Social media listening to understand patient experiences of chronic disease

Industry Insights

23/11/2021

 

Dr Sarah Clifford   Dr Vanessa Cooper    Fernanda Trevisan  

 

 

Today we are delighted to present Sprout’s study on the use of social media listening (SML) for understanding patient experiences of chronic disease at Virtual ISPOR EU, the leading European conference for Health Economics and Outcomes Research. The presentation will be part of the session “Is Social Media Information Useful to Understand Patient Experiences and the Burden of Disease?”

The scoping review, led by Scientist Fernanda Trevisan, aimed to answer three key research questions:

 

  • What social media listening methods have been used to understand patients’ experiences with chronic diseases and treatments?

 

  • What have social media listening studies revealed about patients’ experiences of living with chronic diseases and treatments?

 

  • What ethical issues and limitations have been identified in social media listening studies?

 

We conducted a search of electronic databases and selected studies reporting on primary research using social listening to understand the experiences of illness among people living with a chronic physical condition. We extracted data to MS Excel and conduced descriptive and thematic analyses.

 

 

Overall findings

 

45 studies reporting on various chronic condition were included. The most frequent type of illness reported in the studies was cancer (37%). Most studies did not report any socio-economic or demographic data, either because data were unavailable, or individuals had not consented to share data. In those that did share data, the median age of users was 40 years (range 26-57 years) and the majority were female (75%).   

 

 

What social media listening methods have been used to understand patients’ experiences with chronic diseases and treatments?

 

By far the most studied platforms were Twitter and Facebook, but disease-specific and patient-oriented platforms, such as PatientsLikeMe, were also reported. Data extracted from these platforms included text, images, and videos. Various methods were used to extract data. These included commercially available social media aggregator tools, platform API/search functions, tailored algorithms, and manual extraction.

 

Data were analysed using thematic analysis, data mining (an automated method to quickly extract relevant content from large amounts of data), sentiment, semantic and lexical analyses.

 

 

What have social media listening studies revealed about patients’ experiences of living with chronic diseases and treatments?

 

The most frequent topics reported in the studies were experiences of disease and treatment, unmet needs and decision-making processes. The authors used thematic analysis to organise data into six themes: Impact of illness and symptoms, Treatment beliefs and experiences, Unmet needs, Experiences of healthcare, Milestones, Social support. The content of each theme is discussed in our presentation.

 

 

What ethical issues and limitations have been identified in social media listening studies?

 

Ethical and methodological issues were raised within the studies. For example, only 22% obtained informed consent from social media users – many of these studies analysed blog narratives over time. Most studies used publicly available data. Limitations reported within the studies included selection bias, content bias and questions about the accuracy and reliability of data. Other issues include the absence of confirmed diagnosis and demographic information.

 

 

Conclusions

 

We conclude that SML can generate rich insights into people’s experiences of chronic conditions and can be a key pillar of patient insights work, providing the potential to tap into a broader range of voices and data sources than other methods such as patient interviews. However, there are limitations and clear guidelines for SML research may help guide best practice.

 

View our presentation with a freely accessible download here: ISPOR presentation