Here are 5 important aspects to consider, and tips of what to look for, so you can trust the insights in your SDoH analytic endeavors.

1. How is the data integrated?
2. How specific and granular is the underlying data?
3. What is the social determinant being analyzed?
4. Can you clearly understand the definitions and data sources used for insights?
5. How trusted is the health data itself?

1. TIP: Find out the options you have around individual member address data. Ask questions about the quality and completeness of these fields. Ideally your solution will have the flexibility to use or assemble the most complete collection of member addresses possible.
2. TIP: The best solutions offer a member-level integration to at least census block group level.  That associates people to the social and environmental factors known to a neighborhood level of insight.

1. TIP: Make sure your solution offers data and SDoH indices that meet broad investigative needs.  Most organizations have many questions and require multiple SDoH indices. In a discovery phase – a few options let users understand opportunities to act impactfully based upon different criteria.
2. TIP: Consider ways to allow analytic journeys to mature. Composite indices can be great for initial analysis. As a team starts to work on designing for a barrier or opportunity, a more specific SDoH indice will reveal important nuances or details.

TIP: Enriching claim data delivers fast and intuitive investigations. This makes SDoH analytics easier too.

Enrichment can have many forms: classify claims by episode treatment groups (ETG), apply pharmaceutical classifications, and flag specialty druges. Enrichment processing also identifies gaps in care and low value care and makes it easy to surface these individual moments into analytics.

ER visits that have been classified using the NYU methodology allow you to quickly look at who visited the ER for non-emergent care, just by using a few filters.  Now think how powerful it is to further see these visits by income index.

What’s a point solution?

Most firms offer programs to identify health issues and manage chronic conditions (health risk assessments, biometric screenings, and health promotion programs).

83% of large firms offer a program in at least one of these areas: smoking cessation, weight management, and behavioral or lifestyle coaching.

Source: Kaiser Family Foundation study

Use a cohort strategy to evaluate point solutions.

Cohort comparisons are the ultimate analytic strategy for proving value. Without a direct comparison within the same population, there are so many factors that introduce doubt on what the numbers truly capture. Alternatively, by looking at well defined and specifically differentiated groupings of people, we can directly compare performance take away concrete and specific learnings.

1. TIP: Look at related costs across your cohorts: Determine if there is value beyond just the immediate program financials. For instance, we have looked at disability claims, to measure the influence of a point solution program.
2. TIP: Look at related health concerns: Investigate other aspects of wellbeing to see if there are notable halo effects.  For instance, we have investigated if there are mental health differences across maternity program types, short and longer term.

Here’s a good example from our client base: This national retailer wanted to measure the value of a Center of Excellence strategy for heart conditions.  The metric strategy compared a well-defined pair of cohorts that looked beyond traditional utilization and cost metrics.  We helped them also include mortality rates (COE – lower), returns to work (COE – faster), outcomes (COE – better), and company satisfaction (COE – higher).  Yes, that’s right – employees actually reported a higher employee satisfaction rate on the survey following a major episode of care.

Ask the right analytic questions.

Often “What’s the value?” is the wrong question. The correct question is “Who is this valuable for?” or “What’s the incremental value?”

There will always be a portion of a population that is engaged in their health and wellness. Your data can tell you who this population is, and provide insights that help you identify more people “like them” that you can target and pull along, therefore increasing program value. Also consider if the engaged audience would have been healthy or well without the special program, in some other way. Is it the program – or the people – that are providing the results you see?

Use ALL the data we have available in today’s analytic world.

Understand how social determinants of health influence engagement and utilization.  Then optimize the point solution to meet broader needs by removing barriers.  The data can show you where actions will be impactful.

Data that provides insights into social determinants of health can be time consuming to assemble into an analytic environment and then align to member health data. And yet it’s so powerful for insights. Your analysts time is better spent using this data as opposed to prepping it manually.

We evaluated medical and dental claims for diabetics after the introduction of a new Virtual PCP program.  The solution was selected after seeing a statistically significant difference in PCP utilization across various household income segments.  We created a specific scope around diabetics to study impacts on utilization, medication adherence, medical costs, and co-morbidities in mental health.  Not all investigation can rely solely on data.  The task force team worked with “Voice of the Member” groups, formed based on specific demographics. They focused on understanding context and color behind the numbers.  Transportation, time away from work, and caregiving themes arose in the care access category.  Other reasons were also presented, but offered less immediately actionable solutions.

