Merative Blog | Technology, Data, and Analytics in Healthcare

Optimize health benefits with trustworthy data

Written by Ronda Daugherty | Jul 16, 2025 7:19:01 PM

It’s hard to make confident decisions based on limited or untrustworthy information. That’s especially true in health and benefits, where the stakes are high and benefits professionals are under pressure to keep program costs manageable while maintaining attractive benefits. To design and manage great programs, you first need data you can trust. 

Through their insurance claims, employers have access to a lot of healthcare data – clinical data, admissions, outpatient claims, pharmacy data, lab data, and much more. On top of healthcare data, employees may also have access to data related to employee productivity, leave and absence, disability, level of engagement with health and wellness vendors, potentially even socioeconomic data.   

In a perfect world, it would be easy to bring all that data together to provide a complete and accurate picture of healthcare program performance, enabling better decisions around employee care, program offerings, and overall cost. But as anyone who’s tried grappling with these issues has likely discovered, it’s anything but easy. 

An L.E.K. Consulting survey, commissioned by Truven, found that a lack of quality data and expertise holds employers back from understanding outcomes related to their health benefit programs. Specifically, self-insured employers have:  

  • A lack of high-quality, reliable data on outcomes 
  • A lack of data analytics talent within the company 
  • The inability to benchmark their company’s health benefits program metrics against similar organizations 

With that in mind, how can health and benefits ensure the data they base their decisions on is of the best possible quality? 

Bad data leads to bad decisions 

Without quality healthcare data, organizations can’t ensure that their decisions are going to be the best possible for their employees and their own bottom line. It’s important to know that the data has the following characteristics:  

  • Accuracy – Are the data points correct and compare well to historical data or benchmarks? Does the data match external validation sources? 
  • Completeness – Is all the data present? Pay special attention to key fields and the percentage populated. Do files match control totals? 
  • Consistency – Does the data match across all systems? Are fluctuations across time periods consistent, expected, and within set thresholds? 
  • Timeliness – Is the data up to date? Is it received from the data supplier in the timeframe expected? What is the lag? How much historical data is available? 
  • Uniqueness – Are there any duplicate pieces of data? Is data available from non-traditional data sources? Can missing data, such as for social determinants, be inferred from external data sources? 
  • Validity – Is the data formatted correctly? Can different data files be linked? Are unexpected values received? 

At Truven, we recommend a four-step measurement strategy to make quality data actionable.   

  • Data collection - What data is needed to manage your health and benefits program? What data can monitor success and surface opportunities? Essentially, determine what is needed for an organization to best tell its story. 
  • Data transformation – It’s possible to transform data into one source of truth. Organizations can integrate disparate data sources, standardize with industry guidelines, customize the data to their unique needs, validate data expectations, and protect and secure their data. 
  • Building intelligence – Apply methods and algorithms to make the data smarter. This can include dynamic groupings, clinical classifications, treatment regimens, population categories, and guidelines and benchmarks. 
  • Insights and application – By delivering actionable analytics and developing best-in-class benefits, organizations can better set and monitor KPIs, find saving opportunities, measure ROI and VOI, and identify patterns and predictions. 

Once your measurement strategy is set, there is a cycle of iteration and improvement where you can start collecting new data and integrating it into the system. 

Applying data insights 

So, how can organizations take all that data and put it to good use? Let’s look at three examples. 

Validate program decisions – New York State covers health benefits for over 1 million state and local government employees, retirees, and their dependents. It made some program changes as part of collective bargaining and wanted to measure the impact of those changes before and after they were implemented. Truven’s analysis revealed that the plan changes were successfully able to steer members toward high-quality in-network options, driving steep drops in costly out-of-network services. This kind of third-party analysis can be essential to validate plan decisions, but it's only possible with quality data.  

Put trends into context – The Kentucky Department of Employee Insurance covers 260,000 individuals and has approximately $2 billion in annual claims. With rising prescription and medical spend, it wanted to evaluate the data to see where member needs were most pressing. Truven was able to help determine why there were high rates on certain types of healthcare spend and not others, while also putting rising costs in context – why now? What are some possible long-term impacts?  

The Department was able to study cost impact versus cost avoidance, and was able to evaluate ROI in an ongoing, longitudinal format, rather than just a single point-in-time.  

Identify member needs – The city of Seattle, Washington is comprised of 13,000 employees and 25,000 total members. The city is able to use its own claims data to better identify member needs and evaluate the right solutions. Seattle can utilize long-term forecasting of medical trends to help find potential drivers of higher costs. For example, Seattle brought on virtual physical therapy options to help combat high reports of lower back pain and concurrent heavy use of opiates for that pain. Three years after the virtual option, significant pain reduction has been reported, the average likelihood of surgery has fallen, and millions of dollars have been saved.  

Ready to see what quality data can do for your organization? Check out our Health Insights demo to learn more.