“Big data in healthcare”;
Big data is becoming a trend in healthcare organization. A number of uses cases in healthcare are well equipped for big data solution. Some education- and research- healthcare institutions and government agencies are experimenting with big data or using it in for their advanced research projects. Today most of the healthcare organization are equipped with analytics and reporting and has no need jet to implement big data solutions. But the amount of data generating in Healthcare organization today has brought the need to process all the volume of the accumulating data in more effective and efficient way for future use.
In order to know where you want to be in the future with big data, you must know where you stand now and also you must understand the importance of privacy when handling patient’s data and the common non spoken rules (ethics) in the medical world. In the maturity level of big data – healthcare organization is at the moment very low. Healthcare organization at the moment stand in the level 1 of the maturity table. Healthcare organizations has a few analytics tools at their disposal. Healthcare organizations can not afford anymore to wait for this technology mature in order to diving into analytics. So today, the focus will be choosing a data warehousing solution that can be adapt later in the future to big data. Therefore Healthcare Catalyst recommend health organization to choose for a Late-Binding enterprise data warehouse (EDW) architecture. Late binding EDW is the ideal data warehousing tool making the transition from relational databases to unstructured big data easy. After all big data is the only hope for managing the volume, velocity, and variety of this data.
1. Introduction
For the course Business studio in big data I’m will conducted a research on the topic big data in Healthcare organization. The concept of big data is an evolving term that describes any voluminous amount of structured, semi-structured and unstructured data that has the potential to be mined for information and analyzed. As every other report this report has a structure, in which the research subject is being handled. The structure used for this report is:
– Chapter 1 Introduction
– Chapter 2 Healthcare organization
o Chapter 2.1 Technology
o Chapter 2.2 Organization
o Chapter 2.3 Ethics and privacy
o Chapter 2.4 Level of maturity
o Chapter 2.5 Substantiate
o Chapter 2.6 Business case
– Chapter 3 References
2. Healthcare organization
To understand big data in healthcare, first we need to know and understand what is healthcare. Healthcare can be describe as the diagnoses and treatment one get when their sick, to the professions of one whom practice medicine. But the accurate description is the maintenance or improvement of health via the diagnosis, treatment, and prevention of disease, illness, injury, and other physical and mental impairment in human beings. These professions are being practices in healthcare – chiropractic, physicians, physician associates, dentistry, midwifery, nursing, medicine, optometry, pharmacy, psychology, etc. Healthcare organization like hospitals also provides primary care, secondary care, and tertiary care to the public. Healthcare organizations are established in order to meet the health needs of the public. In some countries and jurisdictions, health care planning is distributed among market participants, whereas in others the planning and executing of the planning is done centrally among governments or other coordinating entities.
2.1 Technology
A brief history of big data in healthcare.
In 2001, Doug Laney, devise the term “the 3 V’s” defining the big data volume, velocity and variety. In healthcare there is a lot of data “volume” coming in (What is big data, 2016). This amount of data’s are being stored in large databases – Electronic medical records. Electronic medical records “EMR” is a digital paper with all the medical history of a patient. EMR collects a lot of data. Most of the data which are being collected are meant for recreational purposes according to Brent James of intermountain Healthcare.
Health Catalyst solutions is a group of healthcare veterans with a decade of knowledge and experience in implementing data warehouse solutions and quality improvement in Healthcare. Regarding the use of big data in healthcare. Health Catalyst argued that the amount of data and the speed in which the data is being created in healthcare, is not high enough to require big data solutions jet. Health Catalyst also argued that healthcare systems show that only a small fraction of the tables in a EMR database are relevant for practices of medicine and its analytics (Adamson, Big Data in Healthcare Made Simple: Where It Stands Today and Where It’s Going, 2016) . So we can conclude for now that a majority of the data collected in healthcare today could be considered recreational. The data can be useful later on in the future but now there aren’t a lot of use cases for much of that data today.
