Tom Lennon, Author at Perficient Blogs

Recap: Data Lakehouse Blog Series

Tom Lennon — Tue, 14 Mar 2023 19:51:41 +0000

Perficient and AWS recently partnered with HIMSS to host a webinar that went in-depth about Perficient’s Healthy Lakehouse solution, specifically:

Driving insights across the enterprise through the pragmatic, scalable implementation of a data lakehouse
Maximizing business value by strategically approaching cloud-based data modernization, not merely as a move from on-premises systems
Understanding the value of choosing a cloud partner that is deeply invested in solving healthcare’s unique business and consumer needs

The webinar had a panel of experts from Perficient and AWS that included:

Tom Lennon: Practice Director, Healthcare Data & Analytics, Perficient

Priyal Patel: Director, Healthcare Strategy & Solutions, Perficient

Rahul Ghate: Head of Worldwide Partner Development for Healthcare & Life Sciences Data Analytics, AWS

You can download the full webinar recording, here.

Blog Series Recap

We have put together a blog series building off the webinar and going more in-depth about Perficient’s Healthy Lakehouse solution. You can find the full blog series, here.

Here is a high-level overview of the blog series:

Blog 1 Summary: Driving ROI in Healthcare with Data and Analytics Modernization

Most healthcare organizations we work with have taken some steps towards modernizing their data and analytics capabilities. However, they’ve not been able to truly take advantage of the full set of benefits and innovation that this modernization vision offers. We’ve seen some movement to the cloud, mainly lifting and shifting data. However, much of this, really, just replicates many of the same legacy challenges that existed in the on-prem world and don’t take full advantage of the momentum that could be leveraged from these initiatives and investment.

Data Modernization-Healthcare organizations need to take steps to increase data security. Historically, healthcare leaders have been understandably hesitant to give up control of patient and member data. Still, the reality is the investment and skill of AWS’ cloud-data security team is multiple orders of magnitude greater than that of any single healthcare organization. So, we should view cloud migration as one step in helping us grow our data security capabilities.

They also need to develop the capabilities to provision and instantiate new environments quickly and reliably – this is a huge benefit for data & analytics project teams, and reduces the effort, timeline, and cost to deploy environments compared to traditional manual deployments. And this eases the burden of spinning-up sandboxes for discovery-based analysis.

Analytics Modernization-Driving analytics value and insight to business users and stakeholders should be the focus for teams. To accomplish this, we need to enable seamless access to the right information. This includes BI scorecards and visualizations of course, but also self-directed analysis that is or should be, table-stakes for any modern analytics program. This means enabling teams to create accessible discovery sandboxes quickly and easily for ad-hoc exploration by power users and data scientists without sacrificing the necessary governance and security.

When we get this right, we can advance beyond simply providing insight to better understand healthcare and achieve the capabilities to automate functions that reduce the load on our completely overburdened HC workers, which, as we know, is a formidable threat to continued success in delivering healthcare to our population.

Blog 2 Summary: Perficient Healthy Lakehouse Accelerates Time and Cost to Value

Through past implementations, Perficient has learned a lot about what adds time, effort, and cost to healthcare data & analytics projects. We’ve also learned a bit about what can be done to reduce the lift involved.

We’ve pulled together a set of architectural design patterns, and frameworks, which can help reduce the time and cost it takes to get to value when building a healthcare data lake, data warehouse, and analytics solution. We call this framework the Healthy Lakehouse. The Healthy Lakehouse is based on field-proven AWS services and solution design patterns which Perficient has developed while implementing solutions for clients in the Provider, Payer, PBM, Pharma, and Clinical Research domains.

Creating this type of framework is something any organization can potentially do on its own, but starting from scratch takes time and money, and it’s increasingly hard for most of our clients to keep up with the pace of innovation by themselves.

The business insight which can be derived from the Lakehouse is really a function of the data which is provisioned within, and the analytics developed based on that data. The Business Insights listed on the right side of the chart provide an overview of just some of the key domains of insight PRFT has developed with past clients, although this should not be seen as a finite or limited list of what is possible.

