Having good data analytics in health programs is not only just about measuring performance, it can also affect program experience for both employers and their employees. For any program to be successful, collecting and analyzing data is of the utmost importance. Integrated health programs lead traditional programs when it comes to data collection, analysis, and action.
In this episode, Matt Doctrow, SentryHealth SVP of Analytics, discusses how analytics plays a key role in health program development and engagement. When companies invest in data-driven employee heatlh solutions, both the employee and employer see positive outcomes.
Measuring return from employee health programs is tricky. While measuring return on investment, or ROI, can be difficult, it doesn’t mean that programs aren’t important. Measuring value on investment, or VOI, can be just as important. Read this blog post to learn more about VOI, including how it’s different from ROI and why it’s important to measure.
In This Podcast
With more than 30 years of analytics experience, Matt is a data expert passionate about applying analytical results to support evidence-based decision making and performance improvement. He’s responsible for applying artificial intelligence and machine learning models to personalize participant journeys for enhanced experiences and clinical outcomes.
Hi everyone, thanks for joining this episode of For Your Benefits, I’m Meghan Henry, Director of Marketing for SentryHealth and the creators of WellOnMyWay, a turnkey employee health and wellbeing solution.
We recently hosted a webinar with Matt Doctrow, SentryHealth’s SVP of Analytics, where he talked about how digital health and wellbeing programs are at an advantage over traditional programs when it comes to data collection, analysis, and action.
He also discussed how using analytics in health and wellbeing programs can create positive outcomes, ROI, and VOI. We thought this information was so valuable that we wanted to make it available to you as a podcast episode. So here it is.
What we’re going to do today is discuss how we at SentryHealth use analytics in health and wellbeing, so we’ll start with identification. The goal is to determine the individual’s risks, needs, and goals. We primarily reduce inefficiencies by being proactive instead of trying to tackle the entire employee population.
Let’s start with the employees and the family members that have the largest potential for health improvement and cost savings. If you look at the top 5% of individuals or family members, they’re typically responsible for about 50% of the claims. So, for example, you have a company with a thousand individuals on their health plan, the top five percent would represent 50 people. Those 50 people out of a thousand would make up about half of the medical and pharmacy claims. We want to focus on those with chronic conditions because typically they have many comorbidities.
If they have diabetes, they typically have hypertension, hyperlipidemia, possibly depression, obesity, etcetera. So, it’s an important part to manage chronic conditions, both those that have been diagnosed and those that we are predicting with our risk models to develop those chronic conditions.
We want to focus on brand-name drugs and find where generics are available, but they’re not being used, those are easy savings. And then specialty drugs, there’s a lot of programs to help reduce the cost of these very high-cost specialty drugs. But the employees are not going to find those, we as an organization have to do that for them, so we will help you do that.
Preventive care, we want to know, are the employees and their family members visiting the dentist, having their preventive care for mammograms, paps, vision tests, etc? So, the system tells us which members have not been getting those preventative tests and exams. The risks and interests expressed in what we call the health risk assessment. It’s a self-assessment that each member will take where they self-identify what they believe are concerns. So, if they answer those questions to lead us to determine that they have a risk financially or they’re at risk through their mental health, say through depression, we’ll pick up on that through the data and target our activity towards those folks.
Frequent flyers — those are people that are at high costs year after year. So, if you see somebody have high cost in one year and then the next year it goes down, you know, maybe it was just a one-time incident. But if you’re seeing people with year over year very high claims, that’s somebody that is a different story that we want to dig in and see how we can help lower the claims costs, improve their health. Some people think if you have zero claims, that’s a good thing. To me, it’s not. To me, it means those members, employees, are not getting their preventative exams and tests.
So, if you’re not going to get your teeth cleaned twice a year, for example, you’re likely to have teeth problems, dental problems, down the road. So, we want to point those out and be able to determine or direct certain campaigns towards those people. Will also communicate on general health and wellbeing topics to individuals that don’t fall into any of these categories so that everyone in the population is receiving communications throughout our engagement.
The second agenda item was the targeting of the members. And the goal is to provide individually focused guidance and recommendations for action. So, we must personalize the communications in order to maximize member engagement.