With less time prepping data, the team had more time to dig deep, address quantified specific barriers, and is now measuring impact.

You are more likely to be on-time and on-budget if your plan is thoughtful and reflects your reality. Have confidence and know what you’re really getting into!  Health data is very complex.

Here are three Best Practices for a successful
health data and analytics platform implementation:

One-size does not fit all. It’s unlikely your implementation is the same as other organizations.  Why?

Organizational culture is a HUGE factor.

Dig in.  What are the details behind YOUR implementation plan?

Discuss what will be problematic or painful based on your experience and what you are moving away from. Are those complexities appropriately addressed, cared for, or resourced? Think about metric definitions and consensus, data quality, data reconciliation, matching and integration across sources, and slowly changing history.

Identify what is- and is-not in your control. If something is beyond your direct control, is there a named resource and escalation path?  What risk does that pose to the project timeline based its nature.  For instance, your health analytics implementation is reliant on data from others.  How are your relationships and service level agreements with those partners and vendors?  How does that affect your plan and what’s the back-up plan?

Before your implementation starts, refresh your knowledge of the day-to-day contacts, authorities, and any contractual SLA’s you have in place. If there will be costs associated with establishing new feeds or data interfaces, identify those early.

Top down, bottom up, or an interesting mix? Think about the approach that will work better for your organization.  What process works for you – here’s my data – what can I do with it?  Or here are my objectives – what data do I need?  There are pros and cons to each but thinking about this as you prioritize is invaluable for setting internal expectations and getting the right resources lined up.

Use phase 1 for quick wins. Standard sources generally seamlessly populate the most common views. Users feel like they get a lot out of the gate and that helps tremendously with adoption.

Vice President of Employer Customer Experiences
Health Data & Management Solutions (HDMS)

Jason Elliott is Vice President of Customer Experience for Employer clients at HDMS.  A true public health enthusiast with a Masters in Epidemiology, he spent over a decade delivering dedicated clinical analytics and leadership at BCBS.  Since then, Jason has managed the managed the Employer practice area.  He brings very structured thinking into the types of problems his clients are trying to solve, and what can be done with the insights discovered.

Wellness programs have become a staple of employer benefits offerings. According to one KFF trends report, nearly 9 in 10 employers with a workforce of 200 or more offered some sort of workplace wellness initiative in 2019.

While many employers are willing to invest in wellness programs—which often are offered through third-party vendors—they aren’t always clear on the goals for these benefits. Multiple surveys and studies across the industry attest to this. No clear goals mean no systematic approach to defining and consequently measuring ROIs.

Most employers rely on wellness vendors’ claims about the potential to improve health outcomes and reduce health care costs. They do not necessarily have the means and/ or expertise to independently verify the proposed advantages either prior to or after implementing them.

Metrics used by vendors to illustrate their successes are not always applicable to all populations or groups. For instance, let’s say a vendor’s “expected outcomes” include a 20% increase in smoking cessation rates. Is that 20% over three months or over three years since the last smoking incident? Or, is it based on a one-time pledge by the participant? What was the size of the overall smoker population in their sample data? Then there is always the question, “Is that the right metric for your population?”

Data analytics can provide objective insights to evaluate such partnerships before beginning, renewing or expanding a wellness program.

Like any strategic endeavor, effective ROI measurement requires diligent groundwork before actual data analysis can begin. It is essential to ensure that the right metrics are chosen for measurement of “before” and “after” states.

Concrete objectives will vary by employer as well as by program—but beware of setting goals focused solely on short-term “dollars-in” vs. “dollars-out.” An effective wellness program aimed at promoting better rates of preventive care with active engagement may actually increase expenses in the immediate and short term. In such cases, the true long-term objective should be to shift health care services from unpredictable, high-cost settings like the emergency department (ED) to more predictable, lower-cost settings like primary care physicians’ offices.

Today’s wellness programs tend to be more holistic in their approach to employee health than the offerings of just a few years ago. Many employers are looking for more than isolated reductions in smoking rates or ED visits. They are starting to understand the overall health and financial benefits to their businesses possible through programs that integrate physical health with mental health and well-being. This adds obvious layers and complexity to the ROI conversation.