The Healthy Lakehouse framework leverages the AWS Cloud Platform and Amazon Data and Analytics Services including:

- AWS Glue, Kinesis, S3 Buckets, DynamoDB, Redshift, Athena, Elastic Map Reduce (EMR), Amazon Quicksight, and Sagemaker

Blog 3 Summary: Use Cases and Successes for Perficient’s Healthy Lakehouse Solution

We shared real-life examples of projects and solutions we’ve executed with our clients including a home health success story.

One of the country’s largest home healthcare providers asked Perficient to assist them to define, architect, and implement a modern data & analytics solution. They wanted to become an operationally integrated, data-driven organization and realized they needed help getting there.

The solution was a fully functional cloud-based data lakehouse and analytics solution that provided financial and operational performance analytics across operating units, and included analytics like:

A staff-hours & revenue dashboard
A Skilled vs Unskilled worker performance scorecard
and Billing trend-analysis visualizations

This solution eliminated the need to have 4 highly skilled analysts spend upward of 2 weeks per quarter collecting and correlating this information just to create these executive dashboards. These analysts can now focus on identifying ways to improve performance as opposed to wrangling data and performing swivel chair integration in excel.

Perficient + AWS

Perficient understands the complexities of the healthcare industry and the unique challenges healthcare organizations face. Our healthcare practice delivers strategic business and technology consulting insights that help our clients transform with today’s digital consumer experience demands. This strategic guidance is then transformed into pragmatic technology solutions that improve clinical, financial, and operational efficiency.

As an AWS partner, we have extensive experience with Big Data, data warehousing, business intelligence, and analytics, and work with our clients to transform data into timely and actionable insights using AWS services such as AWS Glue, Amazon Redshift, Amazon Quicksight, and Athena.

Perficient’s healthcare practice is dedicated to helping healthcare organizations leverage data and analytics to improve care quality, access, and delivery, and to better manage the costs associated with providing that care.

Use Cases and Successes for Perficient’s Healthy Lakehouse Solution

Tom Lennon — Tue, 28 Feb 2023 21:03:14 +0000

In blog 2 of our data lakehouse blog series, we discussed how Perficient’s Healthy Lakehouse solution accelerates time and cost to value and can help crack the code to success in delivering solutions.

Next, we’d like to share some real-life examples of projects and solutions we’ve executed with our clients. These charts show just a few of the use cases we’ve built for past provider, payer, and life sciences clients, to paint a picture of the actual analytics value we’ve delivered.

Now let’s take a more in-depth look at a few specific projects and solutions which Perficient has executed with our clients.

Home Health Success Story

For some background, this client experienced significant growth through acquisition over the past few years, which left their data sprawled across many systems. This data sprawl required significant manual effort to create operational and financial performance dashboards across operating units. Creating these dashboards involved first locating and collecting the necessary data, then stitching it together within spreadsheets, and then manually creating the dashboard itself.

There was also a real risk of a HIPAA data incident occurring due to limited controls around PHI and PII data, given the data sprawl mentioned.

To meet the client’s needs, we began by discovering the current data & analytics capabilities and challenges across people, processes, and technologies, and worked to define a realistic future state vision, strategy, and plan to get there. This included:

First understanding and prioritizing the business and IT needs and challenges
Defining the platform and program architecture, AND selecting the cloud platform and tools,
And defining the program structure, project organization, and execution plan to implement the roadmap.

Once the architecture and plan were in place, the hard work of building the solution began.

Building a Solution

This started with a technical proof of concept and an initial MVP which was brought to production, followed by a series of quarterly releases which added incremental value as the data and analytics capabilities grew.

A staff-hours & revenue dashboard
A Skilled vs Unskilled worker performance scorecard
and Billing trend-analysis visualizations

As an additional benefit, combining the data from multiple operating companies has allowed the organization to enhance its data governance capabilities and program, and help ensure data is being managed and protected from a HIPAA and HiTrust compliance perspective.

Academic Medical Center Case Study

This was an academic medical center that was running into challenges to integrate large data sets securely and quickly to make the data usable and valuable to its providers and researchers.

They initially had a traditional, on-premises enterprise data warehouse to store and analyze data. But this solution was too costly and could not scale to meet their business needs. They were spending a ton of time and money just to produce basic reports and dashboards to draw the simplest of correlations among their patient populations.

