Measure your identity data management maturity. Take our 5-minute maturity assessment now

The Top Three Problems with Tuning Your EMPI

Thought Leadership

Ever try to tune into your favorite radio station when reception wasn’t the best? Trying to home in on the right frequency is futile. Verato often encounters teams that try to fine-tune the data on their legacy enterprise master patient indexes (EMPIs) only to struggle and make things worse. Since the EMPI is the backbone of so much essential data within a healthcare organization, the data tuning process often consumes resources and may degrade matching performance, as I discussed in my previous blog.

Let’s look at where tuning goes wrong and what healthcare organizations can do instead. Verato will often complete an analysis of the data quality of our customer’s existing EMPI. When our prospective customers have legacy systems in place, we discover the following 3 common issues:

1. Introduction of overmatches – Verato regularly completes a pre-deployment analysis, called Verato Diagnose, of a customer’s master person index data. Among large healthcare systems, we routinely identify a false positive rate of over 200 per 1 million patients. This exceeds the common benchmark of 1 in 1 million or even more conservatively 1 in 10 million duplicates, a standard set to uphold patient safety.

When a health system looks to improve their EMPI accuracy they largely start with trying to address missed patient record matches or false negatives. Their IT teams often scope tuning/re-configuration projects from their vendor or consultants, to address the missed matches. Within scope of a project like this, the technology consultants tune and tweak algorithms and weights to address the specific missed patient record matches. While the process resolves previously observed missed matches, the unintended consequence is the introduction of overmatches. Ultimately, this is a temporary fix because as data continues to change, the root of the problem is never really addressed. While the legacy EMPI produces fewer false negatives, the rate of overmatches/false positives remains unacceptable or becomes worse.

2. Massive numbers of tasks and potential overmatch issues – When we conduct pre-deployment analysis, we often see hundreds of thousands of unresolved tasks and/or patient record stewardship issues. These are situations where HIM or other data stewards are asked to manually investigate and decide on the patient identity.

Both situations leave duplicates and disconnected data in their wake. We learned recently about a health system that was still left with over 200,000 tasks after completing a consultant led project that tuned their algorithm. The initial thought was to automate more matches with select tuning. What resulted was more ambiguous matching scenarios, more work for HIM!

The opposite of that result happens when consultants are overly aggressive in their tuning. The project can introduce thousands of false positives in which a patient is mistaken for another patient, increasing the risk of the problem while making no effect on the stewardship workload.

Taking a deeper look at the data situation, this project went astray when the professional services teams extracted tasks at the upper bound of the automatic linking threshold and looked for a new tolerance level. The result, unfortunately, was that task volumes were reduced by lowering the required score to make an automated match. This comes at a price—the cost of overmatching, leaving the bulk of the problem unaddressed with an even more dangerous outcome.

3. Difficulties adding new data sources – Health systems grow and diversify, prioritizing efforts like patient engagement and care coordination. With this addition of data comes the need to tune and reconfigure a legacy EMPI. Retuning is necessary to deal with the data characteristics and nuances of new incoming data sources. As a rule, anytime the data volume changes by 25%, a solid look at your algorithm and process becomes necessary.

Undoubtedly, organizations look to hire consultants to tune and verify that their bespoke algorithm will work for the new source. This is a time consuming and costly effort that requires expert knowledge of the legacy EMPI algorithm and the new source of data. If ignored, matching performance degrades for the entire system. For example, if you have 1M records and add a source with 250K records, re-tuning must occur. Legacy EMPI artisans must adjust the weights and thresholds so that matching can continue to be reliable–avoiding downstream negative impacts. This is not a simple task and is not without risk.

These tuning projects are time-consuming and expensive. What’s more, they take resources away from strategically important initiatives, such as implementing a data warehouse or efficiently transitioning through a post-merger consolidation.

This is in contrast with the experience that healthcare organizations have with Verato.

No more extensive tuning

With Verato, it’s not necessary to continually tune and re-tune the Universal Master Patient Index (UMPI). Verato Referential Matching(SM) is a proprietary technology that compares patient data against our comprehensive, continuously updated national reference database. We achieve high accuracy patient matching rates and support our customers strategic data initiatives, including data warehouses and sophisticated analytics. We have consistently proven our ability to drive higher accuracy rates than bespoke legacy EMPI algorithms. In as little as two weeks, we conduct analysis and show our customers the degree to which their MPI overmatches or undermatches.

Deploy quickly

Purpose-built for the cloud, Verato UMPI uses modern APIs and is straightforward to implement and requires fewer resources to maintain. These characteristics allow customers to deploy in weeks and onboard new data sources in as little as a week. With this greater flexibility, healthcare organizations can build an enterprise-spanning data warehouse and add new data sources as needed. Our customers are no longer limited in expanding their EMPI footprint by scheduling and coordinating costly algorithm work. Verato customers can simply engage Verato and have new contributing and consuming systems available rapidly.

Learn more about a customer’s experience with Verato’s efficient implementation, cost-effective maintenance, and support of an enterprise-wide data warehouse strategy. Join the leading healthcare organizations who are making the leap to Verato.  This demo shows how replacing your legacy EMPI can save your organization significant costs and advance your strategic goals related to data and analytics.