Health outcomes of Fitbit, Garmin or Apple Watch-based interventions

Abstract

Introduction: A rising number of systematic reviews of smartwatch clinical effectiveness have been published; however, the findings are conflicting and need further investigation. Our purpose was to assess the methodological and evidence synthesis quality of meta-analyses of the therapeutic efficacy of Fitbit, Garmin, or Apple Watch-based interventions. Material and Methods: Systematic searches were carried out with three electronic databases from inception to October 2021. Full text systematic reviews originally published in English in peer-reviewed journals describing meta-analyses of all clinical outcomes of Fitbit-, Garmin-, or Apple Watch -based interventions were eligible for this study. The Amstar-2 scale and GRADE were used to assess the methodological and evidence synthesis quality, respectively. Results: One study with low methodological quality (overall score of 13.5) was identified, which contained five outcomes. None of the outcomes met the GRADE criteria for high-quality evidence (0%). Two outcomes (40%) had moderate-quality evidence synthesis, while three outcomes (60%) had low-quality evidence synthesis. The most prevalent downgrading GRADE factor was a risk of bias (100%), followed by imprecision (80%), and publication bias (60%). Conclusion: Despite the widespread adoption of smartwatches, there is a large gap in the literature as there is limited evidence. This review suggests that further clinical trials and meta- analyses are needed and it formulates research recommendations.