Elsevier

European Journal of Cancer

Volume 49, Issue 13, September 2013, Pages 2910-2918
European Journal of Cancer

Rasch-built Overall Disability Scale for patients with chemotherapy-induced peripheral neuropathy (CIPN-R-ODS)

https://doi.org/10.1016/j.ejca.2013.04.004Get rights and content

Abstract

Chemotherapy-induced peripheral neuropathy (CIPN) is a common neurological side-effect of cancer treatment and may lead to declines in patients’ daily functioning and quality of life. To date, there are no modern clinimetrically well-evaluated outcome measures available to assess disability in CIPN patients. The objective of the study was to develop an interval-weighted scale to capture activity limitations and participation restrictions in CIPN patients using the Rasch methodology and to determine its validity and reliability properties. A preliminary Rasch-built Overall Disability Scale (pre-R-ODS) comprising 146 items was assessed twice (interval: 2–3 weeks; test–retest reliability) in 281 CIPN patients with a stable clinical condition. The obtained data were subjected to Rasch analyses to determine whether model expectations would be met, and if necessarily, adaptations were made to obtain proper model fit (internal validity). External validity was obtained by correlating the CIPN-R-ODS with the National Cancer Institute-Common Toxicity Criteria (NCI-CTC) neuropathy scales and the Pain-Intensity Numeric-Rating-Scale (PI-NRS).

The preliminary R-ODS did not meet Rasch model’s expectations. Items displaying misfit statistics, disordered thresholds, item bias or local dependency were systematically removed. The final CIPN-R-ODS consisting of 28 items fulfilled all the model’s expectations with proper validity and reliability, and was unidimensional.

The final CIPN-R-ODS is a Rasch-built disease-specific, interval measure suitable to detect disability in CIPN patients and bypasses the shortcomings of classical test theory ordinal-based measures. Its use is recommended in future clinical trials in CIPN.

Introduction

Chemotherapy-induced peripheral neuropathy (CIPN) is a frequent neurological complication of cancer treatment.1 The overall incidence of CIPN is 3–7% in people treated with single agents but more severe neuropathy affects up to 30% of those treated with some currently approved drugs.1, 2 Moreover, peripheral neurotoxicity may occur in 38% of those treated with multiple chemotherapeutic agents often leading to dose adjustment or withdrawal of anti-neoplastic drugs.2 CIPN is difficult to treat and may severely affect patients’ daily functioning and quality of life.3 In light of potential therapeutic interventions for CIPN, effective outcome measures are mandatory. These outcome measures should fulfil all classical requirements, i.e. being simple, understandable, valid, reliable and responsive as well as meeting modern scientific objectives, i.e. having an interval or linear construct.4, 5, 6, 7

Outcome measures currently applied in CIPN are mostly ordinal based and some mix impairment and disability qualities.8, 9 Sum scores are generally calculated in multi-item measures assuming a linear pattern of the obtained scores, which is highly unlikely.10, 11 Therefore, data derived from these outcome measures should not be analysed using parametric statistics. The Rasch model overcomes these limitations.4, 5, 12 This method is based on the logical assumption that a person with a higher ability (in the current setting read: less disability) will have a higher probability of completing a more difficult daily item or task when compared to a person with lower ability. Scales based on Rasch analysis have the ability to locate both subjects’ ability and item difficulty along the same metric.4

The current paper presents the results of a Rasch-built CIPN specific outcome measure capturing activity limitations and participation restriction as part of the chemotherapy-induced peripheral neuropathy outcome measure standardization (CI-PeriNomS) study.13, 14

Section snippets

Patients and methods

After two consensus meetings between members of participating centres (20 neurologists and oncologists as part of the CI-PeriNomS study group), 281 CIPN patients aged 18 years and older were recruited from 19 European clinical centres between July 2008 and December 2010. Patients were considered eligible if they met inclusion/exclusion criteria of the CI-PeriNomS study as previously described.13, 14

Description of patients

A total of 281 patients were eligible. The median age was 63.9 years with an equal gender distribution. The commonest malignancy was colorectal cancer (42%). Patient characteristics are presented in Table 1 with further information as previously published.14

Data quality control

Data quality control was performed as the initial step in the analyses. Items scored as (3) ‘not applicable’ were entered as missing data. Items with an arbitrarily chosen cut-off of >15% missing values as well as subjects with >15%

Discussion

The current study provides a new scale, the CIPN-R-ODS, based on the Rasch methodology, measuring activity and participation in patients with stable CIPN.4, 5, 19 This outcome measure fulfilled all modern clinimetric requirements, showed invariance across items and persons (Fig. 5) and bypassed the disadvantages of ordinal based measures that have been previously used in CIPN.10 The constructed final CIPN-R-ODS demonstrated proper validity and reliability aspects. A high personal separation

Conflict of interest statement

Dr. Merkies had served on scientific advisory boards for Immune Globulin Intravenous CIDP Efficacy (ICE) trial, CSL Behring, and Novartis International AG; and had received research support from Talecris Biotherapeutics. The other authors declare no conflict of interest.

Acknowledgements

The patience and availability of the subjects who agreed to participate in the study are gratefully acknowledged. This work was not supported by any funding.

References (29)

  • J.F. Pallant et al.

    An introduction to the Rasch measurement model: an example using the hospital anxiety and depression scale (HADS)

    Br J Clin Psychol

    (2007)
  • G. Rasch

    Probabilistic models for some intelligence and attainment tests

    (1960)
  • D.L. Streiner et al.

    Health measurement scales. A practical guide to their development and use

    (1995)
  • K.A. Griffith et al.

    Measures of chemotherapy-induced peripheral neuropathy: a systematic review of psychometric properties

    J Peripher Nerv Syst

    (2010)
  • Cited by (0)

    View full text