Loading...

Medevac Cases Identifying insights Clinical applications for analytics

Home / Blog / Medevac Cases Identifying insights Clinical applications for analytics
Medevac Cases Identifying insights Clinical applications for analytics

Medevac Cases

Identifying insights Clinical applications for analytics


One of the most significant applications for analytics is using it to identify risk patterns. Predictive models, when used alongside other important health attributes, can help clinicians manage patients more proactively — for potential rehospitalization risk, for example,” says Solow. “There has been success using artificial intelligence (AI) and data from health systems to identify patients at risk for atrial fibrillation and chronic obstructive pulmonary disease who hadn’t been flagged before. There’s huge potential value in identifying patients with emerging risk.” Successful deployment requires the right combination of data, analytics, and clinical expertise.

Analytics can also reveal ways to improve processes and, by extension, patient care, says Zallen. For instance, when patients have avoidable events, like ambulatory care sensitive (ACS) admissions or ED visits, what led to these outcomes? Was it not coming in for visits with their PCP? Not having labs monitored? Not adhering to medications? Analytics can help identify the causes so practices and health systems can change care processes to prevent such events in the future,” he says. “There’s also tremendous opportunity to analyze and improve the care of patients with chronic illness. Analytics can help answer important questions: How well is their care being managed? Are they getting appropriate medications and treatment based on the most current evidence-based guidelines? Do their clinical outcomes show that their condition is in good control? The answers to these questions can guide changes to care processes for chronic condition patients.”

The challenge with trying to use analytics to make improvements is, how do you pick the areas of focus? What are the most important things to go after? Going after three things often gets better results than going after 30.

Complex chronic illnesses, with their often-elusive diagnoses and multiple symptoms, are ripe for transformation through analytics. “When you’re at your doctor’s office, somebody checks your vital signs, asks how you’re feeling, why you’re there,” explains Solow. “The quantitative information — blood pressure, temperature, and so forth — go into discrete fields in the EMR. The qualitative information, such as mood and lifestyle, gets summarized in free-form text fields. Those qualitative responses are a source of potential insights about lifestyle and disease progression, but the data can be difficult to extract and utilize.”

How can we unlock this source of insights? By making previously unmeasurable clinical notes structured and actionable. Solow states, “One analytics tool that’s proving helpful in this regard is natural language processing (NLP), a type of AI technology that allows computers to read, interpret, and organize human language. NLP is one of several approaches that can be used to gather insights from unstructured, ‘qualitative’ EMR content. This data often uncovers vital information, such as how sick a patient is or how likely they are to respond to a new treatment.”

Alignment, acumen, and action: Building a strong foundation for clinical analytics
There are many cultural and operational elements health organizations should consider ensuring they’re getting the most from analytics initiatives with clinical implications. We identified some of the most important:

1. Analytical capacity and acumenThere’s so much data out there says, Solow. One study shows that by 2020, the amount of health care information — genomic labs, imaging, consumer data, what have you — will double every 73 days.2 That’s a lot to sift through. Organizations of all sizes should make sure either they or a partner can make that information actionable. For example, Benevera Health uses an analytics platform because they realize you need to get health economists, epidemiologists, and clinicians working together to interpret and analyze all that data, so you’re empowered to make evidence-based decisions

Baker agrees. The challenge with trying to use analytics to make improvements is, how do you pick the areas of focus? What are the most important things to go after? Going after three things often gets better results than going after 30, so it’s important to know where to focus your efforts. Executives need analytic insights to help them choose because the opportunities are endless.

2. Alignment between C-suite and frontline clinicians
Simply put, you’ll get nowhere if executives and frontline staff aren’t aligned on business priorities,” Zallen says. “Analytics efforts often start with clinicians who want to push quality improvement efforts. But if they don’t receive support from executives, and time to participate in these efforts beyond their regular duties, the momentum dies. On the flip side are situations where executives push performance reporting efforts without aligning clinical schedules, providing incentives, or presenting the data in a focused, actionable way. That creates cynicism.”

It’s important for everyone in an organization to realize the time spent on quality improvement efforts or data analysis is at least as important to the health of the patient population as direct care.

The executive team should make analytics maturity not just a goal for the year or a current strategic priority, but part of the organization’s ongoing mission,” continues Zallen. “It’s about getting people believing data is meaningful and used to receiving and acting on it.” Solow concurs: If C-suite leaders aren’t true believers in data and analytics, it’s not going to reach the care providers, where it can make the most impact.”

3. Training and incentives
Incentives for clinicians who improve quality or adhere to care guidelines are also beneficial for reinforcing an organization’s commitment to using analytics to improve care.

It’s important for everyone in an organization to realize the time spent on quality improvement efforts or data analysis is at least as important to the health of the patient population as direct care. That starts with intentionally allocating time for training and sharing insights,” Zallen notes.

