Analysis of counted behaviors in a single-subject design : Modeling of hearing-aid intervention in hearing-impaired patients with Alzheimer's disease

Clinical procedures related to patients with Alzheimer's Disease (AD) largely fail to address the patient's hearing. Given the challenges of this population, unconventional indicators of treatment efficacy may be required. Palmer et al (1999) reported on caregiver-tracked behaviors as outcome measures for hearing aid intervention. Using these data, hearing aid use and subsequent behavior was modeled as a first-order dynamic system, characterized by responses following an exponential time course. The results of such modeling suggest predictable outcomes of hearing aid intervention, or at least useful parameters of quantification (e.g. time-constant and steady-state response), permitting critical assessment of effects of intervention on negative behaviors versus hearing aid use, comparisons among behaviors, and/or comparisons of hearing-aid-use patterns and behavior counts among patients. Use in this and other difficult-to-test populations warrant further study to evaluate clinical efficacy of the analysis described. © 2005 British Society of Audiology, International Society of Audiology, and Nordic Audiological Society.