# Determining minimal clinically important differences in ecological momentary assessment measures of fatigue in people with multiple sclerosis

## TL;DR

This study establishes the smallest meaningful changes in fatigue levels measured in real-time for people with multiple sclerosis, improving treatment outcome assessments.

## Contribution

The study determines minimal clinically important differences for EMA fatigue outcomes in multiple sclerosis using patient-reported and statistical methods.

## Key findings

- Using PGIC and MFIS anchors, changes in EMA fatigue scores averaged 0.94 for intensity and 0.62 for interference.
- Standard deviations and SEM for EMA fatigue intensity were 0.75 and 1.19, respectively.
- EMA provides more reliable fatigue assessments than traditional recall surveys, reducing recall bias.

## Abstract

Fatigue is a common debilitating symptom of multiple sclerosis (MS). Ecological momentary assessment (EMA) provides a more reliable and sensitive assessment of fatigue outcomes relative to traditional recall surveys; however, the minimal clinically important difference (MCID) for EMA fatigue outcomes has not been established.

MCIDs for EMA fatigue intensity and fatigue interference (0–10 numerical rating scale) that were assessed as outcomes in a pragmatic randomized clinical trial of three fatigue interventions were determined using two statistical approaches. The Patient Global Impression of Change (PGIC) and the Modified Fatigue Impact Scale (MFIS) were used within the anchor-based approach, and standard deviations (SD) and standard error of measurements (SEM) were examined within the distribution-based approach.

Pre- and post-treatment EMA data from 336 individuals with MS (76.2% female, 71.1% relapsing–remitting MS, mean age 48.8 (± 11.7) years, mean duration of MS 12.2 (± 9.8) years) were included in the analysis. Percent complete EMA data (4 EMAs/day) for 7 days were comparable pre- and post-treatment for fatigue intensity and for fatigue interference. Using the PGIC and MFIS anchors, change in EMA scores averaged 0.94 and 1.04 for fatigue intensity and 0.62 and 1.04 for fatigue interference, respectively. The SD and SEM for EMA fatigue intensity were 0.75 and 1.19 and for EMA fatigue interference were 0.83 and 1.30, respectively.

Combining two approaches, our study contributes foundational information regarding meaningful change on EMA measures of fatigue, enabling effective use of EMA to assess fatigue treatment outcomes in a person-centered manner.

The smallest changes in treatment outcomes considered important by patients are often determined for traditional recall surveys. However, recall-bias with traditional surveys may contribute to unreliable assessment of the outcome. Evidence suggests that ecological momentary assessments (EMA; real-time self-report) can provide a more reliable and sensitive assessment of symptoms relative to traditional surveys. Applying two analytic methods to investigate meaningful changes of EMA fatigue outcomes involving patient-oriented and sample specific measures of variability, our work addresses the challenge of recall-bias affecting assessments with traditional surveys and provides a practical way to assess meaningful change in fatigue treatment outcomes.