Concurrent Validity of a Physical Activity Vital Sign Used in an Adult Preventive Cardiology Clinic

In clinical settings, counseling patients on physical activity starts by assessing patients' current physical activity levels. Self-report measures of PA are generally easy to administer; however, they may be too long to be convenient and are known to correlate poorly with objective measures of physical activity.

To assess the concurrent validity of a self-report three-question physical activity vital sign with objective Fitbit step counts and the distance walked during a 6-min walk test.

This pilot study tested a best practice advisory embedded in the Epic electronic health record, which was designed to prompt providers in a preventive cardiology clinic to counsel patients reporting low levels of physical activity. Patients were invited to participate in the remote patient monitoring phase to assess the change in their physical activity by wearing a Fitbit for 12 weeks and completing a 6-min walk test at baseline and 12 weeks. This analysis used the cross-sectional data collected in this phase. Pearson correlations were conducted between self-reported physical activity, Fitbit step counts, and the distance walked during the 6-min walk-a measure associated with current physical activity levels. Kappa coefficients were calculated to assess agreement between self-reported physical activity and step counts.

Participants who enrolled in the Fitbit monitoring were approximately 50% female, with the majority identified as White non-Hispanic adults. Their most common cardiovascular risk factor was hypertension. The self-reported physical activity vital signs were significantly associated with step counts at baseline and 12 weeks but were not associated with the distance during the 6-min walk test. However, the distance walked was significantly associated with step counts at baseline and 12 weeks. The Kappa results demonstrate a poor level of agreement between two categories (meeting or not meeting current physical activity guidelines) of self-report physical activity vitals and the objective Fitbit step counts.

There were moderate correlations between the self-reported physical activity vital signs and the Fitbit step counts, but there was lack of agreement when they were categorized. Further validation of this physical activity vital sign is warranted.