Content validity and legacy patient-reported outcome measures in cancer.

Abstract

The publication in 2009 of the Guidance for Industry by US Food and Administration (FDA) [1] ushered in a significant change in the way patient-reported outcome measures (PROMs) are evaluated when used by pharmaceutical companies to support product label claims. In particular, the FDA Guidance emphasised patient input in the development of PROMs. To underline this point, a review of the reasons for FDA rejections of PRO label claims [2] concluded that just under 40 % of all rejections were due to the PROM not being fit for purpose; 31 % of product PRO claims were rejected due to a lack of content validity. This approach adopted by the FDA may be problematic for those legacy PROMs whose development predates the Guidance, one of which is the European Organisation for Research and Treatment of Cancer (EORTC) QLQ-C30 [3]. There is no evidence that the early development of the EORTC QLQ-C30 incorporated patient involvement in the content validity other than in rating the acceptability of items. Despite this, the instrument continues to be widely used in oncology clinical trials: a search on clinicaltrials.gov for the period 1 January 2009 to 31 December 2013 revealed 207 registered trials that included the EORTC QLQ-C30. Furthermore, its inclusion has been recommended in comparative effectiveness research [4], and it has been shown to be of use in facilitating patient–clinician communication [5]. Beyond that the instrument has been part of a successful PRO product label claim granted by the FDA [6] after the publication of the Guidance document. Content validity is one particular psychometric property of the EORTC QLQ-C30 that has not been revisited since the early development of the instrument. Although the EORTC QLQ-C30 has been subsequently validated with other cancer diagnoses, it has undergone little or no change in terms of the original content. However, this needs to be balanced against the fact that the EORTC QLQ-C30 has been completed by thousands of patients with different types of cancers since its development. There are currently 31 studies registered on the EORTC clinical trials database with a total target recruitment of 29,106 patients across a spectrum of cancers including brain, breast, (paediatric) leukaemia, endocrine, gastrointestinal, genito-urinary, gynaecological, head and neck, lung, and melanoma (www.eortc.org/clinical-trials). It also should be stressed that the EORTC guidelines for the development of diseasespecific modules emphasise the necessity of patient input in order to enhance content validity [7]. It could therefore be argued that the EORTC QLQ-C30 has been subjected to a content validation process through exposure in numerous clinical trials, particularly given that the domains of the instrument, such as fatigue, sleeplessness, pain, appetite loss, nausea, and cognitive problems largely correspond with the core set of symptoms recommended for assessment by the National Cancer Institute’s Symptom Management & Health-related Quality of Life Steering Committee [8]. It remains for future research to determine whether the EORTC QLQC30 requires further modification or is—in its current form and owing to the significant post-development use across different cancers and treatments—fit for purpose in oncology trials. A. B. Smith (&) York Health Economics Consortium, Market Square, University of York, York YO10 5NH, UK e-mail: adam.smith@york.ac.uk