So, they decided to invest in a secure and scalable cloud solution that would reduce maintenance costs, increase efficiency, and provide the data and analytics platform they needed to power their journey toward translational and personalized medicine.

The cloud solution ended up providing them with a robust, self-service platform for quickly analyzing complex data sets, compiling large data analytics, providing structured data visualization, and incorporating a data distribution environment.

More importantly, it got data into the hands of the data scientists who were performing the analysis and data modeling, and the clinicians who will ultimately be driving decisions for their patients.

We were able to:

Integrate 6 million patient records
Reduce operating costs by 50%, which helped free up funds for other vital program development
Reduce data query times by 97%, allowing for accelerated research, which in turn was able to get more evidenced-based care into action

Also, the scalable storage easily met the research and clinical demands of the organization.

Perficient + AWS

Perficient Healthy Lakehouse Accelerates Time and Cost to Value

Tom Lennon — Mon, 13 Feb 2023 16:49:37 +0000

In part 1 of our blog series, we talked about driving ROI in healthcare with data analytics modernization.

In this blog, we will discuss accelerating time and money to value with the Perficient Healthy Lakehouse.

Over the past few years, we’ve pulled together a set of architectural design patterns, and frameworks, which can help reduce the time and cost it takes to get to value when building a healthcare data lake, data warehouse, and analytics solution. We call this framework the Healthy Lakehouse. The Healthy Lakehouse is based on field-proven AWS services and solution design patterns which Perficient has developed while implementing solutions for clients in the Provider, Payer, PBM, Pharma, and Clinical Research domains.

The Healthy Lakehouse also borrows data and analytics modernization patterns from other industries, such as Financial Services, Automotive, and Manufacturing.

Please note that we expect the frameworks used within the Healthy Lakehouse to be adapted to your needs, since healthcare source systems, standards, and at times semantic meaning, can vary broadly from client to client.

Perficient brings the Healthy Lakehouse framework to new-client engagements, to accelerate and drive momentum in the early phases of the journey. As most that have been through this process would attest, getting started is often the hardest and slowest part of the journey. And getting it wrong can be disastrous.

Let’s take a moment to walk through the Perficient Healthy Lakehouse at a high level:

It starts with collecting and acquiring data from a variety of sources. There are many different types and flavors of data sources that have different interfaces, formats, standards, and/or semantic meanings. The Healthy Lakehouse Data Acquisition & Ingest Framework provides proven design patterns for the types of acquisition and ingests shown in the chart. Having these patterns at the start of the initiative saves time and money, and also helps drive standardization as data ingest pipelines are designed and built.

Looking to the center of the chart, under the Lakehouse roof if you will, the Lakehouse includes a Raw layer where data is landed, a curated layer that is in fact a processing layer that standardizes and conforms the data, and a publish layer to hold and present data for access… think data marts and views in the publish layer.

- The data lake itself uses Amazon S3 Buckets to store data,
- Curation is accomplished via Amazon Glue-based pipelines which include metadata capture within the Glue Data Catalog, and selective data movement into the Publish layer,
- The Publish layer is an Amazon Redshift data store that holds and presents data for consistent use in analytics.
- There’s also an Operational Framework that acts as a guide to ensure operational best practices such as audit-balance-and control are included from the start.
- The Lakehouse also includes design patterns to enable data tenancy tags to be applied and managed within the Lake and Publish layers.

The data within the lake can be exposed for use in different ways and for different purposes

- - Data is typically used to drive BI style reporting, visualizations, and analytical trending, using Amazon Quicksight as the analytics presentation tool.
  - Data can also be fed to ML models and AI generally, using Amazon Sagemaker, as well as serving data to other applications via APIs
  - We refer to this area collectively as the Data Intelligence and Data Services Layer

As mentioned throughout, the Healthy Lakehouse framework leverages the AWS Cloud Platform and Amazon Data and Analytics Services, as shown on the chart, including:

- AWS Glue, Kinesis, S3 Buckets, DynamoDB, Redshift, Athena, Elastic Map Reduce (EMR), Amazon Quicksight, and Sagemaker

Success Story: Modernizing Data and Analytics in Healthcare

A global health services company needed a more comprehensive view of finances (e.g., distribution revenue, costs, and rebates) associated with the products it sells and delivers. It also needed to ensure the accuracy of financial transactions and enable rebate reporting.

