Can Google Catch Adverse Drug Reactions?

It is inevitable for many in the health care profession to come across patients experiencing unintentional adverse drug reactions (ADRs).

Pharmacists try to ensure that they deliver the right drug to the right person under the discretion of a prescriber, and they do their best to ensure that the drug dispensed to a patient will not lead to medical misadventures that may be life-threatening.

Some ADRs cannot be predicted, however, so pharmacists can only counsel patients on what to watch out for and what to do if something goes amiss during their treatment. Nonetheless, identifying potentially common ADRs is tantamount to help pharmacists understand what to watch out for, especially when it comes to drug counseling or offering clinical insight on possible drug therapy problems.

The FDA relies on pharmaceutical companies to track all potential side effects and ADRs experienced during their clinical trials that will eventually end up in the drug monograph and many drug infobases used by clinicians. Not all problems are caught in clinical trials, however, so data must be collected in post-market evaluations and monitoring.

Pharmacovigilance is a key area of drug safety monitoring, and tracking such issues is the duty of a small cadre of individuals with ties to medical institutions and federal agencies across the world. While data on potentially new ADRs or higher incidences of ADRs and side effects can be collected through case studies and publications, this process is slow and tedious, due in part to the procedures of medical scholarship.

The FDA Adverse Event Reporting System is another system through which health care professionals can submit data on potential drug and medical device issues. Again, however, this process is left up to health care professionals to perform and may not encapsulate the multiple events patients experience, which undoubtedly go unreported.

So, how can health care professionals help collect, collate, and present data that can be interpreted in a meaningful way to identify possible drug ADRs quicker and better? This question has inspired several innovative solutions that are currently being tested and investigated by multiple parties, and the FDA is keen to see how it will all turn out.

Leveraging social media to mine for potential drug-related problems has been a hot area of interest lately. Creating algorithms that allow data scientists to track patients’ posts, tweets, and hashtags has been investigated by multiple scientists seeking to determine whether such data can be mined, interpreted, and used.

While some may question social media and its relevance, it is a brand new way to track patients’ direct input on their health, given that many individuals share everything in their lives through social media. While these posts, tweets, and blog posts may serve as cathartic means of self-expression, they inevitably serve as potential data sources that can be collected on possible ADRs that would otherwise go unreported through traditional channels, suhc as MedWatch.

There is also attention on leveraging more in-depth data provided by patient groups in the growing e-Patient movement. For instance, the FDA is examining the largely successful online patient community PatientsLikeMe as a potential source of untapped data.

Active participants on PatientsLikeMe share not only their health histories, vitals, and labs, but also their treatments, therapies, and successes or failures associated with each. In light of this, the FDA has partnered with PatientsLikeMe to determine whether this data can be collated for a potential pharmacovigilance platform.

Even Google has entered into this realm, as recent news briefs described the growing relationship between Google and the FDA to tap into the search history of patients as another raw data source. If Google can figure out how to tease out important data from patient searches related to ADRs, then the FDA may very well have a significantly large data source directly from patients to monitor and observe.

Are data collected from patients through search histories and online social interactions the future of pharmacovigilance? It could supplement current operations used by the FDA, health researchers, and big pharma. Arguably, the current limitations of these new tools is the accuracy and data presented, which may not provide a complete picture of the clinical scenario or patient characteristics that would spur a more directed research approach.

Nevertheless, Google and social media could very well be used as tools that almost serve as early warning systems, allowing medical researchers and data scientists to uncover suspected or new problems arising from drug therapy. It will be interesting to see where this will take pharmacovigilance and to what extent it will impact the health profession in the future.