OS05.4.A Do neurocognitive deficits explain the differences between brain tumour patients and their proxies assessing the patient’s I-ADL?
Q Oort,L Dirven,S Sikkes,N Aaronson,F Boele,C Brannan,J Egeter,R Grant,M Klein,I Lips,Y Narita,H Sato,M Sztankay,G Stockhammer,A Talacchi,B Uitdehaag,J Reijneveld,M Taphoorn
DOI: https://doi.org/10.1093/neuonc/noab180.021
2021-09-01
Abstract:Abstract BACKGROUND Neurocognitive deficits are common among brain tumour patients, and may impact on patient awareness of deficits in instrumental activities in daily life (IADL). This study aimed to examine differences between patient-reported and proxy-reported assessments of the patient’s performance of IADL, and whether the level of (dis)agreement is associated with neurocognitive deficits. MATERIAL AND METHODS A phase III EORTC questionnaire measuring IADL in brain tumour patients (EORTC IADL-BN32) and six neurocognitive test measures were administered as part of a larger multicentre international study designed to develop a brain tumour specific IADL questionnaire. Bland-Altman plots and Mann-Whitney U tests were used to evaluated patient- and proxy-reported IADL on a group level. Subsequently, Mann-Whitney U tests were performed to compare patient-proxy difference scores (patient IADL score - their proxy IADL score) between patients who were considered clearly neurocognitively impaired (≥2 neurocognitive test measures; ≤2.0 SD below healthy controls) and patients who were not. Furthermore, multinomial logistic regression analyses were performed to examined which sociodemgraphic, clinical, and particularly neurocognitive variables were independently associated with patients and proxies differing in their evaluation of patient’s IADL. RESULTS Patients (N=81) and proxies (N=81), on group level, did not significantly differ on either the IADL individual item or scale scores. However, significant differences were found on patient-proxy difference scores between patients who were (N=37) and were not (N=44) considered clearly neurocognitively impaired for 10/32 individual items and one of the scales (i.e. Scale 4: Administrative tasks), all showing that the proxies of clearly neurocognitively impaired patients reported more problems relative to the patients themselves, compared to proxies of patients not clearly neurocognitively impaired. Furthermore, for each scale, a neurocognitive variable, either impaired information processing speed, cognitive flexibility, verbal fluency or the number of neurocognitive test measures impaired, was found to be independently associated with proxies reporting more problems. For 4/5 scales, a clinical variable was additionally independently associated with proxies reporting more problems. Only one variable was independently associated with patient reporting more problems, namely being in active treatment was found to be associated with patients reporting more problems on Scale 4: Administrative tasks. CONCLUSION Results imply a consistent trend of clearly neurocognitively impaired patients underreporting problems with IADL compared to their proxies. It would therefore be advised to administer both the patient- and proxy-version of the EORTC IADL-BN32, particularly if neurocognitive deficits are presumed.
oncology,clinical neurology