The Autism Impact Measure (AIM): Meaningful Change Thresholds and Core Symptom Changes Over One Year from an Online Survey in the U.S.

Validated outcome measures with the capacity to reflect meaningful change are key to assessing potential interventions for autism spectrum disorder (ASD). We derive clinically meaningful change thresholds (MCTs) of the Autism Impact Measure (AIM) and identify factors associated with meaningful change. Baseline and 12-months follow-up survey of caregivers of 2,761 children with ASD aged 3–17 years from the U.S. Simons Foundation Powering Autism Research for Knowledge (SPARK) cohort were analyzed. Using caregiver-reported anchors for change, the 12-month change in estimated AIM MCT (95% confidence interval) for symptom improvement was –4.5 (–7.61, –1.37) points and 9.9 (5.12, 14.59) points for symptom deterioration. These anchor-based MCTs will facilitate future assessments of caregiver-reported change in AIM scores. Supplementary Information The online version contains supplementary material available at 10.1007/s10803-022-05635-7.


Fig. S1 aa
Fig. S1 Change in AIM frequency score (a), impact score (b), and repetitive behavior domain score (c) versus the primary anchor (overall caregiver impression of change) Fig. S7 Primary and secondary anchor-based within-person meaningful change thresholds for much improvement and much deterioration Fig. S8 Primary and secondary anchor-based within-person meaningful percent change thresholds for much improvement and much deterioration

Table S1
Additional baseline characteristics of the analysis population a The chi-square test was used to compare the distribution of categorical variables between the subset of children for whom an IQ score was known, and the overall group b Continuous variables between the two groups were compared using the t-test Note.Numbers indicate n (%); percent denominators do not include counts of Missing/Don't know/Prefer not to answer.Abbreviations: AIM Autism Impact Measure, ASD autism spectrum disorder, IQ intelligence quotient, SD standard deviation, USD United States Dollar

Table S2
Factors associated with AIM change scores

Table S3
Factors associated with a primary anchor meaningful change threshold response for improvement on total AIM

Table S4
Factors associated with a primary anchor meaningful change threshold response for deterioration on total AIM Note.All factors listed in the table were included in the multivariable model for clinically important responder for deterioration.P-values arrived at by applying the t-test to coefficient estimates.Abbreviations: AIM Autism Impact Measure, ASD autism spectrum disorder, CI confidence interval, ER emergency room

Table S5 .
Meaningful change thresholds for children with IQ data (change and percent change from baseline)

Secondary anchor: caregiver impression of change in communication
"No change": CaGI-C overall ASD assessed as "no change."% Improvement and % Deterioration is % change relative to baseline value.Abbreviations: AIM Autism Impact Measure, ASD autism spectrum disorder, CaGI-C Caregiver-reported Global Impression of Change survey, CI confidence interval, IQ intelligence quotient

Table S6
Factors associated with AIM change scores for children with IQ data Note.All factors listed in the table were included in the multivariable model for AIM score change for children with IQ data.P-values arrived at by applying the t-test to coefficient estimates.Abbreviations: AIM Autism Impact Measure, ASD autism spectrum disorder, IQ intelligence quotient

Table S7
Sensitivity analysis: Meaningful change thresholds estimated assuming equidistance anchors a The equidistant sensitivity analysis assumes that the improvement and deterioration anchors are of equal magnitude; the value of the improvement estimate is negative, and the value of the deterioration estimate is positive.Note.The magnitude of the CIR threshold estimates derived in this analysis falls between the magnitude of the Improvement and Deterioration CIR threshold estimates in the primary analysis.Abbreviations: AIM Autism Impact Measure, ASD autism spectrum disorder, CI confidence interval

Table S8
Variability of AIM score change for individuals identified by their caregiver as having no overall change in ASD (CaGI-C) at follow-up Note.The median change from baseline during the year for all subjects for whom the caregivers reported no overall change in ASD is slightly negative.Abbreviations: AIM Autism Impact Measure, ASD autism spectrum disorder, CaGI-C Caregiver-reported Global Impression of Change survey, IQR interquartile range, SD standard deviation

Table S9
Sensitivity analysis: anchor-based meaningful change thresholds for improvement based on ROC analysis a Empirical confidence interval derived from 5,000 bootstrap samples.Intervals are not bias corrected.Abbreviations: AIM Autism Impact Measure, ASD autism spectrum disorder, AUC area under the curve, CI confidence interval, ROC receiver operator characteristic

Table S10
Anchor-based meaningful change thresholds for deterioration based on ROC analysis

Secondary anchor: caregiver impression of change in social interaction
Empirical confidence interval derived from 5,000 bootstrap samples.Intervals are not bias corrected.Abbreviations: AIM Autism Impact Measure, ASD autism spectrum disorder, AUC area under the curve, CI confidence interval, ROC receiver operator characteristic a