AMDIS/HIMSS: Managing changes in data, technology & patients

NEW ORLEANS—George Reynolds, MD, MMM, vice president, CMIO and CIO for Children’s Hospital & Medical Center in Omaha, Neb., started working on dashboards when the organization implemented a CPOE initiative in 2007. Reynolds spoke during the Physicians' IT Symposium at the Health Information and Management Systems Society (HIMSS) annual convention. The symposium is co-hosted by the Association of Medical Directors of Information Systems (AMDIS).

Reynolds and his team looked at every order on the inpatient side to see what kind of order sets were being used. They built custom-made dashboards and today there are 37 in use which each took from 12 to 16 weeks to develop.

Each dashboard is a partnership, he said, with a clinician/operational leader and a dashboard builder. The key to the organization’s success with analytics has been demonstration of early wins, he said, which helped build trust.

Larger data sets require a robust governance structure, he said, including data quality and data governance and transparency of metrics.

Reynolds discussed the importance of making sure alerts and reminders are useful. For example, he asked physicians if the reminders contained information they could use. If not, they would stop that alert. “We looked at how often alerts changed behavior.” Computerized physician order entry (CPOE) was rolled out over a period of time, allowing them to compare how well a pharmacist entering a handwritten order compares with cope.

CPOE has driven down medication errors. Dashboards have helped with emergency department (ED) throughput. By drilling down to specific months, days and even times of day, “we get rapid insight into what’s going on in the environment and we can make adjustments in staffing to manage.”

Reynolds said no incentives were involved—just data. Physicians “just want to practice good medicine.” Unfortunately, “the data isn’t as clean as you would hope in many cases.” For example, the data don’t show which patients have attention deficit hyperactivity disorder (ADHD) “because pediatricians don’t like that diagnosis.” To manage that, he developed a dashboard that finds which patients are on ADHD medications instead.

Reynolds’s co-speaker, Ari Lightman, MEng, MBA, distinguished service professor of digital media and marketing at the Heinz College at Carnegie Mellon University in Pittsburgh, said “the next generation of physicians will think much differently.” Along with that comes the rise of the e-patient who will have a greater level of involvement in his or her health using digital communication, analysis and awareness. These e-patients also have a high likelihood to share information online, join patient communities and make decisions based on research.

Medicine is moving from qualitative assessments to quantitative analysis, Lightman said. People have a greater capability to visualize complex datasets so by converting to a dashboard, the data are really moving to the next evolution. “I do a lot of natural language processing in my classes. We look at the context behind the content, or not just how many times people are mentioning something but what are they saying. It’s very difficult to do that.”

Generation Z consists of those people born in 1995 and later and now have different behavioral patterns than previous generations. “They value openness over privacy which has interesting implications. The tail end of Generation Z is entering the workforce and going to come into a very regimented system and try to change it. They are very digitally inclined.”

Meanwhile, the “big data” conundrum also is impacting healthcare. “We don’t have a big data problem, we have large volumes of messy data. We need to figure out how to filter it to get at that data.”

Even as the volume of data is rapidly increasing, “all industries are dealing with messy data. Mobility and sensing will exacerbate data analysis limitations. There also is a limitation of skills and resources for mapping data analytics to business intelligence.”

Despite the changes in the industry and the population, “industry disruption is rampant,” Lightman said.