1. Bishop, Mark D. PT, PhD
  2. Romero, Sergio PhD
  3. Patterson, Tara S. PhD
  4. Singletary, Floris F. MS
  5. Light, Kathye PT, PhD


Background and Purpose: One method for defining successful rehabilitation outcomes is to use a threshold of performance on a clinical measure. Patients also have their own criteria for success. The aim of this study was to examine the association between clinical measures and patient criteria for determining successful intervention.


Methods: Twenty-two participants participated in a 12-week intervention program for balance disorders. Participants were tested using the Berg Balance Scale (BBS) and Dynamic Gait Index (DGI) at the first and final visits. They also used the Patient's Perspective Outcomes Questionnaire (PPOQ) at the first visit to rate impairments in the core domains from the International Classification of Function. Participants rated their usual levels of impairment across domains, their desired levels of impairment after intervention (success criteria), and how important it was to change in each domain. At the final visit, all participants used the PPOQ to rate their current impairment across domains and completed a 15-point global rating of change (GRC). The intervention was defined as a success if the participant rated himself or herself as "a great deal" or "a very great deal" better using the GRC. Changes in the BBS and the DGI were dichotomized on the basis of whether the change exceeded the minimal detectable change (MDC) of the instrument. Participants' ratings of impairment at 12 weeks were compared with their success criteria and dichotomized on the basis of whether the success criteria had been met. The Freidman test was used to test differences across domains at baseline with Wilcoxon tests for follow-up. Wilcoxon tests were also used to examine the importance of change across domains. Chi-square tests were used to explore the association among the GRC, performance on clinical tests, and whether success criteria were met.


Results: Significant changes were noted for the group for both BBS (P < .001) and DGI (P = .006). Only 40% of participants exceeded or met their desired or expected change in impairment for mobility. No significant associations were noted between exceeding MDC on the BBS or DGI and any of the domains on the PPOQ (all Ps = .263). Bivariate associations were noted between baseline measures of psychological factors and many of the participant ratings of impairment. Fifty-five percent of participants indicated that they were a great deal better or a very great deal better. Significant association was identified between success and exceeding MDC on the BBS ([chi]2 = 5.84, P = .016) but not the DGI. When considering participants' desired change in impairment, only meeting desired changes in mental function was associated with considering the intervention a success ([chi]2 = 4.55, P = .033).


Discussion: Treatment success from the perspective of this group of older adults was related to improvement on a clinic-based performance measure and not the participants' intrinsic success criteria assessed at the beginning of the intervention.


Conclusion: Making measurable change in gait and balance translates into perceived success for the patient with a balance disorder.