As this client’s trusted partner for AWS-based data and analytics modernization, we developed a solution to meet these needs with a platform through DevSecOps work with AWS account structures. This solution provides better data insights for pharmaceutical distribution, invoicing, and rebates to the client’s internal and external partners.

Perficient + AWS

Driving ROI in Healthcare with Data Analytics Modernization

Tom Lennon — Tue, 24 Jan 2023 22:25:12 +0000

According to a cross-industry study by MIT and Databricks, only 13% of organizations deliver on their data strategy. Now, as we all know, healthcare organizations, especially providers, have been through a lot over the last few years, and this doesn’t make delivering on their data strategy any easier.

This is why we ask ourselves daily: ‘how can we help healthcare organizations modernize their data & analytics capabilities?’ We firmly believe that success is not due just to the systems or applications you buy and implement, or even which cloud vendor you use… it’s about growing the right capabilities across your organization, including IT, within Business, Operations, and Clinical functions and staff.

Most healthcare organizations we work with have taken some steps towards modernizing their data and analytics capabilities. However, they’ve not been able to truly take advantage of the full set of benefits and innovation that this modernization vision offers. We’ve seen some movement to the cloud, mainly lifting and shifting data. However, much of this, really, just replicates many of the same legacy challenges that existed in the on-prem world and don’t take full advantage of the momentum that could be leveraged from these initiatives and investment.

Let’s take a minute to dig into what we really mean by data and analytics modernization and what some of the benefits of embracing this modernization really are:

Data Modernization in Healthcare

For starters, there’s a large opportunity for most organizations to reduce the cost and complexity of their data management environment and programs. The low-hanging fruit to get real ROI is to:

Retire redundant on-prem systems.
Leverage their OpEx budget instead of their CapEx budget, where this makes sense.
Push some of your data & analytics administration and support activities to your cloud provider.
But in addition to direct ROI, this evolution allows you to focus resources and staff on things that truly add value to your core mission instead of using these resources to manage your IT infrastructure.

Next, as we all know, healthcare data is under constant attack. The rate at which ransom attacks and data breaches are occurring is simply out of control, and the number of attacks seems to increase every month. No wonder we all get 4 or 5 emails a week advertising security offerings or services.

Healthcare organizations need to take steps to increase data security. Historically, healthcare leaders have been understandably hesitant to give up control of patient and member data. Still, the reality is the investment and skill of AWS’ cloud-data security team is multiple orders of magnitude greater than that of any single healthcare organization. So we should view cloud migration as one step in helping us grow our data security capabilities.

Many organizations have done a lift and shift of data. Still, they have not fully embraced modernization, leveraging development techniques such as DevOps infrastructure as code, continuous integration and deployment, and containerization.

Organizations need to develop the capabilities to provision and instantiate new environments quickly and reliably – this is a huge benefit for data & analytics project teams, and reduces the effort, timeline, and cost to deploy environments compared to traditional manual deployments. And this eases the burden of spinning-up sandboxes for discovery-based analysis

We should also employ modern DataOps techniques that automate the execution, orchestration, and monitoring of operational workloads. This helps reduce user downtime and provides more predictable data processing operations with less direct human intervention.

Over the past few years, the perspective on which and how much of our data we should curate and publish has shifted, limiting the data we curate and transform to only that data needed to satisfy known, priority use cases, as opposed to the traditional Data Warehousing practice of curating and integrating everything you touch.

There’s a pressing need to ingest and leverage an ever-increasing variety and volume of data. This data comes from many places, including:

Streamed data from devices and patient monitoring… think Internet-of-Medical-Things data, data captured within the digital engagement experience, and alternative digital health modalities like telehealth, not to mention unstructured data from clinical notes, imaging, audio, as well as data generated through social media. This is all in addition to the pile of data that comes from the traditional EMR, labs, and other HC systems.
Capturing and managing this data involves massive amounts of data ingest, storage, and processing capacity.
Although much of this could be handled, in theory, with current on-premises technologies, cloud platforms like AWS have services already built into the platform, such as AWS Glue, Athena, and Redshift to help solve many of these challenges, and they can scale when and as needed.

