Modified Edmonton Symptom Assessment System Including Constipation and Sleep: Validation in Outpatients With Cancer

Abstract

ContextThe Edmonton Symptom Assessment System (ESAS) measures the severity of nine symptoms. Constipation and sleep disturbance are common in patients with cancer, but are not currently included in the ESAS. ObjectivesTo validate the numerical rating scale (NRS) versions of ESAS and its revised version (ESAS-r), with the additional symptoms of constipation and sleep (CS), and to assess patient preference for either version. MethodsOutpatients with advanced cancer (N = 202) completed three assessments during a single clinic visit: ESAS-CS, and an added time window of “past 24 hours”; ESAS-r-CS, with a time window of “now” and symptom definitions; and the Memorial Symptom Assessment Scale (MSAS). Internal consistency was calculated using Cronbach's alpha. Paired t-tests compared ESAS-CS and ESAS-r-CS scores; these were correlated with MSAS using Spearman correlation coefficients. Test-retest reliability at 24 hours was assessed in 26 patients. ResultsESAS-CS and ESAS-r-CS total scores correlated well with total MSAS (Spearman's rho 0.62 and 0.64, respectively). Correlation of individual symptoms with MSAS symptoms ranged from 0.54-0.80 for ESAS-CS and 0.52–0.74 for ESAS-r-CS. Although participants preferred the ESAS-r-CS format (42.8% vs. 18.6%) because of greater clarity and understandability, the “past 24 hours” time window (52.8%) was favored over “now” (21.3%). Shortness of breath and nausea correlated better for the “past 24 hours” time window (0.8 and 0.72 vs. 0.74 and 0.64 in ESAS-r-CS, respectively). The 24-hour test-retest of the ESAS-CS demonstrated acceptable reliability (intraclass correlation coefficient = 0.69). ConclusionThe ESAS-CS and ESAS-r-CS NRS versions are valid and reliable for measuring symptoms in this population of outpatients with advanced cancer. Although the ESAS-r-CS was preferred, patients favored the 24-hour time window of the ESAS-CS, which also may best characterize fluctuating symptoms.