The 131st Annual Meeting (November 15-19, 2003) of APHA

Abstract #63412

Karen Lutfey, PhD and Jonathan Ketcham, PhD. Robert Wood Johnson Foundation Scholars in Health Policy Research Program, University of California--Berkeley, Warren Hall 140, MC 7360, Berkeley, CA 94720, (510)643-0920, lutfey@uclink.berkeley.edu

Variations in treatment patterns are thought to be determined primarily by physician beliefs about the efficacy of the options for treatment. For many medical conditions, chronic conditions in particular, efficacy depends in part on patient adherence to the prescribed treatment regimen. Consequently, the clinical decision-making may depend on the physician perception of patient adherence.

Extensive research demonstrates that patients' social characteristics influence physician behavior, specifically diagnostic and treatment decisions. Other work has suggested that these correlations between patient characteristics and medical treatment are more accurately predicted by provider characteristics. Despite its potential relevance for understanding socioeconomic and race-related health disparities, relatively little attention has been given to the processes by which providers assess patients with regard to social characteristics. Recent work has considered the effect of patient race and socioeconomic status on physicians' perceptions of patients (van Ryn and Burke, 2000), and finds that physicians perceive African Americans and members of low socioeconomic classes relatively negatively on a variety of dimensions compared to white patients; race is also associated with physician assessments of patient intelligence and beliefs about risky behavior and adherence to treatment regimens. We pursue a similar line of analysis, but extend it by analyzing provider assessments of patients as well as patient self-assessments and differences between the two.

We use a unique dataset collected from two diabetes clinics from a large Midwestern city to compare the determinants of physician perception of adherence for an individual patient with the patient's self-reported adherence. We examine the role of race, gender, age, education, and perceived intelligence in explaining both the levels of the physician and patient assessment and the differences between those assessments for 173 patients. We consider overall control of diabetes, overall adherence, adherence to diet, and adherence to medications.

In addition to corroborating earlier evidence that these characteristics influence provider assessments of adherence, we also find that these characteristics result in systematic differences between patient and physician assessments. This suggests that providers use these observable characteristics to infer patient's adherence, but their inferences are systematically biased. Consequently, providers may infer that the efficacy of treatment regimen varies by these characteristics, and they may choose different courses of treatments based on these perceptions alone. Thus even a provider concerned only with patient health may erroneously choose different treatments for different types of patients, regardless of health status or other clinical differences.

Learning Objectives:

• List differences in provider assessments of patient compliance and patients' self-assessments.
• Articulate the importance of compliance assessment for diabetes care.
• Identify potential opportunities for bias in patient assessment and challenge biased assessments.

Presenting author's disclosure statement:
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.