Impact of neurocognitive deficits on patient–proxy agreement regarding health-related quality of life in low-grade glioma patients

on behalf of the EORTC Quality of Life Department and Patient Reported Outcome and Behavioral Evidence (PROBE)

Research output: Contribution to journalArticleAcademicpeer-review

26 Citations (Scopus)


Purpose: Clinical trials in glioma patients with neurocognitive deficits face challenges due to lacking or unreliable patient self-reports on their health-related quality of life (HRQOL). Patient–proxy data could help solve this issue. We determined whether patient–proxy concordance levels were affected by patients’ neurocognitive functioning. Methods: Patient and patient-by-proxy HRQOL ratings were assessed via SF-36 and EORTC QLQ-BN20, respectively, in 246 patients. Data on neurocognitive functioning were collected on a subgroup of 195 patients. Patient–proxy agreement was measured using the Bland–Altman limit of agreement, the mean difference, the concordance correlation coefficient (CCC), and the percentage difference (PD, ±0, 5, or 10 points). We defined patients to be cognitively impaired (n = 66) or cognitively intact (n = 129) based on their neurocognitive performance. Results: Patients rated their physical function and general health to be better than their proxies did, while at the same time, patients reported more visual disorders, communication deficits, itchy skin, and problems with bladder control. The cognitively impaired subgroup reported poorer physical functioning, more visual disorders, headaches, itchy skin, and issues with bladder control. In the cognitively intact group, no statistical significant differences were observed between patients and proxies. Not surprisingly, Bland–Altman plots revealed a high agreement between the patient and patient-by-proxy rating in all HRQOL domains ranging from 95 to 99 %. The CCC was fairly high in all HRQOL domains (0.37–0.80), and the percentage of perfect agreement (PD ± 0) ranged from 8.5 to 76.8 %. In the cognitively impaired patients, the mean difference between patients and proxies was overall larger, and accordingly, agreement based on Bland–Altman plots was lower. Conclusions: The level of agreement between patient and patient-by-proxy ratings of low-grade glioma patients’ HRQOL is generally high. However, patient–proxy agreement is lower in patients with neurocognitive deficits than in patients without neurocognitive deficits.

Original languageEnglish
Pages (from-to)869-880
Number of pages12
JournalQuality of life research
Issue number4
Publication statusPublished - 1 Apr 2017


  • Brain tumor
  • Health-related quality of life
  • Neurocognitive deficits
  • Proxy ratings

Cite this