Healthcare Analytics, Population Health Management, Healthcare Big Data


How to Choose the Right Healthcare Big Data Analytics Tools

Picking the right big data analytics tools can be a major challenge. What are some of the top questions to consider during the process?

Source: Thinkstock

In just a few short years, the idea of “big data analytics” has transitioned from a mysterious new buzzword to an essential competency for healthcare organizations large and small. 

Analytics has moved from a lofty cutting-edge experiment to the foundations of regulatory programs like MACRA, and providers are no longer struggling with the question of how to acquire big data.  Now, they just need to learn how to use it.

The EHR Incentive Programs’ quick-march towards widespread health IT adoption, along with CMS’ ambitious push to expand adoption of value-based reimbursements, have delivered both the basic technologies and the pressing incentives to dive into population health management, predictive analytics, patient engagement, and cost-cutting operational overhauls.

From the most complex, sprawling integrated delivery networks to the solo family physician around the corner, healthcare providers are starting to think about what big data means to them, what problems they can solve with it, and what technology tools they need to get the job done.

The challenge for many organizations – thanks again, in large part, to the fast pace of the EHR Incentive Programs – is how to develop a big data analytics ecosystem that can produce quality improvements, keep users happy, and take advantage of existing technologies while minimizing the need to invest again in purchasing, optimizing, and learning a whole new suite of products.

Providers who are preparing to tackle this difficult task should start by taking an honest look at their infrastructure, their processes, their capabilities, and their goals before signing on the dotted line with any new vendor or business partner.  Here’s how.

Learn about what’s out there

The health IT industry is changing every day, and organizations that have been out of the market for a while may not be up-to-date on the latest and greatest offerings from eager new players in the field.

Real-time location systems that track the movement of medications, staff members, and iPads, clinical decision support tools for precision medicine, patient flow analytics monitoring the admissions and discharge processes, intelligent hospital beds that watch for sepsis, data lakes that predict high-risk patients headed for crisis, and smart hand sanitizer dispensers that keep track of hygiene all serve to solve different problems in the healthcare environment, and not every system is suitable for every type of organization.

The big data analytics technology marketplace can be divided into four different main categories based on the solutions they try to deliver.

Clinical analytics and decision support

Healthcare providers want to make the best possible decision for their patients, and they often need some extra help to do so.  Now that the vast majority of providers have adopted electronic health records, they have access to the basic big data that will allow them to engage in clinical analytics.

Clinical analytics can be patient-focused, such as using the EHR to compare a diabetic individual’s HbA1C readings over the past two years to benchmark data from other non-diabetic patients, using algorithms to create risk scores for post-surgical infections and 30-day readmissions based on vital signs, or using large-scale genomic data to match patients with rare cancers to personalized treatments.

They can also be employed for provider-facing improvement activities, like correlating outcomes and procedure codes to develop best practices for using a certain clinical pathway, or flagging specific clinicians with higher-than-anticipated prescription rates for opioids, MRIs, or antibiotics.

By combining EHR data and claims data in a big data warehouse, providers can develop clinical decision support modules, architect standardized treatment protocols, identify super-users of services, and proactively address patient needs based on trend lines in their clinical information.

Quality reporting and provider benchmarking

Clinical analytics are closely tied to quality reporting and benchmarking provider performance.  In the era of value-based care, financial performance penalties, and accountable care organizations, it is vital for providers to have a clear idea of where the weak links in the quality chain may lie.

The recent proliferation of performance indicators and clinical quality metrics has worried many stakeholders, due to the time and effort it takes to report on them.  But big data has played a significant role in raising the level of patient safety, demonstrating progress towards reducing healthcare spending, and coordinating care across disparate systems.

Quality reporting tools help providers address the following types of questions:

  • How can we reduce hospital acquired conditions and adverse patient safety events to avoid the associated penalties?
  • Which providers are exhibiting poor antibiotic stewardship habits?  Which may be deviating from clinical guidelines for certain conditions?
  • Why are certain patients overusing the emergency department and incurring significant costs for routine conditions?  What other options could we offer this population?
  • Are providers conducting recommended screenings for child development, chronic disease, and mental health?  Are they offering follow-up services and resources to patients?
  • How can we lower the number of preventable readmissions that we experience?  How can we improve coordination between the inpatient setting and primary care?