4. A clear, actionable analytics story
The story data presents should be concise and straightforward, with clear actions clinicians can take right away, says Zallen. “That’s how you get the most engagement.” And if the data and analytics story includes relevant benchmarks, so much the better. “Doctors don’t like to be seen as lagging peers. Presenting insights about how other organizations are performing can be particularly valuable in getting traction.”

5. Actionable data access
The goals and structure of analytics projects are only part of the puzzle. There’s also the question of who receives access to information: executives, frontline clinicians, or a mix of both? And how many reports should they receive? Zallen believes full access to analytics would confuse and frustrate most doctors. They want the kernel of information that lets them know what they need to do. Most don’t want to extract that from the data,” he explains. “Even if clinicians felt comfortable accessing this kind of information, it often means leaving their EMR, the place they live in throughout their workday. That’s understandably a major friction point.

On the other hand, data that appears in the EMR in real-time can be game-changing. Think about paper clinical charts, Solow says. A patient would walk in and I’d have this piece of paper; maybe it has a problem list on the front, maybe not. Either way, it’s a static tool. Today, charts can combine AI and the EMR data to show any care gaps that need closing on that visit in real-time — ‘Oh, they’re due for a cholesterol check, they’re due for a pneumococcal vaccine,’ and so on. That didn’t use to be the case, and it’s thanks to analytics.”

What does success look like?
One way Benevera Health uses analytics is to provide more wraparound, whole-person care for what they term a high-need population — and measure its effectiveness. “We’re asking ourselves if we’re having an impact on that population,” says Baker. “What we’re finding over successive analyses is that when we compare the population we intervene with to look-alike patients we don’t serve, good health care utilization — meaning solid preventative care, pharmacy use, good primary care — goes up by about 10.5%. At the same time, avoidable care is going down. We’re seeing an 11% lower rate of inpatient admissions, an 88% lower rate of ER use, and a 26% overall reduction in utilization.”

Zallen notes the multiplier effect when a clear incentive program is combined with actionable, clinically meaningful data. “When I was at a regional health plan, I worked on outcome-based performance incentives for hospital system contracts. One hospital we worked with had a bloodstream infection (BSI) rate for ICU patients of 10% to 12%. A bloodstream infection is dangerous, particularly if you’re very sick, so a lot of those patients died. In those days, 10% to 12% was considered unfortunate, but a normal rate. I worked with the hospital to institute an incentive in its health plan contracts to lower the BSI rate. Within a year, the infection rate was under 2%. Over the next couple of years, it went under 1% and stayed there.”

“It’s already underway”: The future of analytics in clinical settings
Optum’s most recent annual survey of U.S. health care leaders on the topic of AI and data analytics showed an 88% increase since 2018 in the number of organizations that have implemented an AI strategy.3 “That’s pretty revealing,” says Solow. “These leaders know the future is here. It’s already underway.

Here’s where we believe the next leaps forward could be:

Improving EMR data with medication adherence measures: Zallen sees the potential, saying, “While a lot of important information is built into EMRs these days, an even more valuable piece that isn’t there is information about medication adherence. Say a patient isn’t taking their blood pressure medicine on a regular basis. The doctor is usually unaware of that. If they had the information with the patient in front of them, they could say, ‘What’s the problem? What’s getting in the way?’”

These leaders know the future is here. It’s already underway.

“We have that information in the form of prescription refills Solow adds, but it needs to get to doctors to make it actionable. In my past work on the pharmacy benefits side, we sent millions of letters to doctors, which isn’t effective anymore. If you can get that information in front of them in the exam room, then it becomes valuable

Analytics in clinical trials: Solow is excited about the role analytics could play in clinical trials. The FDA will soon start accepting claims data and EHR data as a virtual medical trial to help medications get approved. You can use data to identify patients for clinical trials and follow them in a real-world setting without doing your traditional radio call for patients, who often don’t really represent the world.” Reduced expenses for clinical trials and improved data sets lower the significant time and financial investment required to bring new medications to the market, which in turn can make medication more affordable for patients.

Expanding the concept of care: Baker believes predictive analytics could be useful for illuminating social determinants of health (SDOH), like a patient’s food insecurity or difficulty with transportation. If you think about what makes a person healthy, plenty of models show it’s about 20% medical care and 80% everything else: lifestyle, housing, clean water, clean air, financial stability, education, she notes. So analytics using medical data alone are just scratching the surface. How do we bring medical and social determinant data together to have a fuller picture of a patient and get even more accurate predictions about their health risk and needs?

She also believes analytics could help health care providers expand their concept of care. We tend to think the patient experience is limited to the time spent in the doctor’s office. But what happens when the patient leaves that appointment with a new prescription, dietary guidelines, or exercise recommendations? Each item in their care plan might have its own barriers. Wouldn’t it be great to be able to know, right there in the doctor’s office, what those barriers are and what we can do now to address them proactively? We need to use data to stretch the notion of patient experience beyond the exam room. There’s so much more we could be doing, and analytics is the key.

For your Medevac needs - call or email us at admin@hiflyingllc.com

Share & Comment:





Contact Us

Call Us
Admin
+1 412 567 2211
whatsapp icon