A study by the management consulting firm of McKinsey, “The New Growth Game: Beating the Market with Digital and Analytics” revealed that companies using digital analytics generate on average 50% more revenue than those not using these analytics.  Digital analytics also provides practical solutions for high-growth industries such as life sciences and healthcare by providing granular insights into customer behavior and purchase trends.

 Increased Data Accuracy

Increased data accuracy is another reason why providers should invest in digital analytics as it will provide accurate data insights that can be utilized for decision making purposes; lead qualifying activities; market segmentation; and prospecting campaigns.

In light of the significant benefits resulting from the use of digital analytics, it is no surprise that healthcare providers are swiftly converting to make use of this technology.

 Insights into Patient Behavior

The number of people in the United States who suffer from chronic diseases is skyrocketing. This increased demand on providers can equate to more pressure on them, especially financially. To make sure they meet the needs of their patients, healthcare providers can employ digital analytics to gain insights into their patients’ behaviors and needs and leverage those insights to spearhead chronic illness.

 Collecting Data is Not enough

Collecting data is not enough. They need to know how to analyze it and use it for their benefit. The right digital analytics can help them understand their patients better; identify potential new services; predict trends; and improve conversion rates on their website.

The right analytics can help providers answer questions like: What is happening with our website traffic?  How many people are calling in for appointments? What is happening with our ad campaigns? The answers to these questions requires a specific skill set in analytics.

 More Efficient Services

Analytics helps healthcare providers in making their services more efficient and effective.

Metrics such as conversion rates; average visit duration; and call bounce rates can assist healthcare providers to identify improvements to their services. Furthermore, analytics can be used to monitor the effectiveness of advertisement campaigns and how it affects the overall revenue and business performance.

Reference:

1. A Look at Challenges and Opportunities of Big Data Analytics in Healthcare https://www.mckinsey.com/~/media/McKinsey/Business%20Functions/Marketing%20and%20Sales/Our%20Insights/The%20new%20growth%20game/The-new-growth-game-Web.pdf

Today, many applications are utilizing machine learning, but we don’t always know how or to what capacity.  How do we integrate machine learning at our business?  How do we get from A, not using any supplemental machine learning tools to B, full integration and adoption?

Although not widely realized or known, many tools we utilize every day already integrate many facets of machine learning, we just don’t see it.  The algorithms behind intelligent search, suggestions on Netflix, Social Platform Censorship, and even Microsoft PowerPoint, are all powered by machine learning.  Most of the time when you see a PowerPoint referencing A.I.  What you really want to know about is the machine learning. So, what’s the difference between machine learning and AI?

Artificial Intelligence is more of a buzzword these days, you can buy a domain and claim to use AI in a presentation, but in all honestly, there’s probably a good amount of fluff mixed in. Many claim to use AI but have minimal achievements.  In application development, especially with AI and Data Science, we can often put the cart before the horse.  Most AI today is a pitch in a slide deck.  More recently, we’ve seen models as a series of regressions or neural networks, attempting to solve some really complex decision-making issues.  But at the enterprise business level these tools have a long way to go to reach unencumbered adoption.

The reality is that in order for AI to be true and robust, it would need to enable in-the-moment decision making with an influx of new data all the time.  The entire artificial intelligence module must be embedded at every layer of the company’s ecosystem.  To be truly effective, it would have to live in the Operating System (OS) of your computer, your e-mail, your devices, and company work product. It would have to integrate back with cloud services in real-time and analyze potentially terabytes of data in seconds. It would need a feedback system that tells it whether it did a good job or not, and some serious looks at how often re-training of the models is necessary.

Today, this is not only possible, it’s frequently happening at every level of the industry.  However, many tools that sit on top of those big players simply don’t have the in-depth subject matter expertise to solve the problem.  Maybe the model is too broad, or maybe the data isn’t internally consistent, it might not get you the answers you need when you need them.  Don’t get me wrong it’s here already but more so for the bigger market players than the niche consulting or legal firms.