Personalization & Preferences
Preferences. we need to learn each individual’s preferences to make it as easy as possible for continual communication, how each person wants to be engaged. For some people, that’s email, for some people that’s phone, or for some people, that’s text or messaging. Every person is different; we want to determine that, record that, and have the system help us remember the best way to communicate with each member.
We want to leverage the data to personalize the communications and recommendations for action. So, if a member is diabetic, then we would want to personalize messages about diabetes to that particular member, but to other members, they’re not going to get that message because it’s not pertinent. And the information in that case, on diabetes, within the system will automatically send educational materials on diet and exercise, for example.
We want to track their platform activity, track their user clicks, and our system so we can learn what they are interested in and we can target our messaging accordingly. On the other hand, we can determine who’s not using a platform program but needs to be using it. So, for example, if you have someone who’s a diabetic and they’re not using our specialty partner program called OneDrop, that is for diabetes and hypertension, we need to know that so we are able to communicate outreach and try and convince them of the positives of using that particular program.
And then, when necessary, a physical connection by a human. We have coaches and clinicians that will add to that digital targeting to bring in that next level when it’s actually needed.
Next, we’ll talk about engagement. How do we engage individual members? We can identify them now through the data, we can target them, but if they’re not responding, then they’re not getting anything out of this platform. So the goal is to raise health literacy and healthy behaviors through individually focused engagements. We’ve identified individuals again. We’ve targeted them for communications. Now we have to engage them to raise their health literacy and to motivate meaningful action. We can do this by using machine language and artificial intelligence.
We can engage individuals digitally using data to determine the subject, the method, and the frequency of personalized engagements. Health coaches are available as needed by individuals, but we need to engage health coaches for the highest acuity patients or the most complex based on advanced analytics. The analytics in health care can drive the employer engagement from the employer side, for example, the employer will be able to see in aggregate that there’s one hundred and fifty diabetic employees, but they’re not going to be able to identify those employees, we don’t share that information with them. But if they know that diabetes or hypertension are very high within their employee population, then they can do things on their side.
For example, schedule brown bag lunches, have lunch and learns with a speaker talking to them about a certain health topic. So that’s one reason that we share that information with the employer.
Another example might be if we can find prescribers, physicians, that prescribe brand names, but not generic drugs. Employer would want to know that, so the next time they put their network together, they can exclude that physician, if at all possible.
Then once we’re engaging the individuals, we need to reward those individuals, encourage meaningful action, and reward success. So we motivate action and then we celebrate success with rewards. We’ll work with each individual employer to determine the rewards that will work best with their employees. The recognition is relational. Showing appreciation is motivating. If you look at HR studies, it’s very obvious that one of the things that most employees want is appreciation. They want to be appreciated for working hard. That comes up time and time again in surveys. This is one way that we can get them to engage is by rewarding that particular behavior.
Foster healthy competition. We can, we as a company, SentryHealth, can create competitions, for example, obese members or those with diabetes or whatever it is, to encourage action. So through those challenges were trying to increase their engagement and make them healthier, at the same time. The types of rewards we’re talking about are premium discounts, paid time off, in some cases, gift cards, and apparel. It’s really up to the employer as to what rewards they want to give and we will help keep track of those particular members.
And then the last stage is measuring. We need to know if we’re improving. We need to know if the cost of the health care claims are going down year to year. And the claims we refer to, is the return on the investment. If you look at the claims savings compared to what the employer’s spending on the program, VOI is more about the value of investment and we’ll talk about that in just a moment. So the purpose is to make sure, again, that we’re achieving our desired outcomes.
Data is used again to measure the success and rewards must be timely. So by doing this digitally, we’ll be able to identify, create challenges and reward those that respond in a timely manner. Success, though, is not only quantitative, as I mentioned before. Improving the quality of life is crucial to attaining company goals. So, yes, it’s about saving the company money through lowering their claims, but it also leads to healthier employees, which means those employees are less likely to be absent and are more likely to be productive and are less likely to leave the company. So what we want to do is use those historical HR metrics from the years prior to them, for those that are using our platform, and compare that to the first year and subsequent years on our platform.
Return on Investment & Value on Investment
We believe that there’s a high correlation between the members that use our platform and are engaged, tend to stick around and be more productive with less absenteeism, for example. So the return on investment, again, that’s easy to calculate using claim dollars.
