Card studies for observational research in practice.

PURPOSE: Observational studies that collect patient-level survey data at the point-of-care are often called card studies. Card studies have been used to describe clinical problems, management, and outcomes in primary care for more than 30 years. In this article we describe 2 types of card studies and the methods for conducting them. METHODS: We undertook a descriptive review of card studies conducted in 3 Colorado practice-based research networks and several other networks throughout the United States. We summarized experiences of the State Networks of Colorado Ambulatory Practices and Partners (SNOCAP). RESULTS: Card studies can be designed to study specific conditions or care (clinicians complete a card when they encounter patients who meet inclusion criteria) and to determine trends and prevalence of conditions (clinicians complete a card on all patients seen during a period). Data can be collected from clinicians and patients and can be linked. CONCLUSIONS: Card studies provide cross-sectional descriptive data about clinical care, knowledge and behavior, perception of care, and prevalence of conditions. Card studies remain a robust method for describing primary care.

Prescription for health: changing primary care practice to foster healthy behaviors.

PURPOSE: The leading causes of premature death in the United States are linked to 4 behaviors: smoking, unhealthy diet, physical inactivity, and risky alcohol use. We report lessons from 17 exploratory projects funded under Prescription for Health that tested the feasibility of innovative behavior change strategies for at least 2 of these behaviors in primary care practices. METHODS: Seventeen practice-based research networks (PBRNs) implemented and evaluated tools, cues, and techniques in 120 family medicine, internal medicine, pediatric, and nursing practices across an ethnically diverse sample of adults, children, and adolescents in rural and urban settings. We reviewed progress reports and notes from site visits and 3 meetings to generate overarching lessons. RESULTS: PBRNs successfully implemented their projects in diverse practices despite reported logistical challenges and practice constraints. The networks showed that distributing the effort across the care team and throughout the practice and community is possible. Although each behavior required specific attention, each did not require its own separate staff and system. Three models emerged as helpful guides for the comprehensive redesign of health behavior counseling, but they require adaptation for use in real-world primary care settings. Traditional methods of collaboration yielded mixed results, making obvious a need for dedicated collaboration funds and a better framework to identify and align high-yield opportunities. CONCLUSIONS: These projects confirm the feasibility of health behavior counseling in primary care practice. They also highlight the need for substantive practice redesign, and the value of models and frameworks to guide redesign and collaborative efforts.

Common measures, better outcomes (COMBO): a field test of brief health behavior measures in primary care.

BACKGROUND: Primary care offices have been characterized as underutilized settings for routinely addressing health behaviors that contribute to premature death and unnecessary suffering. Practical tools are needed to routinely assess multiple health risk behaviors among diverse primary care patients. The performance of a brief set of behavioral measures used in primary care practice is reported here. METHODS: Between August 2005 and January 2007, 75 primary care practices assessed four health behaviors, using a 21-item patient self-report questionnaire for adults or a 16-item questionnaire for adolescents. Data were collected via telephone, paper, or electronic means, either with or without assistance. The performance of these measures was evaluated by describing risk-behavior prevalences, combinations of risk behaviors, and missing data. RESULTS: Of 227 adolescents and 5358 adults, most patients completed all of the survey questions. Two or more unhealthy behaviors were reported by 47.1% of adolescents and 69.2% of adults. Percentages of adults who completed all the survey items varied by health behavior: tobacco use, 98.5%; diet, 98.2%; physical activity, 96.2%; alcohol use, 85.1%. Missing data rates were higher for unassisted patient self-reporting. CONCLUSIONS: A relatively brief set of health behavior measures was usable in a variety of primary care settings with adults and adolescents. The performance of these measures was uneven across behaviors and administration modes, but yielded estimates of unhealthy behaviors consistent overall with what would be expected based on published population estimates. Further work is needed on measures for alcohol use and physical activity to bring practical assessment tools for key health behaviors to routine primary care practice.