Original ResearchThe EORTC CAT Core—The computer adaptive version of the EORTC QLQ-C30 questionnaire
Introduction
Patient-reported outcomes (PROs) are the primary source of information about patients' health-related quality of life (HRQOL). PRO measures (PROMs) are typically static, standardised questionnaires i.e. all patients are asked the same set of items yielding scores that are comparable across patients. To achieve precise measurements for patients at different levels of HRQOL, traditional PROMs often require a substantial number of items; more than may be feasible and/or reasonable to ask patients to complete. Therefore, such PROMs typically represent a compromise between the need to minimise patient burden, while achieving adequate measurement precision.
Item response theory (IRT) provides a family of statistical models to describe the psychometric characteristics of items in multi-item scales [1]. In recent years, there has been an increasing interest in the use of IRT when developing new PRO/HRQOL measures and for enhancing existing ones. A simple search in PubMed® using search terms ‘item response theory’ AND (‘quality of life’ OR ‘patient reported outcome’) resulted in 5 hits for 2000, 21 for 2005 and 69 for 2015 [2]. One of the primary reasons for this increasing interest and one of the major advantages of IRT is that, when a set of items has been calibrated (estimated) to an IRT model, scores based on any subset of the items are on the same metric and hence are directly comparable. This unique feature allows the content of questionnaires to be adapted to the individual patient without compromising the comparability of scores across patients. This is utilised by computerised adaptive tests (CATs) to present the most informative items to each patient, thereby optimising the measurement properties [3]. During a CAT assessment, item selection is tailored to the individual based on responses to prior items, i.e. in each step the choice of item is adapted to the current estimate of the patient's location on a health continuum (e.g. physical functioning or fatigue).
CAT measures have several advantages over measures based on classical test theory. These include increased measurement precision, reduced respondent burden, increased question relevance to individual patients, and increased flexibility. The length of the CAT questionnaire can be adapted to each study or patient, and scores can be generated automatically, facilitating real time feedback of results.
Because of the clear advantages of CAT measurement, the European Organisation for Research and Treatment of Cancer (EORTC) Quality of Life Group (QLG) initiated in 2005 a project to develop a CAT PROM specifically relevant for cancer patients. Since the EORTC QLG's core questionnaire, the QLQ-C30, is one of the most widely used PROMs in cancer research and clinical practice [4], the QLG deemed it particularly relevant to enhance the measurement properties of this instrument. Therefore, the aim of this project was to develop item banks (calibrated collections of items) for CAT measurement of the 14 functional and symptom domains out of the 15 domains assessed by the EORTC QLQ-C30 (additional items have not been developed for the overall quality of life scale) [5], [6]. Development of the CAT required supplementing the QLQ-C30 with additional items to more fully assess each HRQOL domain. Basing the CAT on the conceptual model of the QLQ-C30 ensures maximum backward compatibility with the original instrument. Hence, future studies using the EORTC CAT can compare results with the substantial body of literature of studies using the original QLQ-C30. In particular, a CAT assessment can be set up to ask all QLQ-C30 items (supplemented with additional items for increased precision) if a direct assessment of QLQ-C30 is desired.
We have completed the development of all 14 item banks for the CAT version of the QLQ-C30. The analyses and results of some, but not all, of the individual item banks have been published in details [6], [7], [8], [9], [10], [11], [12], [13], [14], [15], [16], [17]. Here we present for the first time an overview of the complete CAT instrument, the EORTC CAT Core.
Section snippets
Methods
The development of all 14 CAT item banks followed the same general approach, which comprises four phases: (I) literature search, (II) operationalisation (selection and formulation of items), (III) pretesting and (IV) field-testing. This approach has been described in details elsewhere [6], [12]. Following is a summary of the approach.
A literature search was conducted to identify existing items used to measure the domain of focus (e.g. physical functioning or fatigue). These were not systematic
Results
The literature searches in phase I revealed a wide range in the number of items that have been used to measure the different QLQ-C30 domains, ranging from 122 role functioning items to 1729 items for emotional functioning. Across all domains, deletion of redundant items and items not relevant for our measurement model or incompatible with the QLQ-C30 item style reduced the lists by 69–94% leaving, on average, 15% of the items identified in the literature.
