How is Predictive Analytics Used in Healthcare?

Many global changes have pushed healthcare to meet new patient needs, but perhaps none were as significant as the COVID-19 pandemic and the subsequent number of people who needed or preferred to receive care remotely. 

The widening physical distance between the patient and their care provider has increased the need for strong data practices in the healthcare sector, in order to make smarter predictions about what care is needed and how it can proactively be provided. Read on to learn how predictive analytics is currently used across the healthcare industry, and how data analytics use cases continue to multiply in patient care.

Also read: Healthcare is Adopting AI Much Faster since the Pandemic Began

Healthcare and the Predictive Analytics Market

The global healthcare analytics market reached $1.8 billion in 2017 and is expected to grow at an astounding rate over the next several years, reaching a value of $8.5 billion by 2025, according to a report by Allied Market Research.

Although North America currently holds the greatest market share and is expected to retain that status in the coming years, the healthcare analytics market is also growing quickly in Europe and the Asia-Pacific region.

Some of the top providers in the market are listed below:

  • Cerner
  • IBM
  • SAS
  • Microsoft
  • UnitedHealth Group/Optum
  • Allscripts Healthcare Solutions
  • MedeAnalytics
  • Oracle
  • McKesson
  • Health Catalyst
  • WellSky
  • veda

Learn about the greater data analytics market here: Data Analytics Market Review 2021

Top Features of Healthcare Analytics Platforms

If you’re looking to invest in any sort of healthcare analytics platform, make sure you look for the following features while evaluating possible solutions:

  • EHR access: Your chosen solution should not only be able to access patients’ electronic health records (EHRs), but should be able to aggregate, sort, and apply that data to make various predictions.
  • Data governance: Strong healthcare analytics platforms include data governance and compliance features, especially to help users adhere to HIPAA and other regulatory requirements.
  • Real-time alerts: Automated alerts are particularly useful for patient-facing metrics, such as changes in condition.
  • Risk visualization: Risk should be assessed for a variety of patient metrics, ranging from fall risk to hospitalization risk to customer dissatisfaction rates. Risk visualizations can come in the form of risk scores and/or graphs that illustrate different risk patterns.
  • AI/ML: In some cases, AI/ML is used for automated tracking of patient care metrics and automated delivery of notes to care coordinators.

More on governance, risk, and compliance: Top Governance, Risk & Compliance (GRC) Tools of 2022

Benefits of Data Analytics in Healthcare

Healthcare data analytics platforms offer a slew of benefits to organizations that incorporate these tools into their daily operations.

Some of the top benefits are:

  • The ability to provide proactive vs. reactive care with predictive notifications
  • Preventive and transitional care guidance for care coordinators
  • Population-wide health management tracking
  • Optimized scheduling for patients and medical professionals
  • Payment/insurance management analytics to manage payor relations
  • Built-in data compliance strategy and checklists
  • Automated knowledge of care needs with AI/ML
  • Risk management for patients, payors, and care providers

More on healthcare and risk management: Why Healthcare Risk Management Is Important

Healthcare Predictive Analytics Use Cases

The healthcare industry can benefit from predictive analytics in a variety of use cases, not just in one-on-one patient care. Some of the following use cases have increased efficiencies and cost-savings across healthcare organizations.

Hospital Readmissions

Healthcare analytics platforms help care providers to predict the conditions that will most likely lead to hospital readmissions. Hospital readmission data is particularly relevant to senior care providers. Although this is still an experimental field in areas like non-medical home care, research and platforms in this area are growing at a rapid rate.

Population Health Management

Especially when EHR data is provided for a larger population pool, healthcare professionals and data analysts can determine necessary macro-health initiatives with risk scoring across that population.

Payment Automation and Coordination

Payor and insurance relations have traditionally involved significant amounts of manual data entry and user error when determining cost and billing distribution. Predictive tools are now used to predict Medicare, Medicaid, insurance, and private pay costs in advance, resulting in more proactive and accurate payment cycles.

Scheduling Management

Some healthcare analytics platforms look to optimize the schedules of care providers, allowing additional time to fit in emergency cases and making it easier to account for no-shows.

Supply Chain Management

Especially during global supply chain shortages, it’s important to know how frequently you go through your most important medical supplies so you can resupply accordingly. Several predictive platforms now make it possible to predict PPE and other equipment needs based on specific environmental and patient conditions.

