Understanding Claims-Based and MDS-Based Data for SNFs

Best Source of Healthcare Data

Is one source of healthcare data better than the other, and which one should skilled nursing facilities (SNFs) be monitoring?

It has become cliché in today’s rich healthcare data environment to refer to Peter Drucker’s axiom, “What gets measured, gets managed.” While this is certainly true, or at least it would be very difficult to manage anything without some measurement of performance, effective management requires an understanding of the measuring stick.

Data sources, sampling methodology, measurement time frames, and benchmarks are all analytic components that must be understood in order to be used effectively.

In the post-acute care (PAC) space, it’s often necessary to distinguish between data sources that appear to measure the same thing, but actually are different. The challenge for REITs, Accountable Care Organizations (ACOs) and Health Systems managing a network of skilled nursing facilities (SNFs) is to understand the key characteristics of the available data sources without becoming mired in minutiae.

Let’s start with the two commonly referenced data sources for SNF quality measures that have been made publicly available by the Centers for Medicare and Medicaid Services (CMS). Understanding the origin and purpose of the data will give you the necessary context to know what is actually being measured.

In this blog post, we’ll take a look at each in the following ways:

• How do Claims-based data, and MDS-based data, differ?
• Which is better: Claims data or MDS data?
• Which quality measures should you monitor?

How do claims-based data, and MDS-based data, differ?

Claims data

When a SNF quality measure is referred to as “claims-based”, a common misconception is that the data source is billing claims that were submitted by the SNF for services provided in the facility.

In fact, the healthcare data that are used to measure SNFs’ performance in the quality measures for 30-day rehospitalization, emergency department visits, and successful discharge to the community are actually hospital claims for those Medicare beneficiaries that receive PAC services following an inpatient stay.

While clinical data from the patient’s SNF stay are used as covariates in the risk adjustment model for these measures, the actual “triggering” event (e.g. a hospital readmission) is driven by the hospital claim.

MDS data

In contrast, quality measures based on data from the Minimum Data Set (MDS) assessment reflect clinical outcomes that occurred during the SNF stay. The MDS is a standardized, structured assessment that is completed according to a mandated schedule for every patient/resident in a SNF – at minimum, an MDS is required by day 14 of the SNF stay, and upon discharge from the SNF. Additional assessments are completed for Medicare beneficiaries to set the reimbursement rates according to the Medicare Prospective Payment System (PPS).

The MDS assessment captures information on diagnoses, functional status, therapies, treatments, medications, cognition, mood, and behavior. The purpose of the assessment is to facilitate care planning and identify changes over time. This also provides the data which drives the 27 MDS-based quality measures that are currently publicly available.

Which is data better data: Claims or MDS?

A vote for claims-based data

The major item in the “pro” column for claims data is that the claim follows the beneficiary. For example, if a patient is discharged from the hospital to a SNF, and then after 14 days is discharged home, only hospital claims data would be able to capture an ED visit three days after the patient’s transition home. Claims enable us to see the broader post-acute picture rather than only what occurred in the SNF, which is a limitation of MDS-based measures.

On the other hand, claims-based measures are potentially lacking in the breadth of SNF services, since only claims for Medicare Fee-for-Service (FFS) beneficiaries are submitted to CMS. This may lead to an incomplete and possibly misleading picture of SNF performance if the facility has a more varied post-acute payer mix, i.e. Medicare Advantage payers.

Two votes for MDS-based data

What MDS-based data lacks in the ability to track the PAC stay between settings, it makes up for in clinical relevance. Since the MDS is a full clinical and functional assessment of the patient, it captures healthcare data that otherwise would not appear in a claim for Medicare FFS. For example, a SNF would bill Medicare for a PAC patient receiving skilled nursing and therapy under the Patient-Driven Payment Model (PDPM) classification, with the diagnoses on the claim to support that PDPM Clinical category classification.

However, if the patient also has clinical issues such as pain or behavioral changes that do not impact PDPM classification, this would not appear on the claim but it would appear on the MDS. Although PDPM is part of Medicare’s shift towards value-based models from the previous fee-for-service model, the existing claims-based quality measures lack the clinical richness of MDS-based measures.

Another point to keep in mind when you are trying to gauge SNFs’ current performance is the measurement time period of the healthcare data. One potential drawback of claims data is its timing; claims measures reflect a performance period that is generally one to two years old. The MDS-based measures are more up to date, with the publicly reported data covering a four-quarter average that is two to three quarters old.

Which data quality measures should you monitor?

Ideally? Both. Claims-based data and MDS-based data measures can be used together.

Since the claims measures can extend beyond the SNF stay, you can use them to gauge the effectiveness of care transitions between settings – from hospital to SNF, and from SNF to home health services. However, keep in mind that the measures do not distinguish between SNF and community as the origin of a hospital or ED visit.

To dig deeper into this, you would need to work with the SNFs to provide the missing details; many SNFs track their MDS-based re-hospitalization rates either internally or through third party analytics.

Also keep in mind that MDS-based clinical quality measures can identify both excellent performers as well as potential areas of high costs or liability risk (such as pressure ulcers, antipsychotic use, and falls with injury). As you work to maintain a strong, balanced portfolio of SNFs, be sure that the quality measures you use tell the full story of post-acute as well as long-stay resident care.