Both hospital and ambulatory providers have a great deal of room to grow into leveraging performance benchmarking tools for quality improvements, which may translate into lower costs, better outcomes, and fewer penalties.

Revenue cycle, administrative, and operational analytics

Finances are at the heart of every healthcare organization, and providers are keen to invest in revenue cycle analytics to gauge their operational health.  The upcoming MACRA legislation package strongly stresses the role of value-based care as the next phase of progress on the cost-cutting front, and participation in one of the alternative payment model (APM) options will require close monitoring of financial data streams.

Operational analytics can help providers improve their patient collection rates, manage the movement of patients and resources within the hospital, tighten up their supply chain processes, monitor progress towards achieving shared savings, and identify opportunities for operational improvements that could shave dollars off the expenses sheet.

Population health management and patient management

Population health management is a combination of clinical and operational analytics, and relies on risk stratification and grouping techniques to flag patients in need of a specific type of services. 

Population health is considered one of the foundational competencies for participation in accountable care organizations and the value-based environment, and so vendors have been pouring their resources into creating the technology tools to support providers as they embrace this new way of thinking.

EHR vendors, recognizing the role that population health will play in the near future, have started to offer patient management features as easy add-ons or even as part of their out-of-the-box packages. 

Care coordination and health information exchange environments – provided by state entities, local exchanges, and to a large segment of the marketplace by Epic Systems’ CareEverywhere system – connect providers across the continuum to smooth transitions, ensure informed decision-making, and make it easier for patients to receive the full spectrum of services that they need.

Population health management analytics help providers complete stratification and identification tasks such as finding all the diabetic patients in their attribution pool that have missed their check-ups in the past year, sending automated reminders to parents to bring their children in for their next immunizations, and flagging admission, discharge, and transfer (ADT) activities for chronic disease patients who may be in need of a subsequent primary care appointment. 

Understanding Your Big Data Analytics Technology Options

Leveraging Risk Stratification for Population Health Management

Figure out what you’re working with

The majority of providers may have one or two of the technology tools listed above, but might not have the complete set of systems required to become a fully data-driven organization.  Or they may have most of the key areas covered, but the systems they’re using are ten, twelve, or twenty years old and are reaching the end of their useful lifespan.

Organizations should conduct a thorough self-assessment in order to truly understand what they have, what they need, and what needs to be replaced. 

Systems that generate lots of data but can’t produce meaningful and digestible reports may need to be replaced by “smarter” tools that get to the crux of the matter more quickly.  It’s likely time to retire tools that haven’t received vendor support, security patches, or upgrades for the past five years. 

And those that were built on outdated proprietary data standards that preclude interoperability with new innovations?  It may be in the organization’s best interests if those end up in the trash.

That doesn’t mean that providers have to send everything to the scrap heap if it’s more than a few years old.  Many systems, including EHRs, can still be rescued with an optimization overhaul and a refresher on previously unused features that may stretch the value of the investment a little bit further.

In addition to taking stock of the IT department, providers should review their users’ workflow habits to sniff out dangerous workarounds, identify frustrations, and compile a clinical wish-list for changes and updates.

Decoding the Top 10 Buzzwords of Healthcare Big Data Analytics

The Difference Between Big Data and Smart Data in Healthcare

Decide what you want to do and who’s going to do it

Once healthcare organizations have a clear idea of their needs and their options, they should pick a few high-priority goals to target. 

These may include patient management tasks like improving diabetes care, developing a patient-centered medical home, reducing discharge and transfer times, or doing a better job of connecting patients with behavioral health resources.

They may also be financial in nature, such as joining an ACO, lowering emergency department utilization by ten percent, eliminating waste in the supply chain, or cutting lost revenue from unsuccessful collections by a certain amount.

