Positive Registration Rate as a Key Determinant of COCOA Effectiveness: Empirical Evidence from Individual-Level Key-Match Data during the Sixth and Seventh COVID-19 Waves in Japan

BackgroundCOCOA, Japans Bluetooth-based COVID-19 contact tracing app, was widely regarded as ineffective due to persistently low key-match counts. However, this assessment may have conflated two distinct phenomena: (1) a structurally suppressed positive registration rate caused by administrative friction in the HER-SYS linkage, and (2) genuine epidemiological inefficacy. ObjectiveTo empirically examine whether the correlation between individual COCOA key-match counts and regional COVID-19 case numbers depended on positive registration rate, using a unique longitudinal dataset from a single observer with a rigorously controlled behavioral pattern. MethodsThe corresponding author (S.N.) recorded daily key-match counts from his personal iPhone from January 10 to October 8, 2022, encompassing the Sixth Wave (January 10-April 20, 2022) and Seventh Wave (July 9-September 2, 2022). Daily reported COVID-19 cases in Tokyo were obtained from publicly available NHK data. Pearson correlation coefficients were calculated for each wave period separately. ResultsDuring the Sixth Wave, no meaningful correlation was observed between key-match counts and daily case numbers (r2 = 0.018, p = 0.059, n = 194). In contrast, during the Seventh Wave, a strong positive correlation emerged (r2 = 0.530, p < 0.001, n = 56). This correlation disappeared abruptly after September 12, 2022, coinciding with Japans revision of the mandatory full case reporting (Zenshu Todokedashi) policy, which substantially reduced positive registrations in COCOA. ConclusionsCOCOAs utility as an individual infection risk indicator was critically dependent on positive registration rate rather than app installation rate. These findings provide the first real-world empirical evidence supporting the threshold effect predicted by prior simulation studies, and offer important lessons for the design of digital tools in future pandemic preparedness.