Dissertation Defenses

Doctoral Dissertation Announcement

---

Candidate: Jonathan C. Baker

Degree of: Doctor of Philosophy

Department: Psychology

Title: Acceptability of Interventions to Staff in Long Term Care Settings for Older Adults: Comparing Ratings and Hierarchical Selection

Committee:
Dr. Linda A. LeBlanc, Chair
Dr. James E. Carr
Dr. R. Wayne Fuqua
Dr. Leilani Feliciano

Date: Friday, December 12, 2008 1:00 p.m. - 3:00 p.m.
1509 Wood Hall

Abstract:
Older adults and their caregivers generally prefer behavioral interventions over medications in treatment acceptability studies (Osterkamp, Mathews, Burgio, & Hardin, 1997). However, previous acceptability studies have primarily examined ratings, which did not force the responder to select between treatment options. Additionally, recent advances in behavioral treatment technologies (Carr & LeBlanc, 2003) and pharmacotherapy (Schneider, 1999) create a need to revisit treatment acceptability for older adults. The present investigation examined treatment acceptability of behavioral, pharma-cological, and sensory interventions using a treatment acceptability rating scale, treatment selections, and direct report of treatments used in a six-month window. Fifty-three staff from nursing homes in the Midwestern United States who actively participate in treatment decisions (e.g., directors of nursing, program directors, case managers) participated in the study. One-way ANOVAs indicated no statistically significant differences in acceptability ratings for three interventions. Treatment selections were significantly different (f = 33.485, p < .000), with behavioral, sensory, and medical selections being selected significantly more often than referral and selecting no treatment. Pearson-Product Moment Correlations indicated that ratings and selections were not significantly related. Point Bi-Serial correlations indicated behavioral intervention selections (but not ratings) were significantly correlated with using a second behavioral treatment if the first behavioral treatment was ineffective (r = 0.460, p = 0.014). Thus, unlike previous research, treatment ratings did not result in significant differentiation between treatments and this finding was confirmed with treatment selections. Of further significance was the fact that treatment selections, which were significantly correlated with actual treatment use for at least one treatment, were not significantly correlated with treatment ratings. Although this study does not permit an analysis of whether ratings or selections are a better predictor of treatment adoption, these results provide initial evidence that treatment ratings, hereto a staple of treatment acceptability and treatment adoption research, may not be an accurate measure of the behaviors they are employed to sample.