Powering Unbiased Outcomes Research with Longitudinal Data

In preparation for a new launch, a medical device company needed to conduct a health economics and outcomes research (HEOR) study on cemented and uncemented knee arthroplasties. The company wanted to know if readmission and infection rates differed for arthroplasty patients depending on the fixation method used. By conducting this research, the company hoped to generate outcomes data to support the clinical efficacy of its new medical device.

Improved HEOR studies with integrated longitudinal data
Reduced bias with linked claims and EHR data
Generated stronger real-world evidence for product launches

Key Results

Reduced Bias in Real-World Outcomes Research

Challenge

When the company used claims data to investigate feasibility counts, they were surprised to see a steady downtrend in the number of knee arthroplasties explicitly specified as either cemented or uncemented. As the number of knee surgeries in general remained constant, the company realized it needed a deeper data analysis to resolve this discrepancy.

The company partnered with Norstella to generate robust data for its HEOR study. Norstella’s data analysts explained that the apparent downtrend in cemented/uncemented arthroplasties was due to the overall trend of arthroplasty procedures shifting to the outpatient setting. While inpatient medical coding uses ICD-10-PCS codes, which specify the fixation method, outpatient physicians use more generic CPT codes, which do not.

Solution

The Norstella team recommended linking its more granular, unstructured EHR data, powered by NorstellaLinQ, to its closed claims data for a more comprehensive and accurate analysis. As an HEOR study that relied solely on claims data would capture only inpatient procedures, which may have an associated higher risk of adverse events, this approach would mitigate any bias in the data.

The data analysts used curation to identify surgery dates and arthroplasty types within Norstella’s EHR dataset. By extracting this data into structured fields, the analysts created a starting cohort of cemented vs. uncemented arthroplasty procedures. From there, the deidentified EHR data was tokenized and linked to the deidentified claims data. Norstella also performed the statistical process of expert determination, which ensures that any identifying values are obfuscated to prevent patient reidentification after the data is merged.

Outcome

Norstella’s EHR data was then linked to the claims data in the Instant Health Data (IHD) platform. As IHD tracks continuous enrollment, the claim data provides a history of every medical encounter that a patient has within a certain time period. This longitudinal data is especially important for an outcomes study, as it shows both the short-term and long-term impact of a specific procedure on a patient’s overall health.

Instead of sourcing disparate datasets and waiting months for a third-party vendor to harmonize them, at considerable expense, the company was able to leverage Norstella’s fully integrated real-world dataset to achieve fast results. Within eight weeks, the client had access to years of longitudinal arthroplasty data that represented both inpatient and outpatient procedures. The company was able to complete an impressive HEOR study to support a successful device launch.

Get Started

Get Closer to the Patient Than Ever Before

NorstellaLinQ links claims, EHR, lab, and electronic prescription data into a single patient view — decision-grade and AI-ready. See what it can do for your team.

Frequently asked questions

Everything You Need to Know

How does longitudinal data improve health economics and outcomes research (HEOR)?

Longitudinal data tracks patients over time, allowing researchers to evaluate both the short- and long-term outcomes of treatments and procedures. This provides a more complete view of patient health and supports more robust HEOR analyses.

Why is linking claims and EHR data important for real-world evidence?

Claims data alone may miss important clinical details or outpatient procedures. Linking claims with structured and unstructured EHR data creates a more comprehensive dataset, reducing bias and generating stronger real-world evidence.

How can integrated real-world data reduce bias in outcomes research?

Integrated longitudinal datasets capture patient care across multiple healthcare settings, including inpatient and outpatient encounters. This broader view helps eliminate gaps that can occur when relying on a single data source, leading to more accurate and unbiased research findings.

What challenge did Norstella help solve in this case study?

A medical device company needed to compare outcomes between cemented and uncemented knee arthroplasties, but claims data alone underrepresented outpatient procedures. Norstella integrated EHR and claims data to create a complete longitudinal dataset that enabled an unbiased HEOR study.

What business outcomes did the project deliver?

Using Norstella’s integrated longitudinal dataset, the company completed its HEOR study within eight weeks, generated stronger real-world evidence, reduced bias in its analysis, and supported a successful product launch.