The retained items formed the basis for
Discussion
CAT measurement adapts in real time a PROM to the individual patient while retaining direct comparability of scores across patients. The EORTC CAT Core presented here measures the same HRQOL domains as the widely used EORTC QLQ-C30 questionnaire. This ‘intelligent’ version of the QLQ-C30 allows clinicians and researchers to assess each patient's level of symptoms and functioning using only the most informative items from the item banks that have been developed. Thereby, measurement precision
Conflict of interest statement
None declared.
Funding
The study was funded by the EORTC Quality of Life Group (grant no. 005-2007). The work of Johannes M. Giesinger was funded by the Austrian Science Fund (FWF #P26930).
Role of the funding source
The funder of the study had no role in the study design, data collection, analysis or interpretation or writing of the article. The executive committee of the EORTC Quality of Life Group read the article and approved it as being ‘on behalf of EORTC Quality of Life Group’. The corresponding author had full access to all the data in the study and had final responsibility for the decision to submit for publication.
The EORTC Quality of Life Group's PROMs can be used free of charge in academic
References (32)
- et al.
Development of computerised adaptive testing (CAT) for the EORTC QLQ-C30 dimensions – general approach and initial results for physical functioning
Eur J Cancer
(2010) - et al.
The EORTC computer-adaptive tests measuring physical functioning and fatigue exhibited high levels of measurement precision and efficiency
J Clin Epidemiol
(2013) - et al.
An emotional functioning item bank of 24 items for computer adaptive testing (CAT) was established
J Clin Epidemiol
(2016) - et al.
Development and preliminary testing of a computerized adaptive assessment of chronic pain
J Pain
(2009) - et al.
An item bank was created to improve the measurement of cancer-related fatigue
J Clin Epidemiol
(2005) - et al.
Handbook of item response theory modeling: applications to typical performance assessment
(2015) MEDLINE search
(January 5th 2017)Computerized adaptive testing: a primer
(2000)Quality of life research within the EORTC-the EORTC QLQ – C30
Eur J Cancer
(2002)- et al.
The European Organization for Research and Treatment of Cancer QLQ-C30: a quality-of-life instrument for use in international clinical trials in oncology
J Natl Cancer Inst
(1993)
Psychometric evaluation of an item bank for computerized adaptive testing of the EORTC QLQ-C30 cognitive functioning dimension in cancer patients
Qual Life Res
Development of an item bank for computerized adaptive testing of self-reported cognitive difficulty in cancer patients
Neuro Oncol Pract
The EORTC emotional functioning computer adaptive test (CAT): phase I–III of a cross-cultural item bank development
Psycho Oncol
Development of an item bank for the EORTC role functioning computer adaptive test (EORTC RF-CAT)
Health Qual Life Outcomes
Cross-cultural development of an item list for computer-adaptive testing of fatigue in oncological patients
Health Qual Life Outcomes
Development of computerized adaptive testing (CAT) for the EORTC QLQ-C30 physical functioning dimension
Qual Life Res
Cited by (64)
Equipercentile equating of scores from common patient-reported outcome measures of physical function in patients with cancer
2024, Journal of Clinical EpidemiologyPatient-reported outcome measures in cancer care: Integration with computerized adaptive testing
2023, Asia-Pacific Journal of Oncology NursingDevelopment of an EORTC Item Bank for Computer-Adaptive Testing of Nausea and Vomiting
2022, Seminars in Oncology NursingCitation Excerpt :This approach allows optimization of measurement and flexibility of the basic EORTC tool. On one hand it allows adaptation to studies or patients themselves, and on the other, it still allows comparability of results between individuals or the studies.23,24 The statistical basis used in CAT measurement is item response theory that permits the uniform comparison of questionnaires between patients.20,21
Iterative Item Selection of Neighborhood Clusters: A Nonparametric and Non-IRT Method for Generating Miniature Computer Adaptive Questionnaires
2024, Educational and Psychological Measurement