PURPOSE
Primary care clinicians have a unique opportunity encourage to identify health risks in their patients and to healthy behaviors, such as smoking cessation, physical activity, proper nutrition, and moderation in the use of alcohol. Yet, even brief evidence-based interventions are inconsistently used by primary care clinicians.1 The 5 A’s model (ask, advise, assess, assist, and arrange) is a tool to assist clinicians in asking patients about their health behaviors and, if patients are found to be at risk, advising them to modify their behavior, assessing their interest in doing so, assisting in their efforts to change, and arranging appropriate follow-up.2,3 This article describes baseline data from a larger interventional study testing a nurse consultation model for improving health promotion in primary care practices, and presents the implications of moving the concept of the 5 A’s into systematic interventions for multiple unhealthy behaviors. A transition from an organizational construct to a set of evidence-based 5 A tools is essential for primary care clinicians to assist their patients with health behaviors as suggested in the Healthy People 2010 goals for the nation.4
METHODS
Practices in the study selected their own goals for practice improvement based on the 5 A’s for 1 or more of the specified health behaviors. Twenty Michigan family medicine and general internal medicine practices belonging to the Great Lakes Research Into Practice Network participated: 10 were rural, hospital-owned practices (69 clinicians), and 10 were metropolitan, largely independent practices (25 clinicians). Practices were provided several services, including determination of their current preventive care activities, identification of practice improvement opportunities, assistance with planning interventions, and support for implementation and evaluation.
Data collection methods included chart audits and nurse-consultant field notes based on practice observations and interviews with practice clinicians and staff, including their awareness and use of the 5 A’s model. A 3-month time frame was identified for the preintervention audit, and all visits by adults for chronic disease (hypertension, cardiovascular disease, diabetes) and health maintenance were selected for that period. Fifty charts per practice were randomly selected for audit, and 1 reference visit was audited per selected chart. Trained nurses conducted the chart audits using a specifically designed chart audit tool and identified guidelines for determination of each of the A’s. Relevant items from the reference visit progress note, visit-related laboratory and radiograph reports, and flow sheets were noted and copied verbatim onto the audit form. As a quality assurance measure, the principal investigator reviewed all of the chart audit forms for consistency across abstractors.
LESSONS LEARNED
From the nurse-consultant field notes, we discovered it was rare that anyone in the practice, other than the physicians, had ever heard of the 5 A’s. Most of the clinicians who participated in our study were aware of the 5 A’s as an organizational construct, but none used it as a functional tool in practice. The rate of documentation of the 5 A’s in the medical records at baseline varied by practice and by behavior (Table 1⇓). When present, ask was usually documented on a flow sheet completed at a previous visit. Clinicians most frequently asked about tobacco use and smoking but intervened only in about one third of the patients at risk. Queries about diet were least frequently documented, but dietary interventions were most commonly documented for those at risk. More of the 4 A’s beyond ask were documented for dietary interventions than for any of the other behaviors. Dietary interventions were typically documented in the reference visit progress note and were often paired with recommendations for increased physical activity. Advice was the most commonly documented of the remaining 4 A’s for each behavior. For most practices in the study, documentation of assess, assist, or arrange was minimal. It appeared from the documentation that most clinicians were unaware of the importance of assessing the patient’s readiness to change as a key step in assisting with behavioral change and arranging appropriate follow-up.
DISCUSSION
The 5 A’s have been widely used in high-quality, controlled clinical trials in tobacco cessation and brief primary care interventions for a variety of behaviors.3,5–7 Unfortunately, there are few validated, functional supportive resources for screening, monitoring, and intervening for unhealthy behaviors.4,8 Current research also indicates that successful practice interventions involve systematic processes using multiple members of the practice team, and not just relying on the physician alone.9–11 This finding has important implications for the practical use of the 5 A’s tool in that practice staff members must be skilled in its use as well.
Chart audits may either overreport or underreport actual clinician behavior. Previous research shows that chart documentation often underestimates what actually occurs during the office visit.12 Had the actual intervention rates by clinicians in the present study been double those found in the medical records, however, the level of intervention for assess, assist, and arrange would fall short of the ideal. Because assist and arrange were the least frequently documented of the 5 A’s, patients who are ready to change would receive little help in reaching their goals. On the other hand, simple advice, assistance, and follow-up arrangements in the absence of an appropriate assessment of a patient’s readiness to change may not be effective in changing patient behavior. The chart audit does not capture the dynamic nature of the 5 A’s and thus may overstate the effectiveness of an intervention (ie, produce a false-positive result).
There are additional limitations to this study. The sample of practices selected for this study may not be representative of other primary care practices in their use of the 5 A’s for these behaviors. Some arbitrary decisions were made in coding each A within the 5 to maintain consistency in coding. These decisions may not reflect the true intervention delivered to the patient. Finally, these data represent documentation rates, not rates of actual delivery of services, which may have been higher or lower.
CONCLUSIONS
Despite widespread support for their use, the 5 A’s remain an organizational construct with limited practical penetration into the primary care practices in this study and most likely other primary care practices nationwide. This study suggests that practices need clinical staff with a better understanding of the 5 A’s framework, including brief assessment strategies, systematic approaches to the use of the 5 A’s supported by the whole practice team, and development of supportive resources such as chart prompts and other practical tools.
Acknowledgments
We gratefully acknowledge the Robert Wood Johnson Foundation (especially the National Program Office of the Prescription for Health Initiative), the Agency for Healthcare Research and Quality, the Prescription for Health analysis team, consultant Leif Solberg, and the participating health care systems, physicians, and staff of Marquette General Health System and Genesys Health System.
Footnotes
-
Conflicts of interest: none reported
-
Funding support: This project was supported by Prescription for Health (grant No. 49057), a national program of The Robert Wood Johnson Foundation with support from the Agency for Healthcare Research and Quality, and the Michigan Department of Community Health, Lansing, Mich.
Presented in part at the 32nd Annual Meeting of the North American Primary Care Research Group, Orlando, Fla, October 10–13, 2004.
- Received for publication December 20, 2004.
- Revision received February 27, 2005.
- Accepted for publication March 16, 2005.
- © 2005 Annals of Family Medicine, Inc.