Abstract
Objective: To evaluate evidence on the effectiveness of behavioral interventions using single-case experimental design (SCED) methodology and to identify behavioral interventions with sufficient evidence for possible inclusion in the development of guidelines for the management of challenging behaviors in adults following moderate to severe traumatic brain injury (TBI).
Methods: As a subinvestigation of a larger systematic review process designed to identify evidence for guidelines development, the current review focused on studies using SCED methodology applied to persons with challenging behaviors following moderate to severe TBI. Articles were identified from a search of the published literature through January 2021, identifying studies in CINAHL, Cochrane Database of Systematic Reviews, EMBASE, MEDLINE/Ovid, and PsycINFO. Articles meeting inclusion criteria were assessed for design rigor to allow for effect size determination. The identified cases were then critically appraised using the RoBiNT (Risk-of-Bias in N-of-1 Trails) Scale to determine strength of evidence for causal inference.
Results: Thirty-four studies met inclusion criteria, with a total of 44 cases evaluated for effect of the treatment intervention on defined target behaviors. Seventeen cases had effect sizes rated as large, 22 cases as medium, 3 cases as small, and 3 as no effect. An observed trend was for large and medium effect sizes to be associated with lower RoBiNT Scale internal validity scores. Randomization, blinded provider and assessor, and assessment of treatment adherence were the internal validity items unlikely to meet criteria.
Conclusions: SCED methodology was found to produce large and medium effect sizes for behavioral interventions targeting challenging behaviors following moderate to severe TBI. However, the strength of the evidence is limited because of weaknesses in study designs. Most of the studies failed to meet established internal validity criteria designed to reduce risk of bias in SCED studies as such rigor is difficult to establish or often not practical in clinical settings. Suggestions and recommendations are outlined for improving the quality of published cases using SCED methodology, which, in turn, will improve credibility of evidence and better inform the development of treatment guidelines for behavior regulation.