Healthcare Informatics: Ready to Unleash a New Wave of Advanced Analytics?

healthcare informaticsTo understand the state of healthcare informatics at the present time, it’s helpful to think of one of the last scenes in the movie Raiders of the Lost Ark. Specifically the scene where Indiana Jones is being told the government has top men working on discovering the secrets of the Ark of the Covenant when in fact the Ark is being slowly wheeled into an obscure corner of a dark, gigantic warehouse amid what looks to be thousands of other such treasures. “They don’t know what they’ve got there!” Indy later exclaims.

Healthcare informatics is facing a similar situation: massive amounts of powerful data are sitting unused in storage. Health systems have done a great job of acquiring data, especially over the last few years since electronic health records (EHRs) became widespread. But for the most part, the data just sits in a dark warehouse without creating value for those who store it.

Fortunately, healthcare executives aren’t as short-sighted as the government bureaucrats Indy was facing. They do understand that their data stores are extremely valuable.

Healthcare leaders also know the next frontier, the next challenge, is to make the stored data (clinical, genomic, and other data) available for analysis, so they can learn from it and fold that knowledge back into the EHRs and other healthcare data tools to improve patient care. Not just for individuals, which is important, but also across populations. By extracting the insights of the stored data, health systems will be able to reduce outcome variations and raise the level of quality and patient satisfaction while driving down the cost of care — the goal of the Institute for Healthcare Improvement’s (IHI) Triple Aim.

The Birth of Healthcare Informatics

Every great story has to start somewhere. In the case of healthcare informatics, the story starts with some very bright people, many of them clinicians who moved into academia. During the past three decades, these visionaries saw a need and opportunity for technology and information to support the decision-making process in delivering improved patient care. With a lot of work and research, pioneers at informatics think tanks, such as the University of Utah, Indiana University, Columbia, and others, made significant breakthroughs in terms of using data to understand and manage populations.

Now the technology is finally catching up to the work of these healthcare informatics pioneers — and healthcare leaders now understand the need to tap into their data by using informatics technology.

Healthcare’s Current “Recreational” Data Situation

Thanks to federal and state incentives to implement electronic health records (EHRs), health systems are now very good at capturing all sorts of clinical data about patients. This captured data provides a longitudinal view of each patient over the entire course of his or her treatment. The data also helps improve care because it’s now possible for clinicians to see what is occurring in all the different care settings for their patients, review all the medications that have been prescribed, and note whether or not patients are adhering to their treatment plans.

Healthcare’s current data situation is still fairly reactive and historical, however. Brent James at Intermountain Healthcare calls it “recreational data collection,” i.e., collecting data for its own sake — the way some people collect coins or stamps. Or the way Indy collects ancient relics to display in a museum.

Moving from Data Acquisition to Data Extraction

The next frontier, and the one that’s really exciting, is moving from data acquisition to data extraction. This new phase will make it possible for analysts to extract the data out of the EHR, and combine it with data from other systems to provide proactive care to populations of patients. Data extraction will also make it possible for health systems to improve the overall quality and cost of care. But to do this, organizations need to move from the current data capturing phase into a data provisioning and mining phase. In many ways, this next phase is similar to the way Meaningful Use was staged. The first step was to get everyone using EHRs. Meaningful Use Stage 2 and 3 are about using the data to improve the health of populations of patients.

However, for health systems to achieve these goals, they will need to change the way they allocate their budgets. Instead of focusing primarily on capture and storage, an equal amount also needs to be budgeted for data extraction and using the data — a significant change in strategy.

The Next Frontier of Healthcare Informatics: Using the Data

It is important to understand that health systems need more than clinical data for decision-making because this type of data doesn’t provide enough information to answer every quality improvement or patient care question. Health systems also need to analyze financial data, quality data, pharmacy data, and other factors that can affect the quality and cost of care. By connecting all of the various types of healthcare data, it will then be possible to answer questions such as these:

  • Which hip prosthetic delivers the best outcomes, both from a cost and a quality of life standpoint?
  • How can we find and eliminate variations in care delivery in order to improve outcomes and lower costs?
  • Why did this population of patients, treated at this hospital, have better quality and financial outcomes than those patients at other hospitals?

Understanding the entire picture of patient care is best accomplished by aggregating data from the individual source systems into a Late-Binding™ electronic data warehouse (EDW). Then informatics analysts can use powerful analytics applications to make new knowledge discoveries. Extracting the data out of the EHR (and other) systems, analyzing the data, and folding the results back into the EHR will help support clinical decisions and adherence to standards across populations as well as improve the care provided to individuals — bringing the data full circle.

