Digital measurement of SARS-CoV-2 transmission risk from 7 million contacts

The study titled “Digital measurement of SARS-CoV-2 transmission risk from 7 million contacts” offers significant insights into the dynamics of COVID-19 transmission and the efficacy of digital contact tracing.

Here’s a detailed summary:

  1. Study Overview: The research analyzed 7 million contacts notified by the NHS COVID-19 app in England and Wales. The aim was to determine how app-measured proximity and duration of exposure translated into actual transmission of SARS-CoV-2.
  2. Methodology: The NHS COVID-19 app in England and Wales, active on millions of smartphones, recorded proximity and duration of exposure using a privacy-preserving framework. This study focused on exposure notifications from April 2021 to February 2022, involving 23 million hours of cumulative exposure and 240,000 positive tests reported post-notification.
  3. Key Findings:
    • A strong correlation was found between app-computed risk scores and the actual probability of transmission.
    • Longer exposures at greater distances posed similar risks to shorter exposures at closer distances.
    • The probability of transmission, confirmed by a reported positive test, initially increased linearly with exposure duration (1.1% per hour) and continued to rise over several days.
    • While most exposures were short, transmissions typically resulted from longer exposures (median 6 hours).
    • Households accounted for about 6% of contacts but 40% of transmissions.
  4. Risk Assessment Method: The app assessed risk by partitioning the full exposure event into 30-minute ‘exposure windows’, calculating a risk score based on proximity, duration, and infectiousness. This risk assessment method proved accurate in predicting the likelihood of transmission.
  5. Probability of Infection: The study found an increasing probability of reported infection as the maximum risk score increased. The cumulative risk score and the total duration of exposure were more discriminatory than the maximum risk score in predicting infection probability.
  6. Contributions to Overall Risk: The overall risk was determined by contributions from each exposure window and background risk. The probability of reported transmission per window was proportional to the app’s risk score, validating the app’s risk calculation method.
  7. Classification of Contacts: Contacts were classified into four categories to reflect different contexts, such as household and non-household contacts. Household contacts, though a small percentage of the total, accounted for a substantial portion of transmissions due to their longer duration and closer proximity.
  8. Effectiveness of Predictors: The study demonstrated that duration and background risk were effective predictors for binary risk classification. This has implications for optimizing public health strategies, like using ‘amber’ notifications for intermediate-risk contacts as an alternative to quarantine.
  9. Study Limitations:
    • Lack of contextual data on exposures, such as setting, immunity level, and ventilation.
    • Exclusion of very low-risk exposures and potential under-reporting of infections due to non-compulsory testing for contacts.
    • Biases in testing or reporting could have affected the results.
  10. Implications and Recommendations:
    • The study underscores the potential of contact tracing apps in developing quantitative epidemiological understanding and improving interventions.
    • It suggests the need for more systematic collection of contextual data and more precise risk assessments.

In conclusion, this research provides valuable insights into the effectiveness of digital contact tracing in assessing COVID-19 transmission risk, emphasizing the importance of exposure duration and the potential of digital tools in managing future epidemics.

Read More:

Leave a comment