Contributions
Abstract: P355
Type: Poster
Abstract Category: Clinical aspects of MS - Clinical assessment tools
Introduction: Neuropathic pain (NP) refers to pain arising as a direct consequence of a lesion or disease affecting the somatosensory system [1]. It may manifest in different ways, for example as tingling or burning. It is important to have confidence in the assessment of such pain, particularly where Patient Reported Outcome Measures (PROM"s) are used for such a purpose. One widely used PROM is the Neuropathic Pain Scale (NPS) [1]. It consists of 10 Numeric Rating Scale (NRS) responses, as well as additional questions asking the patient to describe aspects of the pain. Previously, classical test theory approaches have been used to demonstrate the reliability and validity of the NPS in MS [2]. The current study examines the scale from a modern (Rasch) psychometric perspective.
Methods: 722 people with MS completed a questionnaire booklet including the NPS. Data from this scale were fit to the Rasch model to ascertain if the NPS is compliant with modern standards for measurement, including transformation to interval scaling.
Results: Initial fit of the NPS to the Rasch model among those reporting pain was poor (Chi Square 355 (df(90); p=< 0.001). However, several pairs of items breached the local independence assumption (for example intensity and sharpness), indicating some redundancy in the scale. Consequently, the items were grouped into two testlets (to absorb the dependency) and fit to the model was then good (Chi-Square 21.8 (df 18); p= 0.24). The unidimensionality test showed a significant difference of 5.19% (95%CI: 3.6-6.8%) in the estimates derived from the two sets, supporting unidimensionality. The latent estimate derived from the procedure accounted for 93% of the total variance. Reliability (α) was 0.83.
Conclusion: The NPS showed good internal construct validity and reliability when its data were fit to the Rasch model. This analysis provides an interval scaled latent estimate which can be used to examine the relation of neuropathic pain to other disease factors, such as disease type and severity, and in models exploring the relationship of neuropathic pain to quality of life.
1. Galer BS, Jensen MP. Development and preliminary validation of a pain measure specific to neuropathic pain: the Neuropathic Pain Scale. Neurology. 1997;48(2):332-8.
2. Rog DJ, Nurmikko TJ, Friede T, Young CA. Validation and reliability of the Neuropathic Pain Scale (NPS) in multiple sclerosis. Clin J Pain. 2007;23(6):473-81.
Disclosure:
Carolyn Young has received honoraria and travel expenses for scientific meetings and advisory boards, or grants from Bayer, Biogen Idec, Merck Serono, Genzyme, Motor Neurone Disease Association, MS Trust, National Institute for Health Research, Novartis, Roche, Teva, and Wellcome Trust.
Roger Mills has received conference expenses from Biogen Idec and Teva.
Alan Tennant: nothing to disclose.
Abstract: P355
Type: Poster
Abstract Category: Clinical aspects of MS - Clinical assessment tools
Introduction: Neuropathic pain (NP) refers to pain arising as a direct consequence of a lesion or disease affecting the somatosensory system [1]. It may manifest in different ways, for example as tingling or burning. It is important to have confidence in the assessment of such pain, particularly where Patient Reported Outcome Measures (PROM"s) are used for such a purpose. One widely used PROM is the Neuropathic Pain Scale (NPS) [1]. It consists of 10 Numeric Rating Scale (NRS) responses, as well as additional questions asking the patient to describe aspects of the pain. Previously, classical test theory approaches have been used to demonstrate the reliability and validity of the NPS in MS [2]. The current study examines the scale from a modern (Rasch) psychometric perspective.
Methods: 722 people with MS completed a questionnaire booklet including the NPS. Data from this scale were fit to the Rasch model to ascertain if the NPS is compliant with modern standards for measurement, including transformation to interval scaling.
Results: Initial fit of the NPS to the Rasch model among those reporting pain was poor (Chi Square 355 (df(90); p=< 0.001). However, several pairs of items breached the local independence assumption (for example intensity and sharpness), indicating some redundancy in the scale. Consequently, the items were grouped into two testlets (to absorb the dependency) and fit to the model was then good (Chi-Square 21.8 (df 18); p= 0.24). The unidimensionality test showed a significant difference of 5.19% (95%CI: 3.6-6.8%) in the estimates derived from the two sets, supporting unidimensionality. The latent estimate derived from the procedure accounted for 93% of the total variance. Reliability (α) was 0.83.
Conclusion: The NPS showed good internal construct validity and reliability when its data were fit to the Rasch model. This analysis provides an interval scaled latent estimate which can be used to examine the relation of neuropathic pain to other disease factors, such as disease type and severity, and in models exploring the relationship of neuropathic pain to quality of life.
1. Galer BS, Jensen MP. Development and preliminary validation of a pain measure specific to neuropathic pain: the Neuropathic Pain Scale. Neurology. 1997;48(2):332-8.
2. Rog DJ, Nurmikko TJ, Friede T, Young CA. Validation and reliability of the Neuropathic Pain Scale (NPS) in multiple sclerosis. Clin J Pain. 2007;23(6):473-81.
Disclosure:
Carolyn Young has received honoraria and travel expenses for scientific meetings and advisory boards, or grants from Bayer, Biogen Idec, Merck Serono, Genzyme, Motor Neurone Disease Association, MS Trust, National Institute for Health Research, Novartis, Roche, Teva, and Wellcome Trust.
Roger Mills has received conference expenses from Biogen Idec and Teva.
Alan Tennant: nothing to disclose.