Piloting an Intervention to Provide Tailored Feedback on Health Behaviors to Adolescents in Pediatric Primary Care

Background: The pediatrician’s office frequently provides the first opportunity for behavioral health intervention. However, pediatricians are limited in time, tools, and training to assess and treat behavioral health problems. Even a brief intervention Motivational Interviewing intervention more than doubles the length of a physical exam appointment. Studies investigating the usefulness of technology to assist with behavioral health interventions in pediatric primary care are limited, but one study found that technology was a feasible approach for use in pediatric primary care (Harris et. al., 2012). That study assessed for alcohol with computer technology and provided the findings to the attending physicians who then delivered the MI intervention (Harris et. al., 2012). A next step approach would be to include a computerized intervention. One such computer intervention is by delivering electronic feedback messages targeting a specific behavior, such as smoking. Additionally, tailored feedback interventions based on unique characteristics of an individual have been demonstrated to be more successful than generic informational feedback (Kreuter, 1999; Noar, 2007). Another innovative approach to creating feedback is may be to take a harm reduction approach with adolescents to encourage an increase in healthy behaviors rather than solely focus on discouraging risk behaviors (Mauriello, 2010; Velicer, 2013). Furthermore, tailored feedback messages are more successful when supported by an underlying theory of behavior change. For example, The Transtheoretical Model (TTM), which is based on the decision making of an individual for intentional change, identifies change as a process involving progress through a series of stages primarily seen as related to Decisional Balance and Self-efficacy. In sum, the best evidence tells us that a theory-based tailored feedback intervention using computer technology evokes successful behavior change in a manner that is feasible for in primary care (Noar, 2007). To date, no one study has completely integrated behavioral assessment with a feedback intervention based on the TTM in a pediatric primary care setting. The primary aim of this project was to use a step-by-step approach to Program Evaluation to develop, pilot, and test the feasibility of a computerized assessment of behaviors followed by brief stage-tailored feedback promoting health behaviors to patients and physicians in a pediatric primary care setting. Methods: Key informant interviews were conducted to engage key stakeholders. Pediatricians and staff were asked to discuss their normal standard of care to understand how best to integrate the program into the office practice, and to gather input into the development of the computerized assessment and feedback system. A literature review of validated measures was conducted to construct an assessment measure and tailored feedback. The behavioral constructs that were found to be most prevalent in the literature and of concern to pediatricians were included (alcohol, marijuana, nicotine use, caffeine intake, sleep habits, disordered eating behavior, exercise, and stress management). Tailored feedback was based on two specific targeted unhealthy behaviors (alcohol use and marijuana use) as well as two specific healthy behaviors (stress management and exercise) to increase healthy activities for all participants. Microsoft Access software was chosen based on the requirements of the Information Technology team for the pediatric office. A computer programmer was hired to load the program titles Multiple Assessment Symptom Checklist Of Teens (MASCOT) onto the tablet. This project IRB approved. Results: Patients between the ages of 13 and 21 years visiting Narragansett Bay Pediatrics for a well-visit were recruited to participate in this study. One pediatrician in a large pediatric office and her patients (N=55 total) participated in this study. Patients between the ages of 13 and 21 years visiting Narragansett Bay Pediatrics for a well-visit were recruited to participate in this study. Participants completed a cognitive assessment (n=10) to test the system for time duration of administration, to ensure that instruction sets and content of feedback were understood, evaluated concerns with user interface, and programming errors, typos, and other minor edits were made to the system. A second sample (n=8) was assessed in the same manner and minor adjustments were made. The remaining participants (n=37) used the revised system. Patients denied any difficulty using the system. They reported that the feedback was helpful and the system prompted them to talk to the doctor about something they might not have otherwise. Patients reported high alcohol use (57%) and marijuana use (46%). Physicians reported that they intervened on all behaviors except stress management. Two of the interventions included the patient’s parents. There were no statistically significant differences in behavior change, but it did appear that some patients moved from precontemplation to contemplation and preparation for the reduction of alcohol use