Big Data: what does it mean for patients and health policy?

Research in Big Data: drug safety, public health genomics, electronic disease surveillance, demography, health and reproduction databases

In and around global health there is a lot of talk about «big data» — very large sets of complex data that become difficult to process using database management tools. Big data is emerging in public health because the health sector is collecting large amounts of data every single day. The data comes for a variety of Settings: clinical, billing, scheduling, drug usage, syndromes and so on. Unfortunately, in the past, a lot of that data was not leveraged to make population health more efficient. Nowadays, researchers are proposing a shift to change that.

Public health is transitioning to more data-driven policies for efficiency, cost-savings, equity and improved outcomes. Providers, patients and the entire healthcare industry may realize a variety of beneficial implications from the use and analysis of big data. However, public health community is still in its infancy in dealing with big data. There are a variety of factors increasing the amount of digitized healthcare information:

  • The push for electronic health records  and the number of hospitals as well as providers who use them
  • Social security and medical insurances with a need for large amounts of information for understanding what occurs with patients
  • New technology in general, including devices, implants and applications on mobile devices
  • Pressure to become evidence-based and predictive with public health services.

Big Data for disease surveillance

When medical records were maintained mainly on paper, it could take weeks to find out that an infection was brewing somewhere in the world. Nowadays, the growing prevalence of electronic medical records has had an unexpected benefit: by combing through the data  flow from hospitals and other medical facilities, some health departments are spotting and combating outbreaks with unprecedented speed.

Two major articles on early warning infectious disease surveillance have been published:

These two studies show that the number of queries on internet search engines (Google and Yahoo) can be used to detect influenza epidemics in the USA one to two weeks earlier and 5 weeks ahead of the real time mortality observed by the Atlanta Centers for Disease Control and Prevention (CDC). A basket of 45 search queries monitored on Google (as «indications of flu») provides a means of flu Monitoring for the USA in real time.

There is also an urgent need for surveillance systems for chronic diseases such as cardiovascular diseases or cancer. The analysis of trends in chronic diseases is often based on mortality data which will not pick up early changes in disease morbidity.

Discuss these issues with conference delegates – register now for the Careum Congress 2014 in Basel

Big Data for drug safety and computed assisted pharmacovigilance

It may take years to find out that a drug is causing undesirable effects after it has been put on the market. In a recent study, reported in the Journal of the American Medical Informatics Association March 6th, 2013, scientists at Microsoft, Stanford and Columbia University were using data drawn from queries entered into Google, Microsoft and Yahoo search engines. For the first time, they detected evidence of unreported prescription drug side effects before they were found by the Food and Drug Administration’s Adverse Event Reporting System. The data-mining techniques used in this study are similar to those employed by services like Google Flu Trends (see above). They give early warning of the prevalence of influenza to the public. How did the authors do it? They used automated software tools to examine queries by six million internet users taken from Web search logs in 2010. They looked for searches relating to the antidepressant paroxetine and the cholesterol lowering drug pravastatin and were able to find evidence that the combination of these two drugs caused high blood sugar.

Such examples seem to herald a new era that will configure the need for new skills for drug safety teams, both in industry and in the public domain: applying these methods is not trivial and requires extensive knowledge of computing and statistics that were not required by standard monitoring of adverse drug events. Specialist software will now be developed to provide integrated systems in the field. It will be difficult not to use them to focus the attention of pharmacovigilance teams more closely on the risks identified as having a high probability of occurrence. However, even with the best software in the world, nobody can predict the future with certainty, certainly not so far as drug safety is concerned: this will always call for close vigilance, including what appears improbable. Computers will not replace pharmacovigilance (also known as drug safety). But they could be of considerable assistance.

Big Data for tracking nosocomial infection and reducing hospital readmissions

In hospitals, big data will change how  health professionals care for patients at an individual level by fostering more personalized bedside Support. Big data also provides predictive models for the likelihood of readmission within 30-days, Without large data sets showing trends and patterns in huge groups of patients, this type of predictive modeling would not be possible. Big data is emerging in the healthcare sector, and it will presumably magnify over time. Healthcare organizations will continue to collect massive amounts of data, so it will be permanently challenging to aggregate and analyze the data. However, these efforts are worth it as we begin to see the implications big data promises.

«Big Data is about how to shift through the data to convert it to useful information — a more usable format and with the right visualization. At the end of the day it is about making sure providers can do the right thing for the right patient at the right time. We are trying to improve healthcare for everyone.» Rooney, Kathleen (December 21, 2012). The rise of big data in hospitals: Opportunities behind the phenomenon. {Blog entry – Becker’s Hospital Review}. Retrieved, February 10, 2014. 

Author: Antoine Flahault, Institut de Santé Globale, Université de Genève, Suisse

Acknowledgements to Avner Bar-Hen (Université Paris Descartes, France) and Antoine Geissbühler (Université de Genève, Switzerland) for their contribution in writing this blog post.

3 thoughts on “Big Data: what does it mean for patients and health policy?

Schreiben Sie einen Kommentar

Ihre E-Mail-Adresse wird nicht veröffentlicht. Erforderliche Felder sind mit * markiert.