Determining the significance weights of ALSFRS-R items using analytic hierarchy process

Abstract

Objective: The objective of the study was to determine the significance weights of the Revised Amyotrophic Lateral Sclerosis Functional Rating Scale (ALSFRS-R) item scores for better evaluation and classification of amyotrophic lateral sclerosis (ALS) status. Methods: We used an analytical hierarchy process (AHP) to determine the weights of the ALSFRS-R item scores based on the opinions of two neurologists. We analyzed data of 51 patients with ALS to group them into four ALS severity classes based on their (i) total ALSFRS-R score and (ii) total weighted ALSFRS-R scores using the proposed weight values. We analyzed the performance differences between the two classification approaches based on the outcomes of these 51 patients. We also used twenty additional patients' data to analyze the accuracies of the total and weighted ALSFRS-R score approaches compared with physicians' actual assessments. Results: The AHP analysis assigned the highest weights to the ALSFRS-R items for respiratory insufficiency, orthopnea, and bed rotation/covering. ALS status classification based on the total and weighted ALSFRS-R scores differed for about 27.5% (confidence interval [CI]: 15.2%–39.8%) of the 51 patients. The classification based on the total weighted ALSFRS-R complied with the actual assessments in 85% (CI: 69.4%–100%) of the patients in the comparison sample; the compliance rate was 60% (CI: 38.5%–81.5%) for the total ALSFRS-R-based classification. Conclusions: Assigning weights to the 12 ALSFRS-R criteria/questions may improve ALSFRS-R's ability to represent ALS severity. This finding requires further investigation.