INS NYC 2024 Program

Poster	Poster Session 10 Program Schedule 02/17/2024 09:00 am - 10:15 am Room: Majestic Complex (Posters 61-120) Poster Session 10: Neurodevelopmental \| Congenital Conditions Final Abstract #88 Improving Precision and Accessibility of ADHD Assessment with Digital Measures Amber Schaefer, Massachusetts General Hospital, Boston, United States Christopher Nicholls, The Nicholls Group, Scottsdale, United States Hana Gross, Massachusetts General Hospital, Boston, United States Jessica Steir, Massachusetts General Hospital, Boston, United States Cole Hague, Massachusetts General Hospital, Boston, United States Category: ADHD/Attentional Functions Keyword 1: attention Objective: Despite a 5-10% global prevalence, there is debate regarding universal diagnostic standards for Attention-Deficit/Hyperactivity Disorder (ADHD). Outdated criteria overlook executive control nuances, emphasizing basic attention and overlooking higher-order functions. Traditional neuropsychological evaluations are resource-intensive, but digital tools like the NIH Toolbox Cognition Battery (NTCB) and Test of Variables of Attention (TOVA) provide equally accurate and efficient measurement with robust validity across diverse samples (Weintraub et al., 2013; Leark et al., 2020). This study investigates executive function profiles in an ADHD sample using the NTCB and TOVA. Participants and Methods: Data from 213 clinically diagnosed ADHD subjects aged 5-26 years (57.7% male; M_age = 12.21, SD = 4.26) informed latent profile analysis (LPA) clusters using NTCB (LPA 1), TOVA. (LPA 2), and combined measures (LPA 3), with varying subsample sizes by test age criteria. Analytics used MPlus with full-information maximum likelihood for missing data. Model selections were based on theory, fit indices, and likelihood ratios (Nylund et al., 2007). Results: LPA 1 (n = 189): Using NTCB Flanker, List Sorting, Dimensional Change Card Sort and Pattern Comparison subtests, a two-profile model was selected based on high entropy (0.829), lower/comparable fit indices (AIC = 6261.479; BIC = 6303.622; aBIC = 6262.444), and likelihood ratio p-value<0.001. Clusters were labeled “Impulsive and Inflexible” (n = 144; x̄_Flanker = 83.5, x̄_DCCS = 84.95) and “Relatively Variable Impulsivity” (n = 45; SS ≥ 15 mean score difference between x̄_Flanker = 99.18 and x̄_DCCS = 114.46). Both profiles demonstrated intact working memory (x̄_List) and processing speed (x̄_Pattern). LPA 2 (n = 185): Using TOVA response time variability, response time, commission errors, and omission errors, a four-profile model was selected based on high entropy (0.882), lowest fit indices (AIC = 6435.521; BIC = 6509.589; aBIC = 6436.741), and likelihood ratio p-value<0.001. Clusters were “Inefficient, Inattentive, and Impulsive” (n = 62; x̄_TovaRTV = 47.52; x̄_TovaRT = 73.69; x̄_TovaCE = 74.46; x̄_TovaOE = 42.39), “Inattentive and Impulsive” (n = 27; x̄_TovaCE = 74.67; x̄_TovaOE = 44.69), “Variable Processing” (n = 34; x̄_TovaRTV = 59.29; x̄_TovaRT = 78.90), and “Developmentally Appropriate” (n = 62; consistently average scores). LPA 3 (n = 189): Using NTCB tests and TOVA scores, a four-profile model was selected based on high entropy (0.871), lowest fit indices (AIC = 9376.744; BIC = 9483.722; aBIC = 9379.193), and likelihood ratio p-value<0.001. Clusters were “Subtly Impulsive with Diminishing Focus” (n = 78; x̄_Flanker = 82.30; x̄_DCCS = 83.93; x̄_TovaRTV = 54.98; x̄_TovaOE = 42.96), “Inconsistent with Waning Focus” (n = 14; x̄_TovaRTV = 65.64; x̄_TovaOE = 42.84), “Variable and Inefficient Processing” (n = 65; x̄_Flanker = 84.13; x̄_DCCS = 86.51 x̄_TovaRTV = 74.58), and “Brief Proficiency with Relative Inconsistency” (n = 32; significant score difference between x̄_TovaRTV = 96.95 and x̄_List = 110.06, x̄_DCCS = 114.46, and x̄_Pattern = 111.69). Conclusions: Integrating NTCB and TOVA scores revealed nuanced profiles beyond dichotomous inattentive and hyperactive clusters. Digital assessments may promote equity by expediting diagnosis and supporting targeted interventions.

