Minimal Clinically Important Difference in the Physical, Emotional, and Total Symptom Distress Scores of the Edmonton Symptom Assessment System

Abstract

Context: The Edmonton Symptom Assessment System (ESAS) is one of the most commonly used symptom batteries in clinical practice and research.

Objectives: We used the anchor-based approach to identify the minimal clinically important difference (MCID) for improvement and deterioration for ESAS physical, emotional, and total symptom distress scores.

Methods: In this multicenter prospective study, we asked patients with advanced cancer to complete their ESAS at the first clinic visit and at a second visit three weeks later. The anchor for MCID determination was Patient's Global Impression regarding their physical, emotional, and overall symptom burden ("better," "about the same," or "worse"). We identified the optimal sensitivity/specificity cutoffs for both improvement and deterioration for the three ESAS scores and also determined the within-patient changes.

Results: A total of 796 patients were enrolled from six centers. The ESAS scores had moderate responsiveness, with area under the receiver operating characteristic curve between 0.69 and 0.76. Using the sensitivity-specificity approach, the optimal cutoffs for ESAS physical, emotional, and total symptom distress scores were ≥3/60, ≥2/20, and ≥3/90 for improvement, and ≤-4/60, ≤-1/20, and ≤-4/90 for deterioration, respectively. These cutoffs had moderate sensitivities (59%-68%) and specificities (62%-80%). The within-patient change approach revealed the MCID cutoffs for improvement/deterioration to be 3/-4.3 for the physical score, 2.4/-1.8 for the emotional score, and 5.7/-2.9 for the total symptom distress score.

Conclusion: We identified the MCIDs for physical, emotional, and total symptom distress scores, which have implications for interpretation of symptom response in clinical trials.

Keywords: Neoplasms; outcome measures; pain; sample size; sensitivity and specificity; symptom assessment.

Copyright © 2016 American Academy of Hospice and Palliative Medicine. Published by Elsevier Inc. All rights reserved.

Figures

Fig. 1. Average Edmonton Symptom Assessment System symptom change by PGI categories

(A) Physical symptoms, (B) Emotional symptoms, and (C) Total symptom burden. PGI was strongly associated with individual symptom change in all instances (ANOVA P<0.001)

Figure 2. Receiver-operating characteristic curves for (A) Improvement and (B) Deterioration for the Edmonton Symptom Assessment System physical, emtoional and total symptom distress scores

The area under the curve ranged between 0.69 to 0.76, suggesting moderate discrimination.