While looking at these products and tools there’s a few things we should especially look out for, and questions to ask:

• Vet the Executive Team: Evaluate startups and review backgrounds. Look for technical expertise and many years of statistics and data science background. Bonus points for specializations in certain fields and certifications that are specific to you.
• Is it internally consistent? Ask some basic technical questions, and some tough ones. There are AI products that are totally bogus out there.  Talk to the vendors about where they get the data they use in the models, what models and techniques they use, and how they validate the model.  What would you check in our data to ensure the models work the same?
• Consider the fit within the organization. Can all the people at my organization use the tool, or is it specific to a select few “power users”? It’s not enough to just ask someone in the tech team if this will solve our problems, you need a team if you want it to work for a team. Start slow, with a use case. Some early adopters and supporters who share in your frustration may help bring the solution to life and assist in implementation. You need allies.
• Will the product evolve with the organization?

Now more than ever the importance of explaining business outcomes hits home for me.  Next time you see a company saying they do AI, ask them questions, make them prove it.

Ask about these processes to ensure that the data your next tool plans on integrating is being used in a way that makes sense to the business.  Where does your data live?  How will this tool extract it and how frequently?

Finally, are the models robust enough to actually get you an answer?  Many models out there are broad in nature and only provide probabilities.

The National Institute of Health (NIH) defines benchmarking as  “a strategic and analytical process of continuously measuring an organization’s products, services and practices against a recognized leader in the studied area for the purpose of improving business performance.”¹

Nearly every KHC client asks, “How is our performance compared to your other clients?  Do you have a benchmark for that?”  Hospitals and other providers have troves of data and a seemingly endless way of measuring their organization’s performance.  The other day, I worked through a limited dataset live with a compliance officer and in one hour and we identified five new metrics for observation services.  However, clients always want to know how they compare to others in order to measure success.

When making the leap directly to an external benchmark to determine if you perform better or worse than a peer, the context is overlooked.  Every company and situation is unique and external benchmarks do not contain enough information to determine if the data that supports the metric was determined in a manner that is consistent with how the company will utilize it that benchmark.

When considering a benchmark, there are key questions to help determine if it is meaningful to you:

1. What is the source of data? Self-reported data is subjective as different respondents may interpret questions or data points differently.
2. What is the lag time between data collection and reporting? Is a benchmark that is reported in June based on data from the prior year helpful or even relevant?
3. What key factors were considered in developing the benchmarks?  Payer mix?  Payer contract terms?  Geographic location?   Provider specialty?
4. What industry or market circumstances might be reflected in the data that does or does not apply to my organization?
5. What is not included in the benchmark?

I was recently digging through the Centers for Medicare and Medicaid Services (CMS) Part D Prescriber Provider Utilization and Payment (PUF) Data to find some contextually relevant benchmarks for a client.  The data is accompanied by a 28-page methodology document that outlines what is included and what is NOT included in the various data sets.²  Although there was a sufficient amount of data available, what was not included impacted any benchmarks in such a way that they were no longer meaningful.

I have also found that a significant amount of time is spent justifying the benchmark and explaining why it does not consider the specifics to the internal organization.  Importantly, benchmarks do not help identify the impact of changes made to the organization.

Operationally, it is more effective to measure the current performance against the desired results.  External benchmarks can be used to provide guideposts as to what the desired result is and a month over month analysis against an internal benchmark can demonstrate performance improvement in relevant areas based on the organization’s individual and unique context and goals.  This also considers the reality that the team that performs the work drives productivity, quality, and the final outcome.

Our auditing clients frequently request comparative error rates.  As each audit has a specific goal and each organization has different coding policies; coding team responsibilities; and physician involvement; comparative error rates can only be utilized as a high-level guideline to help establish what the desired result should be for that organization.

Consider physician compensation as another example.  HFM Magazine’s Summer 2021 issue has an article by Stuart Schaff, “We must stop relying so heavily on benchmark tables to set physician pay.” This statement succinctly explains why external benchmarks expose a hospital to risk because the hospital faces “difficulties not only in foreseeing the changes in benchmark data” and also details how the benchmarks generally used in physician compensation are based on a voluntary response and are self-reported.   Mr. Schaff instead proposes a physician compensation model that accounts for the physician’s specialty and other aspects of the physician’s work and pay.

