What is Big Data?
In the current era of rapid technological advancement, digital data is being collected at a previously unprecedented rate. A study by Monteith and colleagues indicates that two and a half quintillion bytes of data are generated every day, and that nearly 90% of all the data available today has been created in the past two years.1 While data has been an essential source of knowledge throughout world history, the digital age has established immense possibilities for its management and analysis.
Big data refers to specific data sets that are extremely large and very complex. They differ from traditional data sets in their volume, velocity and variety, in addition to their complexity and potential influence on human decision-making processes, obtained by providing usable and accurate data in a short amount of time.1 Due to immense storage, processing, and analysis requirements, big data is nearly impossible to manage with traditional software and hardware tools. Instead, it requires innovative solutions that are usually based on pattern recognition, the process of machine learning whereby computers are able to identify complex interactions among different sets of variables, surpassing the ability of traditional statistical methods.2
Utilizing Patient Data
Big data in psychiatry relates to the immense volume of data collected from patients, and also to the speed with which data is collected. It can consist of clinical data, data from clinical decision support systems, patient records data, data generated through machinery, emergency care data, and even data collected from the internet, social media platforms, news feeds and journal articles.3
As in other fields, big data management in psychiatry presents a good opportunity for research and exploration. Clinical trials are often not as useful as those in other healthcare sectors because they are often incomplete, unavailable, or inconclusive. In these circumstances, big data could immensely aid in the management and analysis of observational data, one of the core information sources in the field. Complex analysis of this data might be able to help in establishing new clinical distinctions between cases and symptoms, as well as various phenotypes. Improved understanding of phenotypes could in turn increase comprehension of the variety present in psychiatric diagnoses, and the role of genetics in specific cases.
Furthermore, big data could improve knowledge about subpopulations that are usually marginalized in traditional samples, by utilizing integrative data analysis to ultimately establish sufficient sample sizes of these groups. Additionally, big data could be able to process substantial volumes of multisite neuroimaging data, helping expand our understanding of human brain processes in normal as well as in diseased states. Some of the more prominent big data projects in psychiatry include: a study into the relationship between the development of severe psychiatric disorders following hospital treatment for head injury; research into the relationship between the ages of fathers and mothers and the risk for autism; an investigation into substance use disorders among psychiatric patients; a study to identify suicidal risk in patients hospitalized for a psychiatric disorder; and, an examination of the prescription of psychotropic drugs by specialty.1
Finally, big data analysis is proving to be a useful tool to deepen our general understanding of human behavior. Since it can process and analyze massive quantities of information, it’s able to combine a large volume and variety of personal transactions from social media, as well as more scientific sensor and monitoring information. Consequent big data analysis can offer a very realistic up to date insight into human behavior that was previously very difficult to recognize. Combining both social media postings and clinical data can uncover behavioral patterns that are otherwise intentionally hidden and hard to detect with other available means. Knowledge gathered from such analysis could provide a path to the distinction and recognition of new biomarkers present in many of psychiatric illnesses.1
Suicide Prevention
An important role of big data gained from social media in psychiatry could be in suicide prevention. Successful intervention in this area would require massive samples, involving the complete medical history of millions of individuals, combined with their social media posts. The potential in this area is tremendous. The suicidal phenotype is characterized by extreme heterogeneity, and potentially suicidal individuals are very often excluded from any clinical trials. Big data could help by combining very complex and large data samples to detect patterns, signaling suicidal inclinations.4
In summary, big data is already being used in psychiatry for many different purposes. Velocity of data acquisition continues to be one of the key challenges, because most data is updated on a periodical basis and there are not many real-time options available. It should also be recognized that as big data gains a more prominent role in psychiatry, issues of governance and security will need to be clearly considered, and that there must be a thorough and open public dialogue on ethical issues.5 However, big data in psychiatry can offer substantial benefits for the treatment of patients and comprehension of their disorders. It would benefit psychiatry professionals to recognize the potential of such ever-evolving technological advancements, while also being careful to preserve high standard of professional demeanor, as well as the traditional doctor-patient relationship based on trust.
References
1. Monteith S, Glenn T, Geddes J, Bauer M. Big data are coming to psychiatry: a general introduction. International Journal of Bipolar Disorders. 2016;3(21):1-11.
2. Passos IC, Mwangi B, Kapczinski F. Big data analytics and machine learning: 2015 and beyond. The Lancet Psychiatry. 2016;3(1):13-15.
3. Raghupathi W, Raghupathi V. Big data analytics in healthcare: promise and potential. Health information Science and Systems. 2016;2(3):1-10.
4. Grunebaum MF. Suicidology meets “big data”. The Journal of Clinical Psychiatry. 2016;76(3):383- 384.
5. Stewart R, Davis K. “Big data” in mental health research: current status and empirical possibilities. Social Psychiatry and Psychiatric Epidemiology. 2016;52:1055- 1072.