Assessing the Clinical Response to Treatment in Patients With Insomnia

BALTIMORE—A minimal­ly important difference (MID) of 6 points for the Insomnia Severity Index (ISI) may be necessary to reflect a clinically meaningful response to treatment in an individual patient, according to the results of a study presented at the 22nd Annual Meeting of the Associated Professional Sleep Societies.
Patient-reported outcome measures, including the ISI, have been increasingly used to assess the subjective impact of insomnia on health status and daily functioning. However, an MID for the ISI has not previously been established for making clinical recommendations based on meaningful clinical score differences, said Min Yang, MD, PhD, a Senior Scientist with Outcomes Insight Consulting at QualityMetric, Inc, in Lincoln, Rhode Island.
Calculating the MID
Dr. Yang and colleagues assessed baseline and six-month posttreatment data from a double-blind, randomized, placebo-controlled trial evaluating the long-term efficacy of eszopiclone for treating patients with primary insomnia. A total of 828 patients ages 21 to 64 with primary insomnia were included in the study. An anchor-based method was used to define the MID of the ISI, by relating ISI scores to clinical and functional outcomes.
Outcomes on the Acute Short Form–36 Health Survey, Version 1; the Work Limitation Questionnaire; and the Fatigue Severity Scale were dichotomized to reflect the presence or absence of limitation. Logistic regression models were used to examine the relationship between the total ISI score and the probability of reporting a particular outcome limitation. Odds ratios of the likelihood of outcome variable status were calculated with use of β coefficients generated from logistic regression models, using an ISI score relative to the population mean. The degree of association with ISI score varied across the outcome variables, with the largest association related to vitality and fatigue.
Clinically Meaningful Differences
According to Dr. Yang’s team, baseline data evaluation suggested that 4 and 6 points higher than the population mean (17.9) were associated with a 1.2-fold and 2.3-fold increased risk, respectively, of reporting feeling worn out at least “a good bit of the time” during the past week; 4- and 6-point reductions in ISI score were associated with 55% and 70% reduced risk, respectively, of reporting feeling worn out that much throughout the past week.
In addition, 4- and 6-point increases in ISI score were associated with 1.5-fold and 2.9-fold increased risk, respectively, of experiencing fatigue, the researchers noted. Conversely, 4- and 6-point reductions in ISI score were associated with 59% and 74% reduced risk, respectively, for fatigue.
“Based on our research, we recommend, for population difference assessment, a difference of 4 or more points of ISI as a clinically meaningful difference from the perspective of changes in health-related quality of life, productivity, and fatigue related to insomnia,” Dr. Yang told Neurology Reviews. “However, when evaluating an individual’s response to treatment, a minimal improvement of 6 points might be necessary to reflect a clinically meaningful response.”
Limitations and Future Research
The trial had several limitations, noted Dr. Yang. “Due to data availability, no direct insomnia measures, such as polysomnography, were included. Future research including clinical indicators may be warranted in cross-validating the recommended MID of the ISI in our study.” In addition, the data were obtained from a clinical trial, which may not be representative of the population variance; therefore, further investigation using community-based populations is needed to generalize the findings.
“The use of MID provides a powerful tool for researchers, clinicians, patients, and policymakers who need to interpret mean score differences in the ISI between patient groups and change scores across time for an individual patient,” Dr. Yang concluded. “Application of the MID, in general, can help clinicians monitor treatment outcomes over time and assist researchers in clinical trial designs.”