Estimated Cases Averted by COVID-19 Digital Exposure Notification, Pennsylvania, USA, November 8, 2020–January 2, 2021
Seonghye Jeon, Gabriel Rainisch, A‐Mac Harris, Jared Shinabery, Muneeza Iqbal Rai, Amar Pallavaram, Stacy Hilton, Saugat Karki, Patrick K. Moonan, John E. Oeltmann, Martin I. Meltzer · Emerging infectious diseases · 2023
C ase investigation and contact tracing (CICT) was a pillar among COVID-19 prevention strategies, especially before vaccine availability (1,2).However, standard CICT relies on staff to reach cases and close contacts, which is labor intensive, and CICT programs often become overwhelmed when caseloads surge (3-5).Standard CICT also relies on case investigation interviews to identify contacts; thus, it is prone to recall and participation bias and might not identify all potential exposures, such as interactions between strangers in public spaces.COVID-19 exposure notification smartphone applications (apps) can alleviate those challenges by automatically notifying app users when they have been near other users who reported positive SARS-CoV-2 results (herein referred to as cases).Pennsylvania, USA, and 26 other states implemented digital exposure notifications to complement their standard CICT programs (6).However, few studies have evaluated the effectiveness of digital notifications in the United States (6,7).We estimated the number of cases and hospitalizations averted by Pennsylvania's digital notification system, COVID Alert PA app.We also investigated strategies to increase the system's efficiency and its effects on the estimated number of cases and hospitalizations. The StudyDuring case investigation interviews in Pennsylvania, digital notification app users were identified and given a validation code to enter into their app.The app then automatically sent anonymous notifications to other users identified through smartphone Bluetooth technology as potentially exposed to the person testing positive for COVID-19 (Appendix, https://wwwnc.cdc.gov/EID/article/29/2/22-0959.App1.pdf).The Pennsylvania Department of Health (PA DoH) collected data on the performance of standard CICT and digital notification apps (Table ).We aggregated those data across all counties, excluding Philadelphia County (Appendix), for 8 weeks, November 8, 2020-January 2, 2021 (Table ).We extracted the daily number of COVID-19 cases from the Centers for Disease Control and Prevention (CDC) COVID Data Tracker (8).We used CDC's COVIDTracer modeling tool to estimate cases and hospitalizations averted by digital notifications during the 8-week study period (1,2,9).COVIDTracer uses an epidemiologic model to illustrate the spread of COVID-19 and effects of CICT and other nonpharmaceutical interventions (NPIs).We calculated a summary effectiveness measure for CICT and digital notification apps from the various data PA DoH collected and input this measure to the model (Table).We defined this summary effectiveness measure as the proportion of cases that entered isolation and contacts that quarantined in response to CICT Estimated Cases Averted by COVID-19 Digital Exposure