Sponsored By

Are You Using Leading Indicator Data Intelligently?Are You Using Leading Indicator Data Intelligently?

Leading indicator digital health data can help improve health care delivery, but only if it is used shrewdly at actionable times.

January 14, 2016

7 Min Read
Are You Using Leading Indicator Data Intelligently?

Leading indicator digital health data can help improve health care delivery, but only if it is used shrewdly at actionable times.

John Walmsley

In a world where ubiquitous sensors and processors make more and more health data available, we need to be intelligent in how we use it. With the right analysis it is possible to deliver actionable information to clinicians at the point in their workflow where it can have the most benefit: before an adverse event. While preparing for a panel at MD&M West, I uncovered five examples of digital health leading indicator data being applied to empower doctors, improve patient outcomes, and identify and speed business improvements and opportunities.

Improve Patient Outcomes

For hip replacements, medical centers and insurance providers have long tracked the frequency of revision procedures, length of hospital stay, number of physio treatments, use of painkillers, and more. They then use this data to identify centers and clinicians whose practice is better. Finally, they try to influence the practice of other centers and clinicians to a point where their outcomes are closer to the outcomes of the best performers.

Most centers and clinicians are naturally motivated to become better, especially in the light of the Affordable Care Act provisions for Accountable Care Organizations. But it’s tough to identify what could be better when the indicators lag the procedure. And it certainly is hard to benefit a particular patient when the feedback comes long after the procedure is completed. That's why finding a leading indicator is important. 

Bob Thornberry, MD of Tallahassee, FL, wanted to come up with a leading indicator. He wanted the indicator to make useful, actionable predictions for orthopedic surgeons while they were still able to make changes to benefit the patient in front of them.

[Hear Walmsley discuss value-based strategies and data at the world's largest medical design and manufacturing event—register for the MD&M West Conference, February 9-11, 2016.]

A critical factor affecting the outcome of a total hip replacement is the orientation of the implant as it is placed. Dr. Thornberry identified a method to measure the orientation of total hip replacements as they are being made. His method has significant advantages in set-up time over existing imaging solutions. As well as allowing the surgeon to place the implant more accurately, the data measured during the procedure can be fed forward to hospital informatics systems and used as a pay-for-performance quality indicator that the surgeon has personal control over. It also works as a leading indicator for the other quality indicators that surgeons and medical centers can be measured against. Good results on orientation will result in less required follow up care, while bad results will require more follow up care.

Remote monitoring is another area where leading indicator data can allow early diagnosis and intervention leading to improved outcomes. A recently published manuscript by Varma et al. in the Journal of the American College of Cardiology showed that "remote monitoring (RM) technology embedded within cardiac rhythm devices permits continuous monitoring, which may result in improved patient outcomes." The authors wrote, "This study used 'big data' to assess whether RM is associated with improved survival and whether this is influenced by the type of device and degree of use." Results from more than 260,000 patients implanted with either pacemakers or defibrillators from St. Jude Medical demonstrated that patients with high adherence to remote monitoring had more than twice (2.1x) the probability of survival than that of patients without remote monitoring.

A new generation of implantable cardiac monitor is also helping to manage heart failure patients. The implantable cardiac monitor, CardioMEMSTM from St. Jude Medical, measures pulmonary artery pressure for more effective management of heart failure leading to fewer hospitalizations. New long-term data published by Abraham et al. in the Lancet demonstrates superiority of the CardioMEMSTM heart failure system over standard of care. Prospective data from the CHAMPION study showed that after 31 months of follow-up, heart failure patients managed with the CardioMEMSTMHF system had a 48 percent reduction in heart failure hospitalizations compared to patients managed with standard of care. This is a clear case of improved patient outcomes from correctly applying leading indicator data.

Empower doctors

Orthopedics has historically been an area where data availability has been limited, but heart monitoring ECG data has long been reviewed and collected. The collection can create a lot of data. Doctors are understandably nervous about accepting large volumes of data into their patients’ clinical record with no plausible solution to alert them to potentially indicative events. Converting it into useful information is a challenge tackled by many solution providers. One example is CardioComm, which was selected to take over monitoring for GE’s ambulatory ECG solution, MARS Event.

CardioComm’s Global ECG Management System also supports their own HeartCheck pen which can capture wireless data directly from the patient in real time—avoiding the challenges associated with traditional phone monitoring. Travis Van Slooten of www.livingwithatrialfibrillation.com uses the HeartCheck Pen to help identify his triggers. For instance, he can record an ECG after meals to see if something in his diet has an effect. 

CardioComm applies agreed event filters to this data before adding it to the integrated record. It is important to the physicians involved that only those events agreed to represent actionable information for the physician are reported. The end result is that event reports are delivered at a pace that can be reviewed by the physician in a reasonable time and be used to plan patient care. 

This is an iteration on an existing leading indicator for future heart issues and one that leverages the new availability of lots of data with the algorithmic ability to identify agreed diagnostic or indicative patterns and straightforward communication into the patient’s electronic medical record.

Identify business improvements and opportunities

Compare the immediacy of device data with that of medical claims. A recent article in MD+DI identifies that it can take two years for Medicare claim data to be made available to health economists and reimbursement specialists. This data is essential in making the case for adoption of a new device. Providers often want to see the effect of an intervention on outcomes before moving to adopt it. The best means to demonstrate this is through the evidence of anonymized statistics of actual treatment cycles with and without the proposed new intervention.

It is common to collect data longitudinally with initial clinical trials in order to provide early evidence. Much better would be a leading indicator on outcome that could be used to convince providers that an intervention will be a cost saver and result in better outcomes. No such indicators are in common use at this point.

Whether converting data from devices into forward-looking actionable information or looking for an indicator of efficacy, care must be taken around the unintended consequences. An example of this comes to us from the world of retail. In the example of Target, we can see data being collected and converted to actionable information. We can see the action being taken—and its unexpected consequences.

 “Congratulations on your pregnancy!” Imagine the feelings of warmth that the mother-to-be must feel when she receives such a message. No doubt the marketing department at Target were imagining such warm feelings when they put a particular advertising campaign in place back in 2005. Using the data accumulated by their cash registers, Target had noticed that a pattern of purchases (unscented lotion and soap, calcium supplements, and other items) correlated strongly with a subsequent confirmation of pregnancy. Target used this knowledge to try to make connections with customers who were entering a new phase of their life that might bring new spending habits.

Famously, Target had not considered the reaction of an out-of-the-loop father to a teenage mother-to-be. With some adjustment to their technique, however, Target did grow their business massively by using the ability to connect with newly pregnant mothers before anyone else.

We aren’t looking to predict who is ready to changing shopping habits. We are, however, collecting more and more digital health data and looking to apply techniques to predict how to improve patient health. When it comes to making an improvement, the feedback loop can be shortened with an accurate leading indicator.

Check out the future of medical technology at the world's largest medical design and manufacturing event—register for the MD&M West Conference, February 9-11, 2016.


John Walmsley is vice president of product development at StarFish Medical. He will be moderating a panel on the topic of "Leveraging Value-Based Strategies and Maximizing Data Credibility" at the MD&M West Conference in Anaheim, CA, February 9-11. 


Sign up for the QMED & MD+DI Daily newsletter.

You May Also Like