To assess potential privacy concerns through real-world examples, the contact trace data of 970 confirmed patients was collected. The data listing confirmed cases date-wise from January 20 to April 20 were released by seven major metropolitan cities in South Korea.
The contact trace data was collected from various publicly accessible online websites, such as the official website and social media sites of the local government, and its press releases and briefing information. Since the data was released to the public by the government and any specific individual cannot be identified with it, there is no critical ethical concern for data analyses. As shown in Table 2, the released contact trace data included (1) the patient's demographics (i.e., nationality, gender, age, and residence), (2) infection information (i.e., infection route and confirmation date), and (3) travel log in time series (e.g., transport modes and visited places). The data is processed by the contact trace officer before it is released online (i.e., excluding places which the patient visited but no contact was made), hence the government may possess more information than the public can access.
The contact trace data of a confirmed patient (Patient No. #e06). The contact trace data of other patients from Seoul can be found in Seoul Metropolitan Government (17).
This study covered seven out of eight metropolitan cities in South Korea, namely, Seoul, Incheon, Sejong, Daejeon, Gwangju, Ulsan, and Busan. The city of Daegu was excluded from the data collection process because it did not disclose patient information since the massive contagion outbreak prevented contact tracing.
As the guidelines set by the KCDC recommend the deletion of the outdated information (after 14 days from the last contact), all the sample cases of disclosed patient data mentioned in this study were anonymized by the researchers. For instance, the address and name of a place (e.g., building name) were converted into four character long random strings (e.g., G3A5-gu, D12Z-dong, BQT3 building). Similarly, the identification number of the patient was also anonymized (e.g., #w4p).
Do you have any questions about this protocol?
Post your question to gather feedback from the community. We will also invite the authors of this article to respond.