June 2015
Volume 56, Issue 7
ARVO Annual Meeting Abstract  |   June 2015
Concordance of patient-report and therapist-report estimates of functional ability
Author Affiliations & Notes
  • Theresa Smith
    University of Texas Medical Branch,Galveston, TX, Galveston, TX
  • Lisa Foret
    Evangeline Home Health, Lake Charles, LA
  • Guy Davis
    Evangeline Home Health, Lake Charles, LA
  • Kyoko Fujiwara
    Johns Hopkins, Baltimore, MD
  • Alexis Malkin
    Johns Hopkins, Baltimore, MD
  • Robert W Massof
    Johns Hopkins, Baltimore, MD
  • Footnotes
    Commercial Relationships Theresa Smith, None; Lisa Foret, None; Guy Davis, None; Kyoko Fujiwara, None; Alexis Malkin, None; Robert Massof, None
  • Footnotes
    Support None
Investigative Ophthalmology & Visual Science June 2015, Vol.56, 2107. doi:
  • Views
  • Share
  • Tools
    • Alerts
      This feature is available to authenticated users only.
      Sign In or Create an Account ×
    • Get Citation

      Theresa Smith, Lisa Foret, Guy Davis, Kyoko Fujiwara, Alexis Malkin, Robert W Massof; Concordance of patient-report and therapist-report estimates of functional ability. Invest. Ophthalmol. Vis. Sci. 2015;56(7 ):2107.

      Download citation file:

      © ARVO (1962-2015); The Authors (2016-present)

  • Supplements

Purpose: Evaluate the concordance of measures of functional ability estimated from patient self-report using the Activity Inventory (AI) with those estimated from therapist-report using the Functional Independence Measure (FIM).

Methods: Forty visually impaired home health patients of one occupational therapist participated in this study. Before low vision rehabilitation, patients rated the importance of 50 AI activity goals and rated the difficulty of those goals identified to be at least “somewhat important”. Functional ability was estimated with Rasch analysis of the AI goals rated as at least “somewhat difficult”. AI goal items were anchored to values previously estimated from responses of 3,200 low vision patients. The occupational therapist then rated the same patient-identified AI goals using the FIM without knowledge of the patients’ difficulty ratings. Rasch analysis was performed on the FIM scores with the AI goals anchored to the same calibrated values used for the analysis of patient responses to the AI. Person measures, response probabilities for the 7 FIM categories, and infit mean square distributions were evaluated to determine the validity of FIM-based measures.

Results: FIM category thresholds were ordered except for category “6”, which was used infrequently. Mean square residuals for therapist FIM ratings were distributed as expected by the Rasch model for 33 of the patients, 7 patients were misfitting. The FIM-based person measures are linear with AI-based person measures, but the slope is approximately 2, indicating that the intrinsic variance in the AI-based estimates is about 4 times greater than the intrinsic variance in the FIM-based estimates. The correlation between FIM and AI based measures is 0.57. Relative to average patient response bias, the estimated response bias for the therapist is -0.5 logit (therapist-based estimate of functional ability is consistently less than the patient-based estimate).

Conclusions: Low vision patient functional ability estimated from therapist judgments agree with functional ability estimated from patient reports. Intrinsic variance in the estimated variable is greater for patient-reports than for the therapist report, most likely because of varying biases between patients. The therapist consistently judged patients to have less functional ability than patients judged of themselves.


This PDF is available to Subscribers Only

Sign in or purchase a subscription to access this content. ×

You must be signed into an individual account to use this feature.