Efficacy and Safety of Citalopram Compared With Atypical Antipsychotics on Agitation in Nursing Home Residents With Alzheimer Dementia

We are delighted to know that our article has captured the attention of readers. Regarding the findings of our previous study, Professor Naharci points out that unrecognized delirium could have had a relevant confounding effect and that the differential role of specific antidementia drugs should have been analyzed in detail. Delirium is a common geriatric syndrome with the highest incidence in acutely ill hospitalized older patients. Professor Naharci cited a retrospective study by Dosa et al reporting a very high occurrence of either full-spectrum or subsyndromal delirium among long-stay residents. The study population consisted of a large sample of long-stay residents who returned to their nursing homes after acute hospitalization. In these patients, the short-term incidence of any type of delirium was expected to be higher than in their “clinically stable” counterpart, and often delirium or still present when older individuals are discharged from acute hospitals. Thus, results by Dosa et al are not generalizable to the whole institutionalized older population. The inclusion criteria of our study were the presence of clinically relevant agitation for which a physician determined that medication was appropriate, in the presence of a diagnosis of probable Alzheimer dementia. Decision to treat was made by 2 experienced geriatricians with expertise in dementia treatment. Good clinical practice in such context requires that a long-term pharmacologic treatment for neuropsychiatric symptoms (NPS) in dementia must be initiated after a careful assessment of the patient. To define whether agitation or other NPS are secondary to dementia or to potentially reversible conditions, such as delirium, it must be investigated to determine whether the symptoms persist over time, whether there are fluctuations over the day, and whether there is concomitant acute impairment in attention and cognition. A differential diagnosis can be easily carried out in most cases. However, as dementia-associated NPS might have a fluctuating pattern irrespective of a concomitant delirium, in particular cases, it can be very difficult to determine the extent to which NPS in demented patients are attributable to dementia itself or to a superimposed long-standing delirium. We are confident in asserting that none of the study participants had delirium at baseline. In addition, retrospective analyses of daily medical and nursing notes, the result of a continuative surveillance of the patients, allow us to assume that most of patients remained delirium-free during the observational period. However, we cannot exclude that some participants may have experienced delirium, especially hypoactive forms, and we do not know whether participants who required emergency department transfer have developed delirium during hospitalization. Because we believe that the proportion of participants who may have suffered from delirium, if any, has been irrelevant, it is our opinion that a systematic assessment of delirium cases would have had little impact on the outcomes of the study. However, we admit that the lack of a standardized delirium assessment represents a limitation of our study, at least from a formal point of view. We do agree with Professor Naharci that assessment of delirium is important in the assessment of drugs for NPS in dementia, especially because almost all psychotropic drugs may themselves be a cause of delirium. Professor Naharci also points out that among antidementia drugs, memantine “is a preferred drug in agitation management in Alzheimer dementia.” However, there is no lack of eminent reports that use of cholinesterase inhibitors is an effective and safe option for agitation control in dementia, and the current level of evidence is not strong enough to retain that memantine should be preferred to cholinesterase inhibitors in this context. In this study, results were controlled for use of antidementia drugs (either memantine or cholinesterase inhibitors) not differentiating between them. It was done in this way because of the small sample size and the low proportion of those prescribed with antidementia drugs, that is 25.3%. In particular, 6 participants were receiving antidementia drugs in the citalopram group (2 memantine, 1 donepezil, 1 galantamine, and 2 rivastigmine), 7 in the quetiapine group (3 memantine, 3 donepezil, and 1 rivastigmine), and 6 in the olanzapine group (2 memantine, 2 donepezil, 1 galantamine, and 1 rivastigmine). We thank Professor Naharci for his valuable suggestions, and we trust that these will be taken into account in future studies.