Preparing for any ROI measurement requires assessing all the data sources at your disposal. It’s critical to obtain as close to a 360-degree view of the entire employee population as possible, which typically requires melding multiple disparate data sources. Medical and pharmacy claims, lab values, biometric and clinical data from electronic health records (EHRs) are some examples. Data warehousing and analytics solutions can help this process by aggregating, integrating, enriching and normalizing data along with consultative services to provide the right insights.

Finally, a realistic timeframe for measuring program outcomes is a must, especially when claims are part of equation, to allow for the time lag between services rendered and paid out. Hence 12 to 18 months serves as an optimal window to gauge discernible changes to patterns of care experience and member behavior. That said, periodic measurements throughout this time are essential for tweaking and adjusting workflows and processes to ensure proper data aggregation (e.g. presence of required code sets, uniform cadence in receipt of various data types, etc.).

Be aware of employee engagement factors

Achieving maximum ROI from wellness programs comes from changing behaviors, especially of those who are most at risk for adverse health events and consequently would benefit the most from these initiatives. For example, employees with chronic conditions who struggle with medication adherence or with managing stress due to work and family obligations. Promoting and maintaining engagement in such groups is challenging, but key to the success of the program itself.

On the same token, initial engagement tends to be high among members who are healthier and would likely gain little from a wellness or similar program, especially when there are participation incentives involved. Engagement typically tends to decline once the incentive requirements are met or phased out.

Setting up cohorts of participants with these factors in mind is critical because the metrics chosen to measure success levels in each will vary. Leveraging the expertise of data analytics vendors and consultants to define and set up such study cohorts—with and without comparable controls—goes a long way in these endeavors.

For example, employees who are engaged in wellness programs tend to also take advantage of preventive services and have a primary care provider. Consequently, data typically will show that they have higher rates of primary care and in-network utilization—whereas those who don’t participate have more ED and out-of-network services.

Establish key metrics

t is vital to ask the question, “Are we measuring the right things for each cohort for this particular initiative?” Defining the right metrics for a cohort is therefore an important aspect of the study design. Example: Establishing new primary care provider relationships and closing care gaps would be good metrics for employees who have traditionally not sought regular primary care in the past. On the other hand, keeping pertinent lab or biometric values within normal ranges, or garnering low scores on health risk assessment tools may be better suited for healthier and more engaged populations. Establishing clear baselines for each metric on day 0 is imperative for apples-to-apples comparisons.

Many employers are using non-traditional data sources to track metrics like sick time, other leave utilization, and rates of disability claims to evaluate the effectiveness of a wellness program. Data analytics and warehousing vendors offer tremendous advantages in this area by integrating disparate data sources.

Consider a pilot program

Pilot programs for a carefully chosen group with comparative control groups is always recommended, especially for new wellness initiatives. In addition to ironing out administrative and process challenges, they provide a great means of gauging the operational effort and resources required. This is an often-overlooked expense not featured in ROI calculations.

Results from a pilot program can go a long way toward determining an effective roll-out strategy. It’s essential to compare these results against the total employee population for the same timeframe. Example: An increase in the rates of flu vaccine compliance among employees in a pilot group does not mean much if vaccine compliance also increased in the total employee population due to onsite flu clinics. With successful pilots that show a definite improvement in outcomes for the participants, odds of further success are better when the program is expanded to demographically similar employees.

Let measurable results drive strategic investments

The last few months have brought renewed focus on the overall well-being of the workforce. Employers recognize the importance of the physical, mental and emotional wellness of their employees and their families. It’s not surprising that wellness program vendors, especially those that provide integrated services, are popular.

But rather than jumping on the wellness bandwagon or adding a program just to expand the suite of benefits, employers would be better served to make data-driven decisions. They would do well to engage the many data analytics vendors who provide evaluation services to answer key questions. “Is this right for our company?” and “Will this save me money on health care costs?” are the types of questions that can be answered even before the program is implemented, based on existing statistics or sample data sets.

Dr. Rani is Head of  Clinical Advisory Services at HDMS. She is a physician (specialty – General Medicine) with extensive experience in the EMR/EHR and population health industries with a focus on clinical transformation, workflow design and development, value-based care, risk management and clinical quality and performance reporting. Her strong background in clinical medicine and experience in the HIT industry make her successful in navigating payer, provider and technology vendor landscapes.