The Best Way to Make Payer and Provider Healthcare Data Accessible
Talk to pretty much anyone on either the payer or provider side of healthcare, and it’s likely you’ll get agreement on the importance of getting the spiraling cost of healthcare under control. They’ll also agree that the cure is to focus on quality – determining the current baseline for care, setting goals for improvement, and measuring against those goals. Unfortunately, that’s about where the agreement will end, because when it comes to the specifics of what to measure, the answers are all over the place.
There are, of course, some measurements that are standardized – things providers must measure in order to get paid. Centers for Medicare and Medicaid Services (CMS) has published a robust set of measures for different care settings, including hospital, home health, and more. Since generally Medicare accounts for approximately 30 percent of most providers’ net revenue, it’s important those measurements are reported accurately and completely.
The National Committee for Quality Assurance (NCQA) has also developed the Healthcare Effectiveness Data and Information Set (HEDIS), a set of 75 measures across eight domains of care that are used by 90 percent of America’s health plans to gauge effectiveness across a population. The Joint Commission has a full range of measures, including for hospital settings, and organizations focused on different diseases (such as cancer) have them as well. If that’s all there was for hospitals to deal with it would be relatively simple.
Payer and Provider Quality Measurements: Complications Abound
It gets complicated, however, when you get past these broad external reporting initiatives. For example, each payer has its own set of measurements it requires from providers. While they may look at the same categories of care – for example, they may all have measures for heart failure – the specific parameters for measurement are often different. The result is hospitals have to be able to extract data to meet each of these sometimes-conflicting individual measures.
Why the divergence? Simply put, each of the payers has priorities regarding what drives quality healthcare for their members. If the provider typically works with eight different payers, the result is a need to extract, measure, and report the data eight different ways.
There have been some attempts to standardize measurements among payers, most notably in California, where as part of a pay-for-performance group, several payers worked together to define measurements among themselves. While they may weigh the measurements differently for their own internal purposes, providers are able to report the same data to those payers. So far, though, this kind of coordination among payers is the exception rather than the rule.
Further complicating matters are the provider’s own quality initiatives. For example, in addition to the measurements being provided to CMS and their payers, a hospital may want to determine its own effectiveness and efficiency in delivering services to a population of heart patients. To do so, it develops a completely different set of measurements focused on the information it wants to capture.
The goal of all these efforts is to improve the quality of care and safety for individual patients and across a population. Yet, the result is a bewildering set of measurements – some required by regulations, some by individual payers, and some by the providers themselves – that make reporting performance complex and time-consuming under the best of circumstances.
Difficulties Obtaining the Healthcare Data for both Payers and Providers
Another challenge to delivering timely reporting is that the data sources for those reports come from a bewildering number of sources
Some of the data comes very easily from claims information. Some must be laboriously abstracted from medical records or retrieved from an electronic health records (EHRs). Some is contained in specific software, such as cancer tumor registries or orthopedic implant registries.
Obtaining that data is often a time-intensive process that requires quality assurance nurses and utilization review staff at hospitals to collect the information manually from all these sources, enter it into Excel spreadsheets, and then send those spreadsheets to internal and external constituents as required. In many cases, nurses are being asked to abstract a specific piece of data from a medical record, which is often in the notes the physician wrote rather than a specific field, making it difficult to locate. If the handwriting is hard to read, it becomes even tougher to find–let alone transcribe accurately.
Yet, this work must be done to comply with the various external reporting requirements or if there are going to be improvements in quality of the care processes. While the focus will likely be on measuring the performance of the entire hospital, the ability to drill down further to analyze variation or discover the root cause is essential. For example, sorting the information by individual physician and discovering who is doing things well (and why), enables design and institution of process improvements. That doesn’t happen without quality data. So while it may be painful, the process is critical to improving patient care.
Making the Healthcare Data Accessible with a Healthcare Enterprise Data Warehouse
With so much data required in so many different forms by so many different constituents, and with the degree of difficulty often required to obtain that data, having it all readily available in a healthcare enterprise data warehouse (EDW) will be a huge advantage. Tools available through the EDW to access and analyze the data to create dashboards and reports and automating the process wherever possible will get answers faster.
This is where a Late-Binding™ Data Warehouse makes sense. When a payer or provider is starting a quality improvement initiative, it’s unlikely they will know all the data elements that will be relevant. The provider may ask “What is every piece of data needed to solve a given problem?” and give their best guess, but, until the team really gets into the process, they will be working off assumptions. With a traditional, early-binding data warehouse approach, the provider could end up including a lot of information that isn’t important, and/or miss some that is. If something is missed, it can be difficult to add it to this rigid data model after the fact.
With a Late-Binding™ Data Warehouse, however, the provider can start small with those elements they are absolutely sure they’ll need, and map just that data into the data warehouse. Down the road, if they find some other piece of data that needs to be added to the data warehouse they can do it easily. It provides far greater flexibility, with the ability to adapt as new requirements arise.
Making Payer and Provider Quality Measurements Work
In order to deliver quality, a payer or provider needs define what quality is and then measure it. That can mean using standardized, agreed-upon, regulatory parameters, payer-specific metrics, or even programs developed internally.
To make quality initiatives work, more time should be spent analyzing data and distributing the information to the key stakeholders than collecting it. A Late-Binding™ Data Warehouse gives a central collection point for the data, puts the data in the hands of the decision-makers, and gives them the tools to analyze it themselves rather than waiting weeks for IT to run a report. Of course, those users also have to understand how to work with data in order to gain the insights needed to effect real change within the organization.
With a Late-Binding™ Data Warehouse, all those pieces can be put into place. Then, it’s time to see how using data can drive successful quality improvement efforts and lead to substantial cost reductions. Using the Healthcare Analytics Adopting Model as a guide, payers and providers can take the next steps to fully understand and leverage the capabilities of analytics.
What complications do you see for payers and providers agreeing on quality measures? What is your experience with trying to combine the data?