The value on investment is a little more difficult, but we should include those types of HR metrics when looking at the value. One thing I’ll mention is just to ensure that your formula, for your metrics, are consistent from year to year. Because if I say, how do you calculate turnover, there’s, you know, five companies that can do it five different ways. There’s probably one that’s used predominantly, but you can’t take that for granted, so you’re going to make sure that their formulas are consistent from year to year, so we can lower the claims again. But the icing on the cake is that we’re going to help them improve their HR metrics and have a more healthy, productive staff.
Analytics in Health and Wellbeing
Digital health and wellness programs have an advantage over traditional programs when it comes to data collection, analysis and action. Through the power of information, each employee’s health and wellness journey can be customized to drive maximum engagement. Tracking who is participating, how much they’re participating, and when they’re participating, can help identify program elements that drive the greatest engagement, as well as those that need to be tweaked.
So to summarize, this is what our platform WellOnMyWay brings to the table for wellness programs.
Number one, we can determine which individuals will benefit from specific recommendations and actions or determine which benefits are most likely to benefit from specific recommendations and actions. So that is the key to what we’re talking about, determining who and figuring out how to communicate with them, and then having those messages spread out so to not overwhelm the employee. But let’s start with the areas where we can make the biggest difference first. We’ve created predictive models for 22 chronic conditions, so for those people that have not been diagnosed with a particular condition, what is their likelihood of developing that condition? So if we know that and we know that these 50 people are at high risk for developing diabetes, it’s for us to contact them and help them understand what they need to do in order to not move across that line.
We have three levels of dashboards for reporting. The first is a client-facing dashboard, and as I mentioned, that’s the summary of the population health, there’s no individual identifying data. And to go one step further, if the number of employees falls below the number 10, we won’t even show that data because in some cases, if there was one member showing on the dashboard, you know, they could potentially figure out who that one person is. And we don’t want that to happen. So any time, where there’s less than 10 members will just show up as blank and not show that particular number.
The participant-level facing dashboards are for us as a company. Let’s say that we want to target the top five percent, we could, with one click and identify who, by name, those top five percent members are, that are not getting their dental exams or the people that have diabetes. We then can target them very easily and pull up the list by going down to the participant level. Again that is for us as a company to help drive healthier populations.
SentryHealth also has a summary dashboard that looks at all of our clients’ data. So, for example, if we know we have over 20 companies, we have one hundred and fifty thousand lives represented in our database and we know the average for blue-collar jobs and the average spend per employee is “X” for white-collar jobs it’s “Y” and for small companies, it’s “A” and for big companies, it’s “B”. So we want to be able to compare each company with benchmarks, and those benchmarks will be as comparing apples to apples as much as we can, in this particular case.
We also want to measure success through key performance indicators. We will be doing that on our end as well, we believe. And the data is showing that there’s a high correlation between program engagement and healthy outcomes. We want to find out who the employees are that are using this platform, who are the family members that are using this platform, and we’ll be able to compare their outcomes healthy versus not to those that don’t.
ROI and VOI
And then lastly, we need to calculate at the end of the year the return on investment or the value on an investment in order to show that our program is making a difference in the way of what’s in the future. We will continue to advance machine language and artificial intelligence within our platform. There’s no limit on where that can go, but right now we’re focused on other things. But that’s going to be doing a lot of the things that typically we would have to do manually. The computer is going to be doing it and doing it better than a person can do it. We’re going to incorporate new partner program information into our predictive models.
So if we have a program on diabetes or we have a program specifically for hypertension, we want to bring that information that we can pull from that partner program into our models. Right now, it’s based on our medical claims, our pharmacy claims and the health risk assessment.
And then lastly, we want to add individual metrics into the predictive models, because I believe if we can pull those in and say and see over the last two years, how many sick days, who left the job, how many employees had less than two sick days, I believe those are going to be predictive in certain ways for particular outcomes as well.
Well, that wraps it up for this episode of For Your Benefits. If you’d like to learn more, please visit our website at www.sentryhealth.com. And if you like what you heard today and you want to hear more, don’t forget to subscribe to our podcast. We’ll continue to keep you on the cutting edge of what’s happening in the world of health and wellbeing. Thanks for joining us.
Want to learn more about SentryHealth? Simply fill out this form to request a consultation and learn more about our solutions.