Reply to “limitations of Conclusions of Systematic Review & Meta‐analysis Due to Exclusion of Groups Most at Risk”

Karin J. Neufeld,Jirong Yue,Thomas N. Robinson,Sharon K. Inouye,Dale M. Needham

DOI: https://doi.org/10.1111/jgs.14727

2017-01-01

Journal of the American Geriatrics Society

Abstract:To the Editor: We are delighted that our systematic review provoked interest and reflection regarding the routine use of antipsychotics for the prevention and treatment of delirium in hospitalized patients.1 Dr. Dissevelt raises several points in her letter that we respond to below.2 First, a sentence in the Discussion section, referring to the study by Vochteloo et al.3 is cited incorrectly. We are grateful to Dr. Dissevelt for highlighting this issue. We have notified the publisher of this required correction. Second, Dr. Dissevelt suggested that the study by Vochteloo et al.3 “solely appears to pull the overall results of the effectiveness of haloperidol prophylaxis on delirium prevention in postoperative patients in Figure 2A on page 6 into the nonsignificant zone”.2 This is not a correct statement as we will clarify herein. Given that the study by Vochteloo et al.3 was not a randomized controlled trial, its findings are considered to be at high risk of bias based on the Cochrane assessment tool used in our meta-analysis. Of the 7 studies included in this meta-analysis of patients receiving prophylactic antipsychotics in the perioperative period, a total of 4 were at high risk of bias. Due to the paucity of literature regarding antipsychotics and the prevention of postoperative delirium, we included all relevant studies, regardless of risk of bias evaluation, in the primary meta-analysis. However, we also conducted a sensitivity analysis of the primary results excluding the 4 of 7 studies that were considered to be at high risk of bias.4 This sensitivity analysis produced similar findings as the primary analysis: antipsychotics were not significantly associated with prevention of postoperative delirium (Table 1).5-7 Hence, the study by Vochteloo et al.3 did not solely create findings that were non-significant. Future research, with rigorous study design, adequate sample size, and less heterogeneous patient populations are needed to provide further evidence in this area. Finally, Dr. Dissevelt raises a question regarding the exclusion of dementia, stroke, neurosurgery, and trauma in patient populations, as well as patients in nursing homes and other non-hospital settings. We agree that these patients are at high risk for developing delirium and are worthy of further evaluation. We did include critically ill patients in our meta-analysis as one group at very high risk of delirium. However, these patient populations mentioned by Dr. Dissevelt were outside of the scope of this systematic review and meta-analysis. As indicated in the title of our publication, the focus of this systematic review and meta-analysis was on hospitalized adults. This research was conducted in association with the American Geriatrics Society's clinical practice guideline on postoperative delirium,8, 9 which had a specific mandate to address a knowledge gap for clinicians who care for patients in perioperative acute care hospital settings. Moreover, acute care in-patients with specific particular pre-existing neurological conditions were excluded because of concerns regarding validity of delirium outcome measurement in patients with significantly impaired arousal, cognition, or communication ability. Notably, most studies that we reviewed already excluded patients with such comorbidities in their study design. Hence, despite their importance, there may be a paucity of rigorous studies in these patient populations. We agree with Dr. Dissevelt that these important patient groups need much more study and perhaps future systematic review and meta-analysis in order to improve our knowledge about delirium in these vulnerable populations. Conflict of Interest: Authors have no conflicts of interest to declare. Financial Disclosure: The development of the original paper was supported in part by the American Geriatrics Society Geriatrics-for-Specialists Initiative (AGS-GSI) through an unrestricted grant from the John A. Hartford Foundation (Grant 2009-0079). Dr. Inouye's time is supported in part by Grants No. K07AG041835 and P01AG031720 from the National Institute on Aging and by the Milton and Shirley F. Levy Family Chair. Author Contributions: All who have contributed significantly to this work are listed as authors. Sponsor's Role: The sponsor had no role in the content of this response.