Acceptability of commercial and health data for the purpose of cancer symptom surveillance


Session type:

Hanna Skrobanski1,James Flanagan2,Xin Shi3,Yasemin Hirst4
1University of Surrey,2Imperial College London,3Manchester Metropolitan University,4University College London



Longer patient intervals can lead to more late-stage cancer diagnoses and higher mortality rates. With advancements in machine learning, there is potential to explore the complex relationship between patient’s symptom appraisal and the first consultation at primary care through linkage of existing datasets (e.g. health, commercial, online). This focus group study aimed to explore the acceptability and public attitudes towards using commercial and health data linkage for the purpose of cancer symptom surveillance.


A total of 29 adults aged 25-74 took part  in one of the four audio-recorded focus group discussions. Participants were recruited from various socioeconomic backgrounds and regions in the UK using social media, an online survey and existing networks. Verbatim transcripts were analysed using Thematic Analysis.


Cancer symptom surveillance using commercial and health data linkage was perceived to be a futuristic and an exciting idea; however, it was considered with caution. Willingness to consent depended on the type of data, benefits, risks, beliefs about accuracy, and preferences for direct/indirect feedback of cancer risk.  Participants were more in support of the use of objective data (e.g. mobile apps) rather than subjective data (e.g. social media entries), and wanted greater transparency about the purpose of data linkages and personal control. Participants discussed that if cancer symptom surveillance is found to be effective, it could lead to a positive impact on research and e-health, increase in early diagnosis and ultimately reduction in costs to the NHS. Perceived risks were the misuse of data (e.g. increase in insurance premiums), the validity of data (e.g. shared accounts) and the potential outcomes of unwarranted GP appointments and anxiety caused by an inaccurate prediction.  


The potential use of commercial and health data linkage for cancer symptom surveillance was generally acceptable to the general public with assurances for transparency, security and confidentiality at all times.