In KHC compliance and litigation work, opposing experts frequently lose when we are able to explain why the analytics are not relevant to the case.  A few good examples are when are when an opposing expert used lower prices from 2018 to reprice claims from nearly a decade earlier, using Medicaid volumes from geographic locations with a lower Medicaid population to attempt to prove fraud or comparing nursing home ratings in one state to another to establish performance metrics.  In each of these examples, the benchmarks that were used were not of the same context as the measurement against which they were compared.

As one that lives in data, benchmarking is an analytics tool that I use daily as part of my work.  However, time must be spent refining the benchmark calculation to ensure that it is within the context of the problem.  Frequently, as I spend more time refining the calculation, I’m able to identify more than one benchmark that needs to be measured in order to measure against the goal.

Resources:

1. National Institutes of Health, Office of Management: https://ors.od.nih.gov/OD/OQM/benchmarking/Pages/benchmarking.aspx.
2. Medicare Fee-For Service Provider Utilization & Payment Data Part D Prescriber Public Use File: A Methodological Overview, https://www.cms.gov/Research-Statistics-Data-and-Systems/Statistics-Trends-and-Reports/Medicare-Provider-Charge-Data/Downloads/Prescriber_Methods.pdf.

Data problem?  Analytics?  Regulations?  Data privacy?  Humanity?  Coffee and talk?  You can reach me at jleventhal@kohlerhc.com or 312.933.2752.

Josh Leventhal is an expert in health and is Managing Director with Kohler HealthCare.  He has over 15 years’ hands-on experience in healthcare data and analytics solving problems for providers, payers and life science organizations.  Josh started his career in management consulting analyzing data for the largest joint defense litigations in the country before using his skills and expertise at local startups to assist the Medicaid managed care organization and medical research industries.  His experiences as a consultant, product manager and developer allow him to work effectively with both business and technology stakeholders.

1.     Not understanding data sources and how they work together (mixing data sources).

Data is one of the most valuable resources in the digital age.  However, not understanding how data is collected and processed can have a negative impact on the rest of your marketing efforts.

The most common mistakes made in data analysis are:  the mixing data sources (i.e. combining customer data from different channels) and not understanding how to filter or separate the good data from the bad data (i.e. using ineffective filters or separating irrelevant information).

2.      Not analyzing the data correctly because of insufficient understanding of statistical concepts.

Statistics is an important field of study for any data scientist or business analyst.  It’s necessary to know how to analyze data correctly so as not to misinterpret the results.

I will discuss two common mistakes that people make when analyzing data and the ways to avoid them.  The first mistake is not understanding statistical concepts such as a p-value means or a confidence interval.

The second mistake is misinterpreting the results.  For example, if you are trying to determine whether a company’s marketing campaign led more people to buy their product and the target cities in question are unknown, then the analysis will be misinterpreted because you are analyzing the wrong population.

3.      Too much focus on inputs and not enough on output.

Data is only useful if it is utilized in a way that has an impact on the outcome.  There needs to be a balance between using data and creating data that can be useful.

Data without proper analysis or interpretation will not result in any value and instead will just serve as expensive noise.  Data can be used for many different things; it may even be able to generate more data points that we would have never been able to produce ourselves.

4.      Ignoring visualization techniques (e.g., dashboards) for presenting insights to clients, stakeholders – those who have the power to adopt change.

Many people might not know that data visualization is the process of transforming information into something visual that can be understood at a glance.  Presenting insights in the form of visuals helps people better understand the data.

As data becomes more and more voluminous, it becomes difficult for humans to comprehend it all without the use of some type of visualization technique.  Data visualization techniques can help us detect patterns and trends in data and generate insights that we would otherwise never notice.

The most common types of visualizations include: charts; graphs; diagrams; maps; infographics; and illustrations.  These are quite effective tools as they relay more information in less time and with fewer words than if they were written on paper or spoken aloud verbally.

5.      Gaining insights from only one perspective or angle of analysis.

The only data perspective that people are aware of is their own.  By looking at the world through only their lens, individuals cannot develop an accurate understanding of the whole picture.  There are many ways in which people can gain a better understanding about a topic.  They can read other people’s perspectives; they can be exposed to new experiences and they can be given access to different types of data.  This is why there is a need for more perspectives when it comes to data analytics.