Providers may wish to ask themselves some of the following questions to help find the best starting place:

  • What areas of care and performance are in need of support from a data-driven analytics platform?  Can we use patient and staff feedback to pinpoint a specific weakness?
  • Is there a project that will produce a short-term return on investment to build confidence, experience, and demonstrate success?
  • What are our upcoming regulatory challenges?  Are we being penalized for certain quality or performance issues?  Can we use big data to address them?
  • What parts of our data are readily accessible?  What data integrity or completeness tasks should we work on?  What new sources of data should we consider accessing to meet our analytics needs?
  • Where do we see our organization in five years?  Do we want to prioritize growth, cooperative partnerships, patient care quality, or our reputation within the community?  What data do we need to achieve those targets?

With goals identified, providers then need to build the right team to integrate big data analytics technology with processes and outcomes.

Starting with the executive suite or board, organizations should generate enthusiasm and buy-in by presenting a clear plan for achieving their targets, creating a budget that includes staff education and training, and setting firm timelines for implementation and go-live.

Enlisting the services of multiple departments, including the health information management office, the IT department, the administrative staff, and – perhaps most importantly – the clinical providers, will ensure that all stakeholders have a voice in the process and a seat at the table. 

Since each area of operations may have different needs and priorities, it is important to balance all available viewpoints before moving to what may be the most difficult step: deciding on a vendor and committing to a specific product.

Building the Team for Big Data Analytics, Population Health

10 Healthcare Big Data Analytics Outsourcing Mistakes to Avoid

Key Ways to Improve Claims Management and Reimbursement in the Healthcare Revenue Cycle

Pick the right business partners to help achieve your goals

Choosing a vendor is a process fraught with peril, and it isn’t always easy to know what to look for before making a commitment.  It’s important to start off with the right frame of mind, and to think about the vendor selection process as a cooperative, mutually beneficial partnership instead of adversarial situation in which unscrupulous vendors are out to swindle their customers and lock them in to sub-par technology suites.

That being said, there’s no harm in being cautious.  Interview all potential partners carefully, and know the warning signs around offerings that may be slick and exciting but completely unsuitable for achieving previously identified goals.

Key questions to ask when demoing a product include the following:

  • Is this product built on open standards widely accepted by the industry?  Does it accept or utilize APIs to connect with other systems?  Will it connect with my systems?
  • Are there hidden fees in the contracting structure for this offering?  Are there data exchange fees, storage fees, early termination penalties, or upgrade and maintenance charges?  If the product is subscription based, will the pricing change over time?
  • Who owns the data that moves through or is stored in this system?  What are the procedures for ending a partnership?  How much notice is required, and how will I access my data if I decide to terminate the contract?
  • Do you engage in third-party outsourcing for any of your services?  What are the entities involved, what part do they play in your offering, and do they meet industry security and operational standards?
  • If the offering is cloud-based, what is your downtime percentage?  What are options for continuity of service if the system goes down?  What are your disaster recovery procedures?
  • What do I do if I need technical support or additional training?  Who will be my primary contact for managing my account?  How long does it typically take to get answers to troubleshooting questions?
  • Is this product scalable and adaptable for my future needs?  Can it accept new sources of data for more advanced analytics, or do I need to purchase additional tools or capabilities?  How will this vendor grow with my organization and continue to provide support?

Vendors that act dismissively when asked these questions, or give vague and unsatisfactory answers, should be viewed with a certain skepticism.  Outside of the core EHR market, federal regulation and oversight of health IT products is still developing, and the tempting opportunities of an explosive marketplace may attract less-than-perfect developers to the field.

But even more important than weeding out the wooden nickels is ensuring that when an organization does commit to a product, they are not buying more than they really need or less than they desire. 

Big data analytics in healthcare is a highly personalized endeavor, and the system that may be perfect for one provider may deliver far too little or too much for another.  Vendors should understand the organization’s goals and how to achieve them, and should recommend products that are tailored to their size, specialty, and patient populations. 

Taking the time to conduct a careful internal assessment, engage in a healthy amount of market research, and thoroughly vet potential vendors will ensure that healthcare organizations make intelligent, considered choices for their big data needs. 

What to Look for in a Healthcare Big Data Analytics Vendor

Which Vendors Lead the Healthcare Big Data Analytics Market?

This article was originally published on July 1, 2016.


Join 25,000 of your peers

Register for free to get access to all our articles, webcasts, white papers and exclusive interviews.

Our privacy policy

no, thanks

Continue to site...