This article is freely available to all

Article Abstract

Because this piece does not have an abstract, we have provided for your benefit the first 3 sentences of the full text.

To the Editor: We thank the Editor for the opportunity to respond to Dr Tan’s letter, which used our recently published meta-analysis on cholinesterase inhibitor (ChEI) discontinuation in patients with Alzheimer’s disease to highlight the limitations of statistical indicators of heterogeneity.

We accept Dr Tan’s view that improving statistical management of heterogeneity in meta-analyses with only a small number of studies may be an important area of future research. Indeed, we acknowledged the small number of studies, as well as the small sample sizes in the included studies, as limitations of our meta-analysis.

See letter by Tan and article by O’ Regan et al

This work may not be copied, distributed, displayed, published, reproduced, transmitted, modified, posted, sold, licensed, or used for commercial purposes. By downloading this file, you are agreeing to the publisher’s Terms & Conditions.

Dr Mazereeuw and Colleagues Reply

To the Editor: We thank the Editor for the opportunity to respond to Dr Tan’s letter, which used our recently published meta-analysis on cholinesterase inhibitor (ChEI) discontinuation in patients with Alzheimer’s disease1 to highlight the limitations of statistical indicators of heterogeneity.

We accept Dr Tan’s view that improving statistical management of heterogeneity in meta-analyses with only a small number of studies may be an important area of future research. Indeed, we acknowledged the small number of studies, as well as the small sample sizes in the included studies, as limitations of our meta-analysis. However, Dr Tan’s suggestion that we used inappropriate methodology and unquestioningly presented our findings is not well supported.

The Cochrane Handbook for Systematic Reviews of Interventions2 recommends that authors compare the effects of their intervention using both fixed- and random-effects models in order to identify the potential influence of small study effects. The Handbook suggests that if no differences in the intervention effect are detected between the fixed- and random-effects models, then authors may conclude that small study effects had little influence on the outcome. As such, we investigated the effect of ChEI discontinuation using both fixed- and random-effects models to determine which model was more appropriate to report in the article. Although we did not publish the results of the random-effects models in that article, we present them here to demonstrate the consistency of our findings.

Using a random-effects model, our analysis shows that patients in the ChEI discontinuation group experienced significantly greater cognitive decline than those remaining on ChEIs (standardized mean difference [SMD] = −0.29 [95% CI, −0.45 to −0.13] P < .001; heterogeneity: χ2 = 2.62, I2 = 0, P = .62). Similarly, using a random-effects model, the ChEI discontinuation group experienced significantly greater neuropsychiatric symptoms than those remaining on ChEIs (SMD = −0.32 [95% CI, −0.51 to −0.12] P = .001; heterogeneity: χ2 = 0.94, I2 = 0, P = .63). The results seen when using random-effects models are identical to those when using fixed-effects models, supporting our findings and reinforcing the lack of heterogeneity. This consistency is not at all surprising given the lack of heterogeneity reported in the article.

Contrary to Dr Tan’s claim that we did not follow through on our observation of heterogeneity (despite none being detected using either model), we did explore potential modifiers of the observed treatment effects and published those findings as part of our original article. Specifically, we showed that the effect of ChEI discontinuation on cognitive changes and neuropsychiatric symptoms was not influenced by the mean patient age, sex, previous duration of ChEI treatment, duration of follow-up in each study, or the baseline severity of dementia. As such, we fully explored potential sources of heterogeneity to increase the transparency of our findings in light of the variability in study designs.

In summary, we conducted our meta-analysis according to established guidelines, using established statistical tests. We investigated heterogeneity appropriately and reported our findings accordingly. Therefore, we suggest that Dr Tan’s criticism of our study is not well supported and brings unnecessary skepticism to our findings. Our methods were rigorous, and we achieved our goal of summarizing the available literature and bringing attention to the need for more studies in this area.

References

1. O’ Regan J, Lanct×´t KL, Mazereeuw G, et al. Cholinesterase inhibitor discontinuation in patients with Alzheimer’s disease: a meta-analysis of randomized controlled trials. J Clin Psychiatry. 2015;76(11):e1424-e1431. PubMed doi:10.4088/JCP.14r09237

2. Higgins JPT, Green S. Cochrane Handbook for Systematic Reviews of Interventions, Version 5.1.0. The Cochrane Collaboration; 2011. http://handbook.cochrane.org/. Updated March 2011.

Graham Mazereeuw, PhDa,b

Krista L. Lanct×´t, PhDa,b,c

Nathan Herrmann, MD, FRCPCa,c

[email protected]

aNeuropsychopharmacology Research Group, Hurvitz Brain Sciences Research Program, Sunnybrook Health Sciences Centre, Toronto, Ontario, Canada

bDepartment of Pharmacology and Toxicology, University of Toronto, Toronto, Ontario, Canada

cDepartment of Psychiatry, University of Toronto, Toronto, Ontario, Canada

Potential conflicts of interest: None.

Funding/support: None.

J Clin Psychiatry 2016;77(8):e1006

dx.doi.org/10.4088/JCP.16lr10666a

© Copyright 2016 Physicians Postgraduate Press, Inc.