By Jason Buchanan M.D., M.S., Clinical Informatics Officer, Baylor College of Medicine
Data is the new currency of the times and it will be the reason why organizations either sink or swim. The ubiquitous use of electronic health records combined with the arrival of data-rich players such as genomics, precision medicine, and the internet of things has caused a deluge of data. While maintaining high standards in carrying out their missions, healthcare organizations are treading water in attempting to collect, analyze, and translate data in impactful and actionable ways. In healthcare, as well as most other industries, data is the key driver of the decision-making. It is the compass which guides operational and financial decisions on the journeys that organizations embark upon to satisfy their stakeholders and those whom they serve. This being known, there are five general pitfalls to be mindful of before setting sail on your data analytics journey.
1. Not Determining if Your Journey is One of Exploration or One of Destination
One of the beauties, as well as one of the pitfalls of data and data analytics, is that it can lead you down many different paths. There are two general approaches to looking at data. The first is an exploratory approach in which you have a broad question, with the liberty to see where the data leads you. The second approach is more targeted, where there is a specific question and the analytics can be tailored precisely to answering that question. Each approach has its benefits, drawbacks, and optimal use scenarios.
Additionally, each approach can potentially lead you to the same conclusion, but with greatly differing expenditures of time and resources. Therefore the “captain” (leader(s) in charge of the project) needs to clearly convey the question to be answered and the type of approach expected to be used. This will be particularly important for those leaders who are new to an organization, assuming a new role, or have an unseasoned analytics team.
2. Not Having a Complete Crew
Once a question to be answered has been conveyed and the approach to analysis has been determined, the temptation is to start drafting and implementing the data analytics plan immediately. Before data analysis, consider taking the opportunity to pause and examine your team. Perhaps there is a subject matter expert within, or affiliated with, your organization who can provide great insight and a different perspective that will help focus your mission? Does your analytics team have the resources and expertise for this particular project? Are you using other departments (such as finance, billing and coding, purchasing, social services, etc.) who may have different data sources that can provide a more organizationally holistic outcome to your analysis? Interdepartmental collaboration helps avoid myopia, frequently provides a more robust outcome, and helps ensure that the data analytics project is well aligned with its organization’s mission.
3. Not Checking the Integrity of the Vessel
Fidelity and integrity of data are paramount. Much occurs behind the scenes in the collection, aggregation, curation, and display of data. Errors are always a possibility, and you do not want to set sail on your project with an unrecognized hole in the hull of your ship. Therefore, it is critical to ensure that the data being analyzed is valid (correct, accurate, and reliable) before its use in organizational decision-making. The type of data validation testing will vary widely depending upon many factors, including organizational resources, size of the data sets, project type, and types of data to be analyzed. Most importantly, the organization should have a documented and vetted data validation plan in place, which is being consistently followed.
4. Lack of Communication Among the Crew
It is well known that overcommunication typically is the mantra in successful, high functioning, and high-reliability organizations. Communication breakdowns are one of the top causes of project failure and will derail your mission. Frequent team communication, using multiple different mediums (email, text, video, etc.), helps to ensure clarity and alignment with the goal. Communication is particularly important in highly data-driven teams as you will find that some of your most technically talented team members may need a bit of extra attention when it comes to communicating their progress and needs. This centrality of the importance of communication has only been magnified during the current Covid-19 pandemic, which has caused separations of distance, time, and connectedness among the employees of many organizations.
5. Losing Track of Your Position
Data and all of its possibilities are intoxicating. It is quite easy for the waves of data, and their associated eddies and currents, to sweep your project into unintended areas. This will cause your project to steadily drift away from the initial goal. Therefore, it is important to continually focus on the question presented and let the data be your guide. Just as one does not sail without tracking one’s position, your data team should periodically revisit the initial question presented, to ensure that the data analysis is still heading towards the intended goal. Additionally, a data project may not seem fruitful at its conclusion. However, keep in mind that the analyzed data still can be used to reshape a prior question, or formulate a new question, whose analysis brings your organization one step closer to achieving its overall objective.
Communication breakdowns are one of the top causes of project failure and will derail your mission. Frequent team communication, using multiple different mediums (email, text, video, etc.), helps to ensure clarity and alignment with the goal.
These are fascinating times as we see the data and health IT landscape rapidly change before our eyes. The above pitfalls provide a set of common mistakes to avoid as your organization sets forth on its respective data-driven journey into uncharted waters teeming with rich and untapped possibilities.