Analytics Modernization

Switching gears, we’ve discussed a lot about data modernization but what about analytics modernization?

Many teams get caught up in, or perhaps overly consumed by, the need to deal with all this data. But let’s not lose sight of the real value here: to drive analytics value and insight to business users and stakeholders.

To accomplish this, we need to enable seamless access to the right information. This includes BI scorecards and visualizations of course, but also self-directed analysis that is, or should be, table-stakes for any modern analytics program. This means enabling teams to quickly and easily create accessible discovery sandboxes for ad-hoc exploration by power users and data scientists without sacrificing the necessary governance and security.

Success with self-service analytics requires using Data Catalogs and Business Glossaries to allow users to locate, access, and trust the data they need and provide mechanisms to interpret the semantic meaning of similarly named data, which is so important to use data appropriately and responsibly.

If we look at analytics and insight from a forward-looking perspective, we need to prepare for the expanded use of AI and Robotic Process Automation. Doing so requires providing an increased level of trust in your data and the ability to manage the bias inherent within that data.

We need to grow the capabilities to create and use predictive models and ML algorithms, allowing us to move beyond the typical ‘reactive response’ and enable ‘predictive responses’. Sustained success with AI and ML requires capabilities to manage these rapidly changing and growing predictive models and algorithms through employing ML-Ops tools and techniques.

As we embrace modernization, we should also consider the solution and product development process, meaning Agile, use-case-centered project execution, and treating analytics solutions with a product mindset, which helps focus on the value delivered by that product instead of the process of creating that value.

Success Story: Modernizing Data and Analytics in Healthcare

Perficient + AWS

3 Problems That Data and Analytics Can Help Solve in Healthcare

Tom Lennon — Thu, 09 Jan 2020 14:08:34 +0000

I’m often asked how data and analytics can help to solve key industry problems in healthcare. With that in mind, three key industry issues rise to the top of the list.

1. Cost of Care Delivery

The cost of care delivery is at the center of the problems facing the healthcare Industry. Healthcare spending accounts for ~18% of US GDP. Although industry actors are working to increase the efficiency of care delivery, there is significant pressure on revenue with newer payment/reimbursement models making it difficult to even maintain historical financial parity.

There is a critical need to use data and analytics to identify trends that enable healthcare organizations to increase the effectiveness of care, reduce errors, better understand risk, reduce costs, increase operational efficiency, and capture maximum reimbursements for care delivery. Healthcare has been slow to implement modern data and analytics capabilities, leaving healthcare leaders without the proper information to make decisions and affect positive change.

2. Industry Consolidation

In the quest to increase efficiencies, industry consolidation has been rampant. Although consolidation promises long term operational efficiencies, it typically has a long payoff from an information visibility and insight perspective. Hospitals and payers are complex businesses and organizations, have complex data and applications systems, and are subject to many regulatory rules and hurdles, particularly around data security. When large players are combined, it typically takes years to achieve a reasonable level of consistency and access to data (information), which increases the blind spots mentioned above.

Healthcare organizations need help establishing a common view of data (information) across these complex organizations. If approached in the right way, modern data and analytics architectures, technologies and practices, collectively “Data and Analytics Programs,” can be leveraged to enable significant increases in efficiency and scale of data management and analytics systems, enabling a consistent and trusted view of healthcare data (information). And the relative cost of these modern data and analytics programs is typically well below that of the legacy programs and approaches.

3. Increase in Available Data

The proliferation of electronic health records systems, medical devices, and digital health has resulted in huge increases in the volume and variety of healthcare data, and is still picking up speed – this is truly Big Data. This presents vast opportunities to improve care through clinical research, improved care paths, mobile health and otherwise, however, it also presents significant data management and governance challenges for healthcare organizations.

Healthcare organizations are starved for the architectures, tools, processes, and policies needed to drive consistency, access, security, understanding, trust, and management of this deluge of Big Data. Unlocking the treasure trove of value held within this data requires implementing modern data management, analytics and governance systems, and programs to turn this data into information. This includes modern BI, predictive analytics, and artificial intelligence systems to enable forward-looking insight and action from this information.

What is Data Modernization?

