I found an article online about how big data can increase the quality and profitability of healthcare. I summarized the articles main points and added a little bit of information that I learned from a meeting at M.D. Anderson.
Opportunities:
Big data provides many different applications for patients, providers, researchers, players, etc. These applications are designed to increase the quality of patient care mostly due to the new model of hospitals that focuses on keeping patients healthy instead of keeping patients in care more days. Some of the ways that big data will help patient care is by increasing the amount of data, breadth of data, and communication between the players in healthcare.
Big data will increase the capabilities of globalization of information.The benefits of globalization are that it can increase populations for clinical studies, trending and disease monitoring for epidemics; could create the capability for early detection and potential for improved results. This would lead to more innovation for experimentation better protection and monitoring for epidemics, and various other applications. One other effect of globalization will be the increased accuracy due to more scrutiny.
Another aspect of Big Data that can have huge healthcare implications is social media. An increased used of social media will increase communication between patients, providers, and communities regardless of space. This would also lead to more globalization which will also help increasing data.
Problems:
One of the main problems with gathering data from patients is that there is low accuracy, a privacy concern, and lack of consistency. Different people report data differently and care about their privacy differently. This creates an issue of collecting data that is important because you get out what you put in. If you put inaccurate and inconsistent data in, you'll get inaccurate and inconsistent results.
Another problem with getting good data is that there already is a huge amount of data about healthcare out there but not all of it is useful. One patient generates a ton of data that needs to be processed and integrated to become meaningful for analysis. Also, the people who are recording the data don't always know what data is useful for analysis.***
Also, healthcare is known for being slow to adopt new technology that will greatly affect healthcare systems and there is so much separation of different sectors in healthcare that it is hard for all of healthcare to adopt technology at once. Even within the same hospitals or healthcare centers, different departments are separated to the extent that they don't communicate, collect and store data completely differently, and just work in completely different ways.*** This creates a lot of problems that need to be solved before big data can be successfully implemented.
The article make several recommendations about the implementation of big data technologies in healthcare.
- Establish business intelligence center of excellence with a focus on big data
- Decide appropriate big data strategy based on the organizations current and target business and technological maturity and objectives
- Asses the various big data initiatives that can be deployed to meet overall corporate objectives.
- Work with partners that understands the full range of big data technology and implications.
***This is something that I heard personally from engineers working at M.D. Anderson about the difficulties of working at hospitals.
Especially for medical studies, big number of cases and time period of treatments are very important. More information about patients provides better observation on the statistical studies and so better predictions for the future medical issues. This type of studies requires big data and thanks to huge data storage possibilities, very big amount of data sets can be stored easily.
ReplyDeleteI found a video relating to healthcare by Dr Avey is the co-founder of personal genomics company 23andMe. Think of ways to improve global health and you might suggest increasing the number of vaccinations, finding new drugs or employing more doctors but Dr Avey believes the answer is data.She believes that sensors and test give us the ability to collect data about our bodies that has until now been invisible. This, coupled with advances in computer technology that allow us to store and process massive amounts of data, will have a transformative effect on health care, she says.
ReplyDeleteVideo link and Reference : http://www.bbc.com/future/story/20121212-will-big-data-cure-disease
All,
ReplyDeleteBig Data in healthcare is an area for great discussion. I would highly suggest using the search term: "big data" healthcare in one of the search engines and you will find many interesting results.
I believe Katy touched on an important problem, which is privacy (as well as security). How to ensure that different players in the healthcare industry get access to different parts of the data. For example, you probably do not want your dentist to know too much about your medical conditions (as long as it does not affect his/her job). How would you limit access to the data?
Another important concept is standardization. How would you standardize data protocols such that the data is reported in the same structure and using the same language. Thus, data quality and integrity becomes pivotal to ensure that all information is correct. How would you quickly detect errors?
These are all additional issues to think about.
Fadel
http://www.infolawgroup.com/2013/02/articles/big-data/the-privacy-legal-implications-of-big-data-a-primer/
ReplyDeleteI found an interesting article referring to big data and privacy. The section that was most interesting to me was 2.4 Anonymization of Big Data. The main idea is obvious; make the data as anonymous as possible. This seems like it would be easily doable without losing some information in some cases but not in other. Specifically for healthcare, there are some characteristics that are extremely important to healthcare professionals (age, height, weight, sex, medical history…). Now, things like age, height, and weight aren’t a huge problem on their own because there are only so many combinations and there are bound to be a lot of people with the same numbers for those. However, if there are only 200 people in the world who have a certain medical condition, those specific metrics could identify a person unwillingly.
So even if there is a way that medical professionals could get useful information without getting enough information to identify a person, I’m sure not everyone wants their information shared then. There seems like there could be a lot of issues with consent. This provides a huge problem of an incomplete picture of what is going on (put junk in, get junk out). This could even cause more harm than good in extreme cases. So at this point, it seems that only so much data can be used for healthcare in a positive way but I’m sure we’ll get there eventually. =)
Katy