A good example of bringing the data full circle is the accepted standard of not performing elective inductions on pregnant mothers before 39 weeks. Most healthcare organizations understand the value of this standard in reducing risks to the mother, the newborn, and the provider, and have rules in place to support it. Yet when we analyze the EHR data for our clients, we often discover that 15 percent of the time they are not adhering to the standard as an organization.

In many cases, the reason is quite simple: the health system doesn’t have a single protocol to collect gestational age and they don’t have visibility into their rate of elective inductions. They also may have many different ways of collecting gestational age data in the EHR, which makes using the data in standard calculations very difficult. Once health systems resolve the data collection issues they can begin to work on interventions to improve care, bringing the data full circle.

Determining Which Data Is Useful

In the data acquisition phase, the objective was to capture everything. EHRs had to be built that way because health systems didn’t know which data they would need. Moving into the analysis phase, however, health systems are discovering that not all data is worth capturing. It’s like the saying about strategy: 9/10ths of good strategy is knowing what to throw out.

Of course, some data is captured by mandate so it must be collected whether it’s useful or not. But other data is arbitrary, and health systems have much more of it than they need. Since collecting and maintaining data can be very expensive, now is the time to determine which data systems should continue to collect to improve care and which data is unnecessary.

One of the best ways to determine which data is useful is by looking at which data is being used and which is not. An informatics analyst may think they need to run a certain report, but if no one is looking at the report or using the data to draw conclusions, why bother keeping the data? Because that’s the way it’s always been done is not a good enough reason. Instead, health systems need to focus their data capture on the information that helps with clinical decision-making at the point of care.

Placing Effort Where the Greatest Value Lays

Once a health systems agrees with the premise of using data proactively to improve outcomes, the next question is how should the system begin prioritize areas for improvement. No matter how dedicated a healthcare organization may be to delivering quality outcomes, there is always room for improvement in every area.

The best way to launch a quality initiative is to start where the changes will do the most good. After all, it takes just as much effort to improve clinical outcomes for small populations as it does for large ones. But the return on investment (ROI) — both in terms of improving patient health and driving better financial outcomes — is much higher with larger populations.

In selecting a clinical process for quality improvement, an organization needs to choose the process with high volumes that will result in greatest ROI. For example, a focus on improving the care around hand procedures requires nearly the same amount of effort and resources as the same focus on patients with congestive heart failure (CHF). However, the volumes will be much higher with CHF patients, thus a higher ROI.

The decision of which group to target is now easy. By focusing efforts on the CHF patients, the health system will be able to deliver value-based care to a much larger segment of the total patient population. Then, after there’s been an improvement in the large populations, the health system can circle back to focus on the smaller patient populations.

As health systems need to move towards accountable care and other value-based models, payment will be based on successful outcomes rather than encounters. Using healthcare informatics to analyze the data in an EDW instead of in an EHR and having the ability to study the data across populations will make value-based models more feasible. And affordable — for everyone.

Looking Ahead — The Rise of the Healthcare Consumer

The ability for healthcare organizations to analyze their data to improve care and decrease costs will become even more important as consumerism in healthcare continues to take hold. The day is coming when consumers will shop for providers and procedures, such as hip replacements or even open heart surgery, the way they shop for cars — with an Internet printout and consumer reviews in-hand.

Patients and consumers will have more input into their longitudinal records as well. Health systems will start to capture more data from telemetry, fitness devices, and home medical equipment that feeds directly into their EHR. Think what that means for vital signs. Rather than having a once-a-year snapshot of a patient’s blood pressure or temperature, clinicians will be able to see a daily record that indicates important trends over time. More data will provide a more accurate picture, which can be used to improve the care of populations through analysis.

To survive and thrive in this new world of healthcare informatics, providers will need to move from collecting data to analyzing it for critical insights, and from there to knowledge and finally wisdom — the classic maturation cycle of data. Even if health systems make mistakes, it’s OK. Together, healthcare organizations will learn and get better at closing the analytics loop and leveraging their valuable data to improve patient care.

Are you prepared for the next frontier in healthcare informatics? What have you done to prepare for the transition to data extraction?

Powerpoint Slides

Would you like to use or share these concepts? Download this Best Healthcare Data Warehouse Model presentation highlighting the key main points.

Click Here to Download the Slides

Loading next article...