More on healthcare data management and use cases: What are Key Features of Healthcare ERP Solutions?

An Expert Perspective on Healthcare Analytics

Dr. Bob Lindner, co-founder and CTO of veda, a healthcare data automation company, believes that predictive analytics is an important and growing initiative for healthcare intelligence. Read on for some of his thoughts on predictive analytics and data analytics trends in the healthcare sector:

CIO Insight: How are healthcare companies currently leveraging predictive analytics in their operational models?  

Lindner: Healthcare companies are starting to leverage predictive analytics in many ways, from helping providers make better decisions about patient care and reducing patient risk, to identifying and addressing inefficiencies and lowering costs. 

More specifically, in practice, predictive analytics is being used to: 

  • Lower costs by helping to predict resource needs, like equipment or staffing needs 
  • Integrate patient records with other health data to identify any red flags around potentially serious medical conditions 
  • Collect and integrate holistic data, such as lifestyle and past treatment data, to help providers make better decisions about a patient’s health in the context of their social determinants of health 
  • Determine which treatment options may work best for a particular patient 

For example, one healthcare company used predictive analytics to improve its customer experience. They looked at customers who gave a one-star rating for their final experience with the company and collected all the metadata, touchpoints, and inbound transcripts for this population throughout their entire customer experience. 

This data was used to train a supervised machine learning model to predict if any given customer was likely to end up with a one-star rating. They then ran the model on existing customers, and for any customer where the model predicted that they may end up as a one-star rating, the company intervened with customer support proactively to check on the customer and right the situation. In this way, they were able to act to “prevent” one-star experiences before they happened.  

Data analytics initiatives, like this one, pave the way for members and patients to have the frictionless experience in healthcare that they’ve come to expect from other industries. 

Selecting the right predictive analytics solution: Best Predictive Analytics Software for 2021

CIOI: How has data analytics changed over time in the healthcare industry, and how could it potentially grow in the future? 

Lindner: Predictive analytics has proven its worth in numerous industries, including technology, retail, and even real estate. However, the healthcare industry has lagged behind. It wasn’t too long ago that hospitals housed huge document storage rooms and hired file clerks to sort, alphabetize, and distribute medical documents into physical patient folders. 

With new technologies — from faster ways to communicate with patients to automating complex data processes — collecting and analyzing data has now become enormously important in addressing some of the healthcare industry’s most pressing challenges, such as high costs and meeting new regulatory requirements. 

For example, the No Surprises Act (NSA), which goes into effect on January 1, 2022, requires insurance companies to update their provider directories within 48 hours so patients can easily identify which providers are in-network and avoid surprise bills. While this sounds straightforward, health plans are processing an avalanche of provider data at any given time and it’s not being seamlessly shared between payers and providers — payers are literally receiving hundreds of different provider data from hundreds of different provider groups. 

The data is so messy that all plans have dedicated staff to assess, interpret, clean up, and then enter it manually into the plan’s system — taking weeks to update and resulting in millions of dollars spent and low accuracy rates.

That’s hugely problematic since plans rely on this information not only to update their provider directories, but also to determine which providers should be paid contracted rates, for specialty and licensing updates, and for billing information changes that are necessary to properly pay claims. 

While historically it has been challenging for the healthcare industry to address needs like this, recent advances in data analytics now make it possible. Moving forward, we’ll continue to see data analytics playing an essential role in reducing healthcare spending, as well as simplifying and streamlining the healthcare experience from the patient, to the payer, to the provider. 

CIOI: What trends are you currently seeing for data analytics solutions in healthcare?

Lindner: The COVID-19 pandemic has forced the healthcare industry to rethink how care has been traditionally delivered and data analytics is driving innovation towards improving operations. Here are three trends I’m seeing around data analytics in healthcare. 

Healthcare organizations and vendors are starting to turn their attention to finally solving data problems that have persisted for decades. Rather than continuing to be confined by conventional wisdom or by the traditional systems and processes that have limited healthcare’s potential, healthcare organizations and their partners are connecting innovation with practical problem-solving in ways other industries have done for years.

We now have the ability to take on data problems that simply haven’t had the necessary attention or imagination and were accepted as status quo. For example, by applying a machine learning approach from astronomy and the creativity of our experts, veda is solving complex human-generated data problems in healthcare that were deemed unsolvable. 

