A Medical Assistant-Based Program to Promote Healthy Behaviors in Primary Care

• Proportion of intervention-group vs. control group patients with behavioral risks at follow-up (individual behaviors separately and combined behaviors).
  

• Total number of health risks identified and interventions ordered by medical assistants (measure program impact).
  
• Rate of behavioral risks identified by MA's (risks identified/patients assessed).
  
• Rate of interventions ordered by MA's (interventions ordered/patients assessed).
  
• Time-motion impact of intervention on medical assistants.
  
• Impact on MA satisfaction (qualitative interviews).
  
• Cost of intervention.
  

Unhealthy behaviors are the most common actual causes of death in the United States. The high prevalence of behavioral risk factors in primary care calls for a response of appropriate scale. One strategy to achieve a greater scale of delivering interventions is to relieve the bottleneck of physician assessment and management. Creating a system that engages other members of the primary care team to link patients who are ready to change with practice, health system and community resources is a mechanism to amplify the practice's impact and thus achieve the necessary scale.

The project evaluates a model of primary care that manages population risk with appropriate expertise at the necessary scale. We test the effectiveness of a medical assistant-based program to link patients with behavioral risk factors to interventions at the practice, health system, and community levels. We hypothesize that intervention-group patients will achieve higher rates of behavior change than control patients.

The study is a controlled trial in the PRENSA network, which includes six urban practices serving a disadvantaged Latino population. Using data from a health risk assessment (HRA) routinely collected in PRENSA practices, medical assistants will assess 4500 patients' behavioral risks and apply behavior-specific "assess-advise-agree-assist-arrange" algorithms to engage behavioral interventions in the practices, health system, and local public health department. Outcomes data will be collected at 6-9 month follow-up with repeated completion of a standardized health risk assessment. The primary analyses will compare outcome measures for smoking, problem drinking, sedentary lifestyle, and unhealthy diet in intervention and control patients.

We anticipate the study will show a medical assistant-based program to address smoking, risky drinking, sedentary living, and unhealthy diet in primary care to be feasible and effective. We will evaluate program effectiveness by assessing the number of patients reached, interventions requested and completed, impact on medical assistant workflow and satisfaction, costs to the practices, and impact on health behaviors.