likely, this was both due to nontreatment and to higher frailty compared with treated patients (Figure 1). Therefore, antibiotics are likely effective in curing pneumonia in some patients with end-stage dementia.
van der Steen JT, Ooms ME, Adèr HJ, Ribbe MW, van der Wal G. Medical Treatment of Acute Illnesses in End-Stage Dementia—Reply. Arch Intern Med. 2003;163(4):497–498. doi:10.1001/archinte.163.4.496