Business Intelligence in Healthcare, how does it work?
I believe that healthcare business intel is the answer that many hospitals should be striving to develop not only because of the efficiency that it can bring about, but it can also be used to help both the patients as well as the providers that suffer at the hands of clunky software solutions.
Sure, yes, I do agree that business intelligence is only as good as its foundation, data warehousing. If the data warehouse isn’t configured and managed well, with a purpose in mind, then business intel will suffer and not produce anything of worth. One of the most difficult questions that I came across after reading “business intelligence this, business intel that”… just what the heck is business intelligence? After spending a few….
I realized that Business Intelligence… is just too broad of a term and it contains a broad category of analytics, data warehousing and visualization tools. Sometimes it can be used for long-term and sustainable analytics or to help link the visualization layer (MVC) with a visual representation of the data. Despite the lack of a concrete definition, the greater problem is that many healthcare organizations and vendors often face problems due to a lack of a BI strategy, especially in an industry with some of the most-complex data. By not having a BI strategy or even a poorly developed one, these vendors are shooting themselves in the foot because they simply aren’t making the most out of their wealth of data. As the industry is shifting from fee-for-service to fee-for-value it is important that the industry is adopting these options that can allow them to stay ahead of the learning curve and to ensure that they are meaningfully using the data to their fullest advantage. Business Intelligence is meant to help solve or shine light on “what ifs” or “how so” questions, through the use of complex analysis from the large data set. Business intel is supposed to lead to understanding, insight and provide reasons to take action.
In a 2012 survey to healthcare executive’s, it was found that some of the top concerns were financial performance, patient safety, and outcomes of healthcare reform implementation. Seeing this opportunity, many management teams adopted data warehousing techniques to try and gather reliable and comprehensive information spanning the enterprise. These management teams eventually work with certain departments to try and compile data for their reports, by using an enterprise data warehouse, these teams and the executives can easily compile and generate reports together to see the big picture and draw valid conclusions about financial performance, safety and satisfaction, and outcome and quality. Data warehousing led to a more efficient and scalable reporting process, but even greater is ensuring consistent and clean data that everyone can rely and trust. In a study conducted by Scott Evans et. al. Clinical Use of an Enterprise Data Warehouse, it was specifically mentioned that an EDW is “not only an essential tool for management and strategic decision making, but also for patient specific clinical decision support”, indicating that the investment isn’t only for the administrative goal in streamlining the processes down and making the data information more scalable and usable, but it can also help direct patient standards and clinical decisions that can lead to improved health outcomes.
As more and more data warehouses were being developed, there was also an increase in the related business intelligence tools to help support decision making, both for management and clinical decision making. Although it can provide great benefits to the hospital, it also allows for great mechanisms for information sharing regarding antimicrobial resistance, measuring antibiotic use, infectious diseases, costs etc. In doing so, hospitals can better estimate time and money saved. Another thing to take note is that despite being part of the same medical system, it is advised that each hospital or institution should develop their own data warehouse due to the effects of putting “all of thine eggs into thine satchel”. Instead of a combined central, many have suggested to use a federal model where each “facility can manage and maintain control over their local databases” but also allow these databases to be queried through a standard web API.
A great benefit is that a variety of teams and departments can then use the EDW to identify opportunities for improvement. The organization then can develop and deploy targeted interventions to promote improvements in care, whether it’s reducing septicemia (which is a common problem I saw in the ED) or just to try and eliminate the needless labs and tests. With business intel, there are a variety of data warehouse designs that they are built on, the traditional approach is described as “early-binding”.
Understanding Governance in BI!
No… not that BI (look up iKON’s BI) and you’ll understand more about the GIF choice. I SCOURED for this specific gif because I realized without BI, you might as well as be throwing out the money, time and resources spent on developing a data warehouse. BI is what analyzes it and puts it all together into a format that anyone could understand. However, there are specifics in how this information and data should be used, so it’s really down to the implementation of the data warehouse. So let’s bring about the buzzzzzword. Governance. It’s important to effectively use the resources and the collected data, but the keyword is effectively, and that can only be done with proper planning, research, and management. The best way to go about this is to clean the data and remove any irrelevant data.
First of all, you gotta make sure the data is protected and secure. Certain data, especially in healthcare, is subject to VERY strict regulations and standards. It’s important that these standards are upheld and are securely transmitted when extracted and moved from the databases into data warehouse and into any data marts as well. In cybersecurity, a key concept of security is the CIA… no, not the agency but Confidentiality, Integrity, and Availability. Since then, there have been more of these key concepts but it’s important to keep the data confidential and protected, with integrity to ensure that no one has modified it, and available when needed.
Another thing to keep in mind is to ensure the relevance. Meaningful Use is a very key concept because it ensures that all of the collected data is meaningfully used (woopdeedoo, that was difficult to understand, but it’s KEY because it also restricts what kind of data can be collected if it can’t be used within standards and regulations). This kind of governance is not an IT issue. Well it is, but it is intersectoral, like more technology decisions it is key that there is close communication with the upper level management at C-Men to create policies and standards to define the data usage, strategy, and utilized tools. As I mentioned in an earlier post, policies and standards are only as effective as they are well communicated; if people don’t know, it’s your fault not theirs. Well, that’s about it in business intel. I realize that the last section isn’t the most relevant to BI but governance is an important concept throughout data warehousing and business intel, it’s vital that the data is used and managed properly so that the business intel tools have the most capability with the least liability.