Interview: Nicholas Marko, Geisinger on Building the Analytics Culture for Healthcare

We discuss how to establish credibility of data analytics, recommendations for a data-driven culture, analytics challenges in healthcare and more.

nicholas-markoNicholas Marko, MD is Chief Data Officer for Geisinger Health System in Danville, PA. He is a data scientist with expertise in predictive analytics, organizational data strategy, data-driven collaboration, and data science consulting. He heads the Department of Data Science & Engineering in Geisinger’s Division of Applied Research and Clinical Informatics (DARCI) and co-directs the High Performance Computing Center in the organization’s Institute for Advanced Applications (IAA). His specific academic interests include integration of heterogeneous data sources and application of advanced mathematical methods and modeling strategies to generate measurable value for organizations in the healthcare sector and beyond.

Dr. Marko is also a practicing neurosurgeon and serves as Geisinger Medical Center’s Director of Neurosurgical Oncology. His clinical practice focuses on surgical management of patients with malignant brain and spine tumors.

First part of interview

Here is second part of my interview with him:

Anmol Rajpurohit: Q4. What measures are required to establish the credibility and usability of data analytics for those who still continue to strongly rely on intuition (over data analysis)?

Nicholas Marko: This comes down to consistent messaging and to solid change management. Of course there are some decision-makers who are used to relying on intuition and experience, and those practices aren’t going to change just because data is available. Instead, I think it is a matter of making sure those people get a consistent message that the data is available to help them augment their decisions. The process also involves learning how those people like to consume and to interact with data. Intuitive-driven-vs-analysis-driven
Even the strongest advocates of “intuitive decision-making” generally recognize that having information is a good thing, it’s just that it has often been cumbersome for them to get this data or to understand what it means.

Through consistent messaging, easy availability, and tailored analytics we work to ensure that data is a big part of every key decision.

AR: Q5. What are your recommendations for creating an organizational culture that encourages data-driven decisions?

NM: Much the same as my answer to #4. Make sure that the data is available and is easy to interpret. Make sure that it is a part of every key discussion, and work to get a data-driven-culturedata advocate at the table whenever important decisions are being made. It is also important to foster a culture that is constantly asking “Why are we doing that?” or “How do we know that this is the best approach?” Data is there to answer questions, but if the corporate culture values blind assent over constant questioning and reassessment then there will never even be a space in which the use of data can grow.

AR: Q6. Analytics in Healthcare industry is still lagging far behind as compared to other industries such as Retail and Technology. What are the primary reasons behind this?

NM: There are a variety of reasons. One is that data and information is not the primary business of healthcare – taking care of sick people is our primary focus. Certainly information is an important part of how we do it, but when push comes to shove decisions about patients always get health-analyticsprioritized over decisions about data. And that is a good thing, because it allows us to excel at what we are supposed to be doing. However, it means that sometimes healthcare organizations don’t have as much time or resources as they would like to be focused on the data aspects of what we do. This is particularly true in an environment of rising costs and declining reimbursements, because almost anything that doesn’t directly touch a patient can easily find itself on the chopping block. We are addressing this by trying to focus on those parts of data and analytics that really help us meet our primary goal – patient care – in a more effective and efficient manner.

Another reason is the medical culture in general.

Until relatively recently doctors have been trained with an emphasis on clinical experience. The most senior people in the system have the most experience, and the decisions of these people who have “seen it all” were rarely questioned in the medical culture. That is antithetical to a data-driven culture, where constant questioning is the basis of the mindset. Only relatively recently have medical schools started to emphasize the importance of objective data and decision-making processes that rely heavily on evidence (even when they are counter-intuitive).

There are a lot more reasons and we could talk for hours about this, but overall I’d say it’s getting better. A newer generation of clinicians and business leaders are emerging, and these people have been raised in a data- and information-centric culture. That is bound to spill over into the workplace – even into an academic establishment with a tradition as long and proud as medicine!

AR: Q7. What have been the key lessons from your work on outpatient clinic prediction engine (an on-going project)?

NM: This tool is just one example of the many predictive modeling exercises that we use to try to improve the ways we take care of patients and keep the system running efficiently. In this particular case I think we see how important information can be that is not traditionally stored prediction-enginein an electronic medical record. Trying to predict what clinical volumes will be at any given time, who will show up and who will not, and what systems can be implemented to give patients better experiences with healthcare is largely dependent on how our patients behave under certain circumstances and how they perceive and interact with the world around them. There is a lot more to that than what is recorded in a medical chart, and so we’ve really tried to work on incorporating “ambient” data sources that are not traditionally part of healthcare analytics. In the end it is about knowing our patients as well as we can so that we can make their interactions with the system efficient and positive.

Third and last part of the interview