There is a critical need in most healthcare organizations to modernize their data and analytics programs and capabilities to take advantage of the ever-growing amount of information available.

Data Modernization – Capabilities and Benefits

Common Use Cases

Leveraging data and analytics can be key in helping to make improvements, gain insights and realize efficiencies for multiple healthcare categories and needs, including:

Reduced ED/Urgent Care Wait Times
Readmission/Re-hospitalization Prediction and Reporting
Contract Management (payer scorecards; contract performances, etc.)
Patient Satisfaction
Regulatory Items (HEDIS; Stars; P4P; ACO, etc.)
Population Health (risk management; quality care; registry items)
Leakage Analysis
Utilization (though you could probably fold this into provider performance or service line)
Labor Productivity (nursing hours; RVUs, etc.)
Treatment and Medication Trend Analysis (top conditions, eligibility, risk score, cost-sharing, PMPY trend, Price and Use, High-Cost Claimants)
Provider Performance Analysis – Efficiency ($s) and Quality
Disease Management Analysis
Facility Analysis
Claims Denial Analysis
Preventive Services (gaps in care)
Disease Surveillance
Diagnosis Prevalence
Service Line Analytics

Summary

To achieve Healthcare’s Triple Aim (improving population health, improving patient experience, and reducing the cost of care), healthcare organizations need to take advantage of the insights available within the ever-increasing volume, variety, and velocity of data being produced in our always-on and always-connected world. Turning this data into insight requires leveraging modern data and analytics architectures and capabilities.

3 Hybrid Cloud Considerations for Healthcare Data

Tom Lennon — Mon, 29 Jan 2018 19:34:15 +0000

I recently joined my colleague Jim Kouba, director, Healthcare Solutions at Perficient, and HIMSS Analytics Senior Director James Gaston to present our webinar, Moving to the Cloud: Modernizing Data Architecture in Healthcare. In the webinar, we discussed the benefits and risks of moving data and analytics environments to the cloud and the main healthcare use cases for cloud migration.

My portion of the discussion took an in-depth look at one of two healthcare organizations’ cloud journeys, including the vision, challenges, and key takeaways. Our client, a leader in the academic health center space, needed a data and analytics platform and program to power the organization’s journey to translational and personalized medicine in support of leading edge value-based care delivery models. They wanted to leverage the wealth of information the collective organization has across the full population of pediatric and adult patients to improve care in the short term and provide game-changing innovation over the long term.

Some of the key challenges our client faced:

Complex architecture, infrastructure, and operations
Must scale to handle large amounts of data
Security concerns all around

Key benefits included:

Integrated 6 million adult and pediatric patient records
Reduced operating cost by 50 percent (reallocate into other areas)
Delivered a broader and richer set of tools and technologies to data scientists and clinical decision makers; significantly decreasing prep time to allow a focus on improving patient care
Rapid, iterative development of visually oriented analytics (disease surveillance, diagnosis prevalence)

At the end of the webinar, we received some great questions from the audience, including:

What are the benefits you see to a hybrid approach as opposed to just an on-prem or cloud-only environment?

To answer this question, you first need to consider the following:

What are your organization’s needs, vision, and drivers, and what are you trying to accomplish?
What is realistic given your current environment, processes, capabilities, and culture?
What components of your in-house architecture and processes provide a competitive or strategic advantage, and which do not?

Most organizations have a large investment in their in-house technologies and processes, and a hybrid approach allows a greater degree of control over what leaves the organization’s four-walls and what remains in-house. This is as much a cultural challenge as a technical challenge.

Control over information security, both real and perceived, must, of course, be considered, but this also applies to operations and processes that could be impacted by a shift to the cloud. It’s important to consider what your critical areas of strength and market advantage are to ensure the shift to the cloud increases your leverage in these areas and does not outsource capabilities that your provider cannot truly replicate. So be sure to do your homework to understand your chosen cloud provider’s strengths and weaknesses.

One final note: Regardless of your approach, having a well-vetted strategy and roadmap, which includes why, how, and when you will take actions and see results, is critical to achieving your goals. This allows realistic expectations to be set across the organization.

For all of the details on the project, the other client success story we covered during the webinar, and the entire Q&A session, view the full on-demand webinar here.