This book includes a plain text version that is designed for high accessibility. To use this version please follow this link.
Predictive Modeling


Using big data for big impact


Leveraging data and analytics provides the foundation for rethinking how to impact patient behavior. By Bill Fox


T


he statistics and the costs are staggering. Roughly 27 percent of the American population has two or more chronic conditions, such as diabetes or heart disease, and 15 million more join their ranks


every year. Almost 75 percent of the total U.S. healthcare spend – nearly $1.9 trillion in 2009 – is spent on treating that 27 percent of the population. Our aging population means this number will only increase since almost two-thirds of Americans over age 65 have multiple chronic diseases, and 10,000 more of us will turn 65 every day between January 2011 and January 2030. Our industry will soon eat up a quarter of the country’s GDP.


Bill Fox, JD, MA, is senior director, Commercial Health Care, LexisNexis. For more information on LexisNexis solutions: www.rsleads.com/111ht-209


Other major industries – banking, fi nance, retail – rely on predictive analytics to make important decisions across every aspect of their operations. What will the customer buy? How will the


market react? Healthcare, however, is only now beginning to tap the potential of predictive modeling. Most of that ef- fort has been directed toward clinical data, but the statistics above clearly indicate that something is missing. What’s missing is the ability to understand patients as opposed to diseases. No matter how much comparative effectiveness data we have, no matter how much we can predict the course of the diseases a patient has, until we can understand how to have a serious and lasting impact on that patient’s behavior both wellness and disease management (WM/DM) programs will continue to have unsatisfactory results and costs will continue to rise at a runaway pace. The majority of health insurers and disease management organizations (DMOs) rely solely on data related to the pa- tient’s medical condition to target and manage participation in WM/DM programs. This approach overlooks crucial social and behavioral data that directly impacts whether or not a patient will choose to participate in such a program, what their level of engagement will be, and if that program is the right one for them. We can attack this problem by combining public data related to an individual with data related to his medical condition to better understand how to target and


16 November 2011


retain real patients – people, not medical conditions – in these programs. We can leverage this data to build predic- tive models that can more accurately assist doctors and case managers in actually impacting the behavior of patients with chronic diseases. We can start exploring analytics to achieve what we might call “intelligent case management,” which today includes targeting, recruiting, retention and compliance. By leveraging the ability to utilize “big data” effectively to drive smarter predictive models, we can begin to study the effectiveness of various combinations of clinical, claims and public data to create better, more impactful, predictive models. Leveraging data and analytics that can assist in predict- ing the ability of a given program to impact an individual patient’s outcomes provides the foundation for rethinking how to impact patient behavior. A meaningful “impactabil- ity score” would:


• Identify individuals most likely to benefi t from a par- ticular type of DM/WM program;


• Identify individuals most likely to participate actively in a particular disease management program and what level of outreach will be required to ensure that partici- pation;


• Identify data points that impact an individual’s adher- ence and compliance; and • Identify the specifi c types of outreach and support most likely to impact an individual’s behavior and outcomes. The last bullet is the most important one to the success of the DM/WM program. Impacting behavior change is the holy grail of improving outcomes for those with chronic diseases. It is also a goal that has proven to be highly elusive. By designing analytics aimed at understanding the individual patient, and not just the disease(s) he has, we can begin to better understand how to motivate and support the behav- ioral changes we know need to happen. We know where the runaway costs in our healthcare sys- tem are concentrated. Development of better “impactability” models, that leverage everything we know about the patient, must be a priority if we are to “bend the cost curve.” This is not just a healthcare imperative, but a national economic imperative that must be addressed immediately.


HMT HEALTH MANAGEMENT TECHNOLOGY www.healthmgttech.com


Page 1  |  Page 2  |  Page 3  |  Page 4  |  Page 5  |  Page 6  |  Page 7  |  Page 8  |  Page 9  |  Page 10  |  Page 11  |  Page 12  |  Page 13  |  Page 14  |  Page 15  |  Page 16  |  Page 17  |  Page 18  |  Page 19  |  Page 20  |  Page 21  |  Page 22  |  Page 23  |  Page 24  |  Page 25  |  Page 26  |  Page 27  |  Page 28  |  Page 29  |  Page 30  |  Page 31  |  Page 32  |  Page 33  |  Page 34  |  Page 35  |  Page 36