4122.0: Tuesday, November 14, 2000 - 1:45 PM

Abstract #8903

Reliability of the SF-36 using proxy measurement in the geriatric population: preliminary results

Thomas E. Wasser, PhD, Department of Community Health and Health Studies, Lehigh Valley Hospital and Penn State School of Medicine, 227 N. 17th Street, Allentown, PA 18104, 610-402-2497, Thomas.Wasser@LVH.com

Introduction: Gathering functional status information is important for use in outcomes studies. These studies are increasingly conducted on geriatric populations in both the inpatient and outpatient settings. Often the patient in these studies has difficulty responding to these questionnaires by themselves. We conducted this study to determine if proxy respondents could provide reliable data for the patient which is unable to respond using a common outcome tool the Medical Outcomes Study Short Form-36 (SF-36). Methods: This preliminary analysis was conducted on the first 15 pairs of subjects who completed the SF-36 in both inpatient and outpatient settings. Analysis was the Pearson's correlation (r) between patient and proxy, using a correlation of r=>0.80, for minimum evidence of reliability. Results: The patient population consisted of 8 males, 7 females (mean age 69.4 years). Proxy questionnaires were completed by the spouse (11/15, 73%), child or grandchild (3/11, 20%), or sibling (1/15, 7%). The correlations for the eight SF-36 subscales were; Health Perception r=0.282, Physical Functioning r=0.418, Role Limits in Physical Functioning r=-0.072, Role Limits in Emotional Health r=-0.305, Social Functioning r=0.113, Mental Health r=0.490, Pain r=0.718, Energy and Fatigue r=0.705. Conclusions: The preliminary results indicate two primary findings; none of the subscales achieved correlations that would indicate proxy questionnaires are reliable, and directly observable traits such as Pain, and Energy/Fatigue show much higher correlations than do non-observable traits such as Role Limits in Emotional Health. Based on this preliminary data the use of proxy SF-36 is not recommended.

Learning Objectives: 1) Describe the problem of data collection as it often occurs in the geriatric population. 2) Outline procedures for questionnaire completion by proxy. 3) Describe the problems with data collection and reliability of proxy measurement

Keywords: Assessments, Outcomes Research

Presenting author's disclosure statement:
Organization/institution whose products or services will be discussed: None
I do not have any significant financial interest/arrangement or affiliation with any organization/institution whose products or services are being discussed in this session.

The 128th Annual Meeting of APHA