The Role of Health Terminologies in Applied Population Health

Applied Population Health

Applied Population Health is defined as, “the use of the EHR system to identify patients in need of evidence-based interventions and facilitate action to address care gaps”. [1]  Integration with the EHR is a critical element of this concept because it is through this application that patient clinical data is managed and care decisions are recorded. The EHR is the primary vehicle for delivering just-in-time, context sensitive clinical decision support that is foundational to value-based care. Here we explore the role of health terminologies within an applied population health program for the use of data integration, practitioner support and quality measurement.

Applied Population Health is an integrated approach to delivering value-based care.

Barbara Berkovich, PhD, MA Clinical Associate Professor, University of San Diego Tweet

Data Integration

The role of health terminologies in this context is to integrate clinical information from disparate data sources to create a complete and consistent record of patient health status and health care interventions. At minimum, an EHR system collects patient care records for a single health system. However, to compile a complete and accurate record of patient care, organizations must look beyond their own EHRs to capture care documented in other health systems, legacy medical records, external laboratory results, and Health Information Exchanges (HIEs). In Applied Population Health, external data needs to be “translated” into a common meaningful context using terminology management. The degree of success in data integration impacts how well computers can support the clinical practitioner and report quality measures.

Practitioner Support

As clinical decision support (CDS) continues to play an ever-larger role in the complex care setting, access to external data in human readable form is not sufficient. The data must be computationally accessible which is another way of saying interoperable.  Again, the translation metaphor is useful in understanding the need to convert text, image formats, standard and non-standard clinical codes into a common “language” that the algorithms can process. Clinicians can be heavily penalized for making a single error in treatment if it results in a malpractice claim, and therefore may lose trust in clinical decision support algorithms if they detect even narrow margins of error. 

Quality Measurement

Health quality measures are commonly computed as the ratio of patients who did receive an intervention (numerator) and patients who should receive and intervention (denominator).  The figure illustrates that the diabetes A1C quality measure utilizes over 250 ICD-10 diagnosis codes and possibly 10 or more local lab codes. Use of a single LOINC could conceivably simplify the lab coding, but many organizations have not fully converted to LOINC, having no mandate to do so. LOINC has grown to over 86,000 codes covering the full scope of laboratory testing and a broad range of clinical measurements. [2] LOINC mappings i.e., the relationships between a specific result type and the corresponding code, demand absolute precision and accuracy and are difficult to maintain, and the simple omission of one lab or procedure code in a quality measure calculation can contribute to under-reporting quality performance.

Diabetes A1C Lab Monitoring Quality Measures

Bottom Line

Applied Population Health is an integrated approach to delivering value-based care. Ultimately, the goal of that approach is high reliability health care in which nearly all of the eligible patient population receives the recommended care within set time parameters. That requires a robust feedback loop in the form of real-time quality measure calculations to support process control  for cancer screening, chronic care management, and other innovative measures. Achieving that goal requires rigorous terminology management for data integration, practitioner support and real-time quality measurement.

Terminology errors can have a negative impact on financial reimbursement for care, operational efficiency and patient outcomes. From the patient’s perspective, terminology errors increase the likelihood of an unnecessary clinical intervention or a missed opportunity to identify a care gap.

For these reasons, terminology management is a big deal in Applied Population Health programs. Yet there is an insufficient supply of true experts to meet the collective needs. Now is the time to invest in terminology support services.

Barbara Berkovich, PhD, MA

Clinical Associate Professor, University of San Diego

https://www.linkedin.com/in/barbaraberkovich

References

  1. Berkovich B, S.A., Applied Population Health: Delivering Value-Based Care with Actionable Registries. HIMSS Book Series. 2020, Boca Raton: CRC press. https://www.taylorfrancis.com/books/edit/10.4324/9780367196714/applied-population-health-barbara-berkovich-amy-sitapati 
  2. Bodenreider, O., R. Cornet, and D.J. Vreeman, Recent developments in clinical terminologies—SNOMED CT, LOINC, and RxNorm. Yearbook of medical informatics, 2018. 27(1): p. 129.

Publication date: June 1th 2021

Share on facebook
Facebook
Share on twitter
Twitter
Share on linkedin
LinkedIn

Leave a Comment

Your email address will not be published. Required fields are marked *