All-Payer Claims 837 Data and Longitudinal Analysis

All-Payer Claims 837 Data and Longitudinal Analysis

Making this Data Accessible and Usable for Informed Growth Planning

By Brian Dailey, Morgan Atkins, and Dave Sellers

Over the years as we’ve worked with all-payer claims data (APCD), and built tools on top of this dataset, we’ve often encountered questions about it— what it includes and how it can be used. In this post, we’ll address many of those frequently asked questions. We’ll define APCD and describe the longitudinal analysis that can be done using this data.

What is APCD?

We can’t begin a discussion about longitudinal data analysis without first unpacking APCD. Specifically, 837 APCD, which is also referred to as “submit” data. 

APCD is anonymized patient-level data. It consists of electronic files that contain information about patient visits with a provider. Each patient visit results in one or more claims being submitted by the provider to the payer for reimbursement. (Self-pay or charity care situations are not included.) These files are handed over to a clearinghouse or insurance company so that the provider can be reimbursed for the service they provided to the patient. 

The information included with these claims can include the following:

  • Patient information
  • The condition that the patient was treated for
  • The services that were provided
  • Details about the site of service
  • The physician rendering services

iStock-1197865054

Stratasan’s APCD includes patient claims that are sent to commercial insurers, Medicare fee-for-service, Medicare Advantage, Managed Medicaid, and traditional Medicaid. It provides insight into all sites of care where a patient may seek treatment—from hospitals and ASCs to physician offices and urgent care centers. 

That covers a lot of data—over 1.5 billion claims per year, in fact. 

With claim-level detail and nationwide coverage, the 837 APCD delivered to us from a data aggregator can be inconsistent—it has a lot of variability and “noise.” Each of the clearinghouses our aggregator works with has a unique set of protocols for how they sort and organize their data. This is why, in its raw form, this data isn’t very useful. It must be cleaned and curated before it can be used to draw conclusions. 

Before this data can be loaded into any of our tools for analysis, it’s curated from the claim-level through our proprietary data processing service. A seemingly quiet, behind-the-scenes operation, our data processing experts deliver outstanding results. 

This team will verify trends, connect codes, validate formatting, and QA the data. Additional proprietary cleansing is done, such as facility inference (when no facility info is provided), care setting cleansing (when we think there's other information that produces a more accurate care setting), and determination of the primary procedure and diagnosis.

Once processing is complete, the end result is reliable, consistent, and user-friendly 837 APCD that is ready for analysis. It’s the most sophisticated format of APCD available for quick, digestible insights.

What is Longitudinal Analysis?

Longitudinal analysis is based on 837 APCD. The longitudinal analysis of claims data allows us to identify patterns across cohorts of patients—tracking patient care across the full continuum of care, with elapsed time between steps on the patient journey.

Essentially, longitudinal analysis links individual claims together over time. Each one of a patient's claims is linked to all other claims for the same patient. These related claims can then be summarized and analyzed to answer questions about referral patterns, leakage, shared patients, and more. Viewers of this data can choose to analyze trends as to what happened either before or after a given "originating" event.

The Takeaway

Curated claims data and longitudinal data analysis are extremely powerful tools. But claims data alone, and especially in its raw, unprocessed form can be confusing to navigate and hard to pull intelligence from. With the right oversight, processing, and user-friendly tools that can query and visualize claims data, it can provide immense value.  

To learn more about Stratasan's APCD—how it could help you to analyze referral patterns in your market, increase in-network utilization, and identify which physicians you can re-align with your network—then schedule time to talk with one of our experts today.

Post by Brian Dailey, Founder and CTO, Morgan Atkins, VP of Product and Innovation, and Dave Sellers, Senior Data Analyst at Stratasan

Connect with Stratasan on LinkedIn or follow Stratasan on Twitter and Facebook

all-payer claims data apcd Longitudinal Analysis focus pathway