2.2 Organization
Absence of big data in healthcare
Healthcare organizations are equipped with basic analytics tools for analyzing data. It can be argued that most healthcare organization can do enough today without big data, this includes meeting most of their analytics and reporting needs. They haven’t come close to what healthcare analytics can achieve with traditional relational databases. By focusing first on using the current tools effectively and efficiently. When this needs are met, then healthcare organizations can focus on what for big data solutions (wearable medical devices and sensors) is appropriated for the needs.
Healthcare organization are limited when it comes to use of big data. Because getting valuable and useful information out of unstructured data required special skillset. Data scientist has the knowledge on manipulating the data to get the desired results by using complex algorithms and complex queries. Nowadays these highly qualified scientist are too expensive. And putting the IT personnel’s in training is also expensive. Most of these IT personnel’s has only the basic knowledge of traditional relational database and sql queries. The transition can be too much for them. That’s why healthcare organization has to hire data scientist.
2.3 Ethics and privacy
Beside a patient’s health, there is nothing more important than the privacy and security of patient data. Big data has a few disadvantage one being security. According to Healthcare Catalyst, there aren’t many good integrated ways to manage security in big data now. But security surrounding big data in healthcare is coming along. If hospital only has to grant access to a few data scientists, then there is no need to be worried or alarmed. But when opening up access to a large, diverse group of users, security cannot be a taken lightly. (Adamson, Barriers Exist for Using Big Data in Healthcare Today, 2016).
As mention before, when needs are met, then healthcare organizations can focus on what for big data solutions (e. g wearable medical devices and sensors, cloud) is appropriated for the needs. In the medical field, the demand for these type of devices can be argued by several connecting factors. Overcrowded hospitals and longer hospital stays leads to health cost rising, increase of chronic diseases such as diabetes and other like high cholesterol which cannot be treated in the hospitals and the aging population has created the need for health monitoring, which these devices can handled.
Using technology can change the way physicians practice medicine. But this can also raise a few ethical concerns. For example Autonomy, in medical world autonomy is viewed as preventing one to self-diagnosed them self, because they don’t have the knowledge of medicine. But if the new technology provide the patient with this up-to-date information, patients will have the ability to make informed decisions when choosing the appropriate management for their condition.
However according to a paper by Bauer published in Cambridge Quarterly of Healthcare Ethics in 2007, these technologies must likely create a gap between the patients and the doctor. The bond of trust between a patient and his/her doctor is very important. By taking the bond away one can risk of do-it-yourself type of approach to medicine. While the visit’s to the doctor’s office and hospitals decreases so is the level of confidence in doctors and hospital
Another aspect is, how confidentiality can be affected by these new technologies. In Ireland under the Data Protection Acts, healthcare professionals are required to keep personal data, including the patient’s medical records secure. Since big data isn’t complete security there cannot be guaranteed that the information transmitted digitally is secure. That is why patients must be informed of this potential lack of privacy. Data stored in cloud servers or on premises can be targeted or vulnerable for hacking. They can be withheld or modified for instance delivering the wrong the drugs. Also can be used for identity theft or data corruptions. From legal point of view, according to the Data Protection Acts, the physician has the legal responsibility to keep the patient’s record confidential, but now that a third party is involved (the server) how can we prevent the data getting in the wrong hands. Should the law be changed to now involve this third party?
2.4 Level of maturity
According to the Advisor Board Company in United States, health organization are not far along in Business Intelligence. In the maturity level of big data – healthcare organization is at the moment very low. Healthcare organization at the moment stand in the level 1 of the maturity table. Healthcare organizations has a few analytics tools at their disposal, but have not yet created an enterprise wide culture of making data-driven decisions that drive level 2 organization. that have centralized infrastructure and use physical and logical data modeling.
Picture 1: Big data maturity model for healthcare organization
2.5 Substantiate
Healthcare organizations can begin taking some steps in incorporate big data in their organization by:
– Ensure better security for their big data solution. Big data runs on open source technology with limited security technology. In order to avoid big problems on the way, organizations should be careful about big data suppliers and avoid assuming that any big data distribution they select will have a good security.
– Simplify the tooling for users. People with less-specialized skillsets will be able to easily work with big data in the future. Health Catalyst stated that micorsoft’s polybase is an option. Polybase uses query tool that enables users to query both Hadoop Distributed Files systems and SQL relational databases by using an extended SQL syntax. These types of toolings will bring big data to a larger group of users.
