Published analysis of contraceptive effectiveness of Daysy and DaysyView app is fatally flawed

Abstract

BACKGROUND: In March 2018, Dr. Martin C. Koch and colleagues published an analysis purporting to measure the effectiveness of the Daysy device and DaysyView app for the prevention of unintended pregnancy. Unfortunately, the analysis was flawed in multiple ways which render the estimates unreliable. Unreliable estimates of contraceptive effectiveness can endanger public health. MAIN BODY: This commentary details multiple concerns pertaining to the collection and analysis of data in Koch et al. 2018. A key concern pertains to the inappropriate exclusion of all women with fewer than 13 cycles of use from the Pearl Index calculations, which has no basis in standard effectiveness calculations. Multiple additional methodological concerns, as well as prior attempts to directly convey concerns to the manufacturer regarding marketing materials based on prior inaccurate analyses, are also discussed. CONCLUSION: The Koch et al. 2018 publication produced unreliable estimates of contraceptive effectiveness for the Daysy device and DaysyView app, which are likely substantially higher than the actual contraceptive effectiveness of the device and app. Those estimates are being used in marketing materials which may inappropriately inflate consumer confidence and leave consumers more vulnerable than expected to the risk of unintended pregnancy. Prior attempts to directly convey concerns to the manufacturer of this device were unsuccessful in preventing publication of subsequent inaccurate analyses. To protect public health, concerns with this analysis should be documented in the published literature, the Koch et al. 2018 analysis should be retracted, and marketing materials on contraceptive effectiveness should be subjected to appropriate oversight.