There is a renewed focus on interoperability. With new technologies comes a need to make sure they work together in the service of improving patient care and lowering costs. There’s no silver bullet for the interoperability challenge, and plenty of data silos exist across the healthcare ecosystem. But we are increasingly seeing technologies that can work as “translators,” sitting between disparate networks and enabling seamless data exchange so that one common standard isn’t necessary. 

Companies are rediscovering the value of their business rules. When implementing a predictive analytics solution, sometimes there exists a sentiment that “One shouldn’t use old-fashioned business rules in the system, we should use supervised machine learning models and let the data teach the system how to behave.”

The reality is that to solve a complex problem, one needs to bring every tool you have into use. If you consider the amount of training data needed to “teach” a supervised system a business rule from scratch, then a single business rule can be worth a million rows of training data.

Newer predictive analytics systems are using new algorithms to elegantly combine the flexibility of supervised learning alongside the prior knowledge from business rules to reach the highest performance. 

CIOI: What is the biggest roadblock that healthcare organizations run into when they start using data analytics in their model? 

Lindner: There are a few common roadblocks that I see when it comes to healthcare organizations deploying data analytics, and they’re easily solvable:  

Misunderstanding what data analytics can — and cannot — do. Data analytics cannot completely solve all of an organization’s problems, but if users can put in an upfront effort to codify how humans have been making decisions about processing data in their organization, then data platforms can apply that same decision-making framework in a much more efficient and consistent way.  

Thinking through the holistic process. One roadblock that companies face frequently when implementing data analytics and automation is not having a clear and documented process that they would like to improve laid out beforehand. Automation will draw attention to gaps in the process.

Predictive analytics and automation cannot make your corporate process, but they will supercharge it. Therefore, if your process has holes or dead ends, automation will amplify them to new heights. The way to prepare for adding predictive analytics and automation to a company is to lay out a comprehensive data flow chart of the before and after states. Don’t just focus on the after. 

Having a limited view of the range of data analytics solutions available. Not all data analytics solutions are created equal. Some require a “rip and replace” approach that may be disruptive to an existing IT infrastructure, but other solutions can coexist with current systems. Solutions also vary in terms of the type of data they can process.

It’s important to look for a vendor partner that provides human support in addition to technology. It’s not necessary to have a team of data scientists in-house to leverage data analytics. The right partner can assist with needs assessment and goal-setting, provide guidance on how to go about processing data, and set expectations around the timeline for implementing data analytics.  

CIOI: How can predictive analytics improve patient care? 

Lindner: Predictive analytics will no longer be a luxury, but a necessity to achieve better patient care. Predictive analytics can equip healthcare providers with meaningful insights — such as understanding population health needs and individuals’ social determinants of health, as well as being able to better identify and support care coordination.

This type of technology will help providers succeed in today’s value-based care environment by informing them how, when, and where care and resources are needed, eventually leading to improved clinical outcomes and patient satisfaction, as well as decreased healthcare costs. 

More on predictive analytics and risk management: Best Risk Management Software for 2022

About Dr. Bob Lindner

Dr. Bob Lindner, co-founder and CTO at veda, empowers and guides veda’s scientific and engineering teams.

He provides strategic vision, builds innovative core data science technologies, and connects veda scientists to the Scientific Advisory Board.

Dr. Bob Lindner, CTO of veda

Bob fell hard for data science and has a passion for solving big problems, not just in the greater universe, but also those affecting our everyday

 lives here on earth. With over 10 years of experience, he is a published and acclaimed astrophysicist with experience modeling data and designing and building cloud-based machine learning systems. Bob had a number of impressive and significant discoveries and “first detections” in his years of researching and studying.

Most notably, he created machine learning code that automates and accelerates the ability for scientists to analyze data from next-generation telescopes. That program, Gausspy, continues to increase scientists’ understanding of our galaxy’s origins. He earned his Ph.D. in physics from Rutgers University and was a postdoctoral researcher at UW-Madison, where he led the development of Gausspy. Bob is a contributing member of Forbes Technology Council.

Shelby Hiter
Shelby Hiter is a marketing content writer with more than five years of experience in writing and editing, focusing on healthcare, technology, data, enterprise IT, and technology marketing. She currently writes for three different digital publications in the technology industry: Datamation, Enterprise Networking Planet, and CIO Insight. When she’s not writing, Shelby loves finding group trivia events with friends, cross stitching decorations for her home, reading too many novels, and turning her puppy into a social media influencer.

Latest Articles