Background: Neurocognitive impairment is frequently present in brain tumour patients and is therefore considered an important outcome in brain tumour research. To use neurocognitive outcomes (NCO) in clinical decision-making, neurocognitive evidence should be of sufficiently high quality. We aimed to investigate the level of neurocognitive functioning reporting in randomised controlled trials (RCTs) in brain tumour patients.
Methods: We conducted a systematic literature search in several databases up to August 2017. Of the selected relevant RCTs, the following data were retrieved: basic trial demographics and NCO characteristics, quality of NCO reporting and risk of bias. We also analysed studies that should impact clinical decision-making based on their quality of reporting.
Results: We identified 65 RCTs, of which NCO was the primary end-point in 14 (22%). Important methodological limitations were related to the documentation of statistical approaches for dealing with missing data and to discussing limitations and generalisability issues uniquely related to the NCO components. Risk of bias was high regarding blinding of personnel and incomplete outcome data. Twenty RCTs (31%), eight with NCO as primary end-point and 12 as secondary end-point, satisfied a sufficient number of criteria to be classified as 'high-quality' NCO evidence. Most of these studies did contribute to clinical decision-making.
Conclusion: Investigators involved in brain tumour research should give attention to methodological challenges related to NCO reporting as identified in this review, as 'high-quality' reporting of NCO evidence can be of value in clinical decision-making.
Keywords: Bias; Brain tumour; Cognition; Data reporting.
Copyright © 2018 Elsevier Ltd. All rights reserved.