Poster

02/17/2024
09:00 am - 10:15 am
Room: Majestic Complex (Posters 61-120)

Poster Session 10: Neurodevelopmental | Congenital Conditions

Final Abstract #88

Improving Precision and Accessibility of ADHD Assessment with Digital Measures

Amber Schaefer, Massachusetts General Hospital, Boston, United States
Christopher Nicholls, The Nicholls Group, Scottsdale, United States
Hana Gross, Massachusetts General Hospital, Boston, United States
Jessica Steir, Massachusetts General Hospital, Boston, United States
Cole Hague, Massachusetts General Hospital, Boston, United States

Category: ADHD/Attentional Functions

Keyword 1: attention

Objective:

Despite a 5-10% global prevalence, there is debate regarding universal diagnostic standards for Attention-Deficit/Hyperactivity Disorder (ADHD). Outdated criteria overlook executive control nuances, emphasizing basic attention and overlooking higher-order functions. Traditional neuropsychological evaluations are resource-intensive, but digital tools like the NIH Toolbox Cognition Battery (NTCB) and Test of Variables of Attention (TOVA) provide equally accurate and efficient measurement with robust validity across diverse samples (Weintraub et al., 2013; Leark et al., 2020). This study investigates executive function profiles in an ADHD sample using the NTCB and TOVA.

Participants and Methods:

Data from 213 clinically diagnosed ADHD subjects aged 5-26 years (57.7% male; M_age = 12.21, SD = 4.26) informed latent profile analysis (LPA) clusters using NTCB (LPA 1), TOVA. (LPA 2), and combined measures (LPA 3), with varying subsample sizes by test age criteria. Analytics used MPlus with full-information maximum likelihood for missing data. Model selections were based on theory, fit indices, and likelihood ratios (Nylund et al., 2007).

Results:

LPA 1 (n = 189): Using NTCB Flanker, List Sorting, Dimensional Change Card Sort and Pattern Comparison subtests, a two-profile model was selected based on high entropy (0.829), lower/comparable fit indices (AIC = 6261.479; BIC = 6303.622; aBIC = 6262.444), and likelihood ratio p-value<0.001. Clusters were labeled “Impulsive and Inflexible” (n = 144; x̄_Flanker = 83.5, x̄_DCCS = 84.95) and “Relatively Variable Impulsivity” (n = 45; SS ≥ 15 mean score difference between x̄_Flanker = 99.18 and x̄_DCCS = 114.46). Both profiles demonstrated intact working memory (x̄_List) and processing speed (x̄_Pattern).

LPA 2 (n = 185): Using TOVA response time variability, response time, commission errors, and omission errors, a four-profile model was selected based on high entropy (0.882), lowest fit indices (AIC = 6435.521; BIC = 6509.589; aBIC = 6436.741), and likelihood ratio p-value<0.001. Clusters were “Inefficient, Inattentive, and Impulsive” (n = 62; x̄_TovaRTV = 47.52; x̄_TovaRT = 73.69; x̄_TovaCE = 74.46; x̄_TovaOE = 42.39), “Inattentive and Impulsive” (n = 27; x̄_TovaCE = 74.67; x̄_TovaOE = 44.69), “Variable Processing” (n = 34; x̄_TovaRTV = 59.29; x̄_TovaRT = 78.90), and “Developmentally Appropriate” (n = 62; consistently average scores).

LPA 3 (n = 189): Using NTCB tests and TOVA scores, a four-profile model was selected based on high entropy (0.871), lowest fit indices (AIC = 9376.744; BIC = 9483.722; aBIC = 9379.193), and likelihood ratio p-value<0.001. Clusters were “Subtly Impulsive with Diminishing Focus” (n = 78; x̄_Flanker = 82.30; x̄_DCCS = 83.93; x̄_TovaRTV = 54.98; x̄_TovaOE = 42.96), “Inconsistent with Waning Focus” (n = 14; x̄_TovaRTV = 65.64; x̄_TovaOE = 42.84), “Variable and Inefficient Processing” (n = 65; x̄_Flanker = 84.13; x̄_DCCS = 86.51 x̄_TovaRTV = 74.58), and “Brief Proficiency with Relative Inconsistency” (n = 32; significant score difference between x̄_TovaRTV = 96.95 and x̄_List = 110.06, x̄_DCCS = 114.46, and x̄_Pattern = 111.69).

Conclusions:

Integrating NTCB and TOVA scores revealed nuanced profiles beyond dichotomous inattentive and hyperactive clusters. Digital assessments may promote equity by expediting diagnosis and supporting targeted interventions.