– Big data will become valuable and useful to healthcare in what we now as the internet of things. The internet of things is a growing network of everyday devices connecting to each other sharing information and complete tasks while you are busy with other activities, for example work, or exercise. The internet of things can be used for tracking health information of a person. These information will be monitor by the physician. These device that can generates data bout a person’s health and sends the data into the cloud will be part of this IoT. Wearables for example. Glucose monitor – a sensor can be implanted below the skin and send real-time glucose levels measures into the cloud. The physician will have real-time information, if the glucose level is too high or too low the physician can intervene. Or the information is directed to a non-implanted transmitter that communicates the pager so it can display the glucose level and active the insulin pump.
2.6 Business Case
In previous chapters I have mention the limitations of big data in healthcare to dated and what big data can mean for healthcare organizations in future. But in the meantime what is the possibilities now for healthcare organization. Healthcare organizations has the needs for data-driven quality and cost improvement. Healthcare Catalyst stated that data warehousing solution make the transition to big data easily in the future.
The Late-Binding™ Data Warehouse is balanced between early binding in Inmon, Kimball, and I2B2 with the no-binding environment of Hadoop. The Late-Binding™ Data Warehouse highlights the following fundamental principles in relation to data modeling:
1. The key to success for data warehouses is relating data, not modeling data. Model less, not more.
2. Minimize the use of new conformed data models in the data warehouse by leveraging the data models in the source systems.
3. Apply the data models to subsets of data—in data marts.
4. Early binding to the esta meded 20 core data elements is a best practice.
Health Catalyst recommend this approach because it’s very similar to the big data approach. And is easier to transit to big data form relation database.
In late-binding EDW, data form source systems HER’s financial system, etc. are placed into source marts. In this process it is best practice to keep the data as raw and rely on the natural data model of the source systems. Late-binding methods minimalize remodeling of the data in the source marts. The data remains in its raw state until someone needs it or be able to used it. (Adamson, The Future of Healthcare Data Warehousing and the Transition to Big Data, 2016)
Benefits of Late binding EDW
1. A late binding EDW allows you to get data out of the transactional system at the most detailed, lowest level of granularity with the minimal transformation.
2. With a late binding approach, over time your cost are reduced because you’re putting effort into binding data you actually need.
3. Late binding approach brings flexibility because it is built of a relational database, you can use different type of commercial tooling to access it.
There are a few reasons why healthcare data warehouses fail.
1. There is no business plan. Success of implementing data warehouse must requires a clear understanding of financial and clinical needs and how you expect a data warehouse to address this needs and wishes.
2. Executive sponsorship and the managing board are not (entirely) involved. There must be significant senior leadership involved for the beginning till the end.
3. The users are not involved from start to finish. An organization can have a strong business case and great executive sponsorship but fail to involve the primary users.
4. By don’t implementing in phases. Choosing one area at a time to focus on and slowly improving quality and cost. Making it easier to rollback in case problem occurs.
5. Being too realistic. Starting the design of a data warehouse and don’t stray because it makes it too difficult. And majority of the time it takes longer to finish and cost more money.
6. Worrying about getting data governance perfect, can paralyzed the project. By putting too much effort in data governance is a mistake. An organization doesn’t need the have all the answers before starting.
Investing in Healthcare analytics in four stages
Health Catalyst is one of the leading dataware house companies in United states that provide dataware housing soltutiont to Healthcare organizations. Healthcare Catalyst followed a model with four stages of investment in a data warehousing. The stages include implementing data warehouse platform and analytics applications that run on the platform. Each stage delivers a measurable, solid benefits or ROI within a time period. (Burton, 2016)
Stage 1: Implementing the enterprise data warehouse platform in healthcare.
The first phase is getting the enterprise data warehouse platform in place. Aggregating data from disparate sources into single data warehouse platform.Health Catalyst typically execute this stage is in 3 to 9 months. Healthcare Catalyst argeud that the cost of the first phase is less than a quarter of what the health system will likely invest in the total solution. In other words, this phased reduces investment by 75 precent.