Heartfelt Connections

Title: Simplified Methods for Identifying Subtypes of Automatically Maintained Self-Injury

Authors: Louis P Hagopian, John Michael Falligant, Michelle A Frank-Crawford

Journal: Journal of Applied Behavior Analysis

Volume: 56

Issue: 3

Date: May 2023

DOI: 10.1002/jaba.1005

The study aimed to simplify the subtyping criteria for automatically maintained self-injurious behavior (SIB) and validate the use of a derived level of differentiation (LOD) cutoff as a diagnostic and predictive behavioral marker. The research consisted of three studies: Study 1, Study 2, and Study 3.

Study 1 involved analyzing data from previously published studies to identify subtyping criteria for SIB. The criteria were based on quotient scores and other rules and were used to classify SIB into Subtypes 1 and 2. To determine the optimal LOD cutoff point for distinguishing between the two subtypes, a receiver operating characteristic (ROC) analysis was conducted.

In Study 2, the LOD cutoff obtained in Study 1 was validated by applying it to an independent cohort of clinical cases. This cohort consisted of 21 participants with automatically maintained SIB. The goal was to establish the validity and generalizability of the LOD cutoff by examining its accuracy in classifying the participants into Subtype 1 or Subtype 2. The results showed perfect correspondence between the LOD cutoff and the subtyping criteria in all 21 cases, confirming the accuracy and applicability of the LOD cutoff.

Study 3 focused on evaluating the accuracy of visual analysis, without the use of structured criteria, for identifying automatically maintained SIB. Functional analysis data from 127 cases, including those from previous studies, were analyzed. The LOD cutoff of 62.5% derived from Study 1 was used to classify the cases. The analysis compared the LOD-based classifications with the interpretations made by study authors and clinicians using visual analysis. The findings revealed excellent correspondence between the LOD cutoff and visual analysis interpretations, demonstrating the accuracy of using LOD as a diagnostic behavioral marker for identifying automatically maintained SIB.

The general discussion highlighted the advantages of LOD over quotient scores in simplifying the subtyping process for SIB. LOD directly measures the extent to which SIB is disrupted by alternative reinforcement, providing a more direct and comprehensive measure of SIB dynamics. The LOD cutoff of 62.5% was deemed an excellent diagnostic behavioral marker, capable of accurately classifying SIB into Subtypes 1 and 2. Additionally, LOD showed potential as a predictive behavioral marker, aiding in treatment planning and forecasting response to behavioral interventions.

However, certain considerations were noted. The potential role of social variables in SIB assessment should be assessed and monitored during treatment, as unidentified functions may emerge. The derived LOD cutoff should not be applied blindly and should be interpreted in conjunction with visual analysis, taking into account the overall response patterns and session-by-session data. The LOD cutoff values of 62.5% were specific to the samples analyzed in the study and further replication and research are necessary to establish their generalizability.

In conclusion, the study proposes LOD as a valuable diagnostic and predictive behavioral marker for subtyping automatically maintained SIB. The LOD cutoff of 62.5% simplifies the classification process and has demonstrated accuracy in distinguishing between Subtypes 1 and 2. The findings have implications for treatment planning and intervention development in individuals with SIB. Ongoing research is essential to refine and validate the LOD cutoff values and further understand the complex nature of SIB subtypes.

Leave a Reply

Your email address will not be published. Required fields are marked *