People are inherently social – in all walks and aspects of our lives. This includes patients and patient advocates and their participation in healthcare. Individuals, whether they meet at a conference or have built bonds based on their shared experiences with their disease and treatments, are connected like Legos.
There are many social networks where patients interact online. As we examine below, data scientists and researchers are using publicly available posts on Twitter in new ways.
Sick? Tweet about it!
For those who aren’t feeling well, they often turn to their mobile phone or computer and tweet about being under the weather. Researchers at Brigham Young University are paying attention and have parsed tweets and their location data to help entities tracking disease find out where flu symptoms are popping up and where the disease might be headed next. In addition, folks at MappyHealth have built out the same concept to include the trending of a variety of diseases in different regions of the world!
By utilizing Twitter data and location, those involved in disease management can monitor, in real-time, the current state of an outbreak and thereby develop intervention steps to better manage an epidemic – influenza in this case.
Twitter as a discovery tool
Twitter might be the perfect medium for finding conversations amongst patients. Here’s why:
- The data is publicly available. Most tweets are visible publicly and available immediately. Unlike Facebook where privacy is set differently, Twitter offers a view into real-time communication unlike other social channels.
- The data is tied to location. In some cases, individual tweets are geo-encoded when sent from a mobile phone. In most Twitter users’ profiles, their city and state provide approximate location (this was true in 77% of BYU’s data sample above).
- The data is ready to harvest. Twitter provides an Application Programmer’s Interface, making it possible for clinical researchers and those managing clinical trials to follow tweets en masse.
What we can learn by listening
For those involved in executing and conducting clinical trials, we can learn a lot by listening in social channels, especially Twitter. Patients are already conversing on Twitter – researchers just need to get smarter about how they are listening.
Perhaps the most obvious way to become engaged with patient communities would be to focus on the existing Twitter chats that take place every day. There is much to be learned with Twitter chats, where patients are already discussing their diseases, symptoms and treatment options. These Twitter chats are recurring, making them very easy to tune into and draw connections from. For example, #HCSM takes place every Sunday at 9:00 Eastern.
Patient reported outcomes
Patient reported outcomes are reports from patients on their state of health during a clinical trial. More precisely:
Patient reported outcomes is an umbrella term that covers a whole range of potential types of measurement but is used specifically to refer to self-reports by the patient. (Source: Wikipedia)
Potential Uses in a Clinical Trial
The gold standard for conducting a clinical trial on a new drug is by randomizing and blinding a patient to their regimen, where for the sake of the scientific integrity of the study, patients do not know which control group they are a part of. In addition, patients aren’t permitted to share effects with other patients.
However, observational studies (in this case – listening to patients on Twitter) can also play an important role because unlike randomized studies, they provide information on “real world” use and perceptions of the patients’ regimen; investigators in randomized trials are not always able to see what goes on in the participants’ day-to-day life.
We must develop ways to listen to and connect with patients who are already talking, but also we must consider how to preserve the validity of the clinical trial at hand. Tremendous knowledge and signaling could be gathered if one could mash data from a randomized blinded trial to one that is more observational in nature, without tarnishing biased results.