Everyone can imagine a home health agency, probably the one they work for, and conclude that this definition applies to all of them. In this post, we will explore how HHAs view themselves in order to define them in our data.
A couple blog posts ago, I took a diversion from the cost report data analysis to explain how this data is cleaned by CMS, and myself, to make it relevant for use. In this post, I am going to explore two seemingly simple and important questions about the data. What is a home health agency and how can we compare their financial performance?
As I have been breaking down the aggregate cost report data, I have found significant insights as it relates to subgroups of the national home health industry. Before I relate this to you, it is important that we have a mutual understanding of what this data means by having a shared definition of the units of the KPIs (Key Performance Indicators). In order to share these results associated with an HHA, we have to agree on a definition of these entities and understand the nature of this unit of measure including its limitations. The same is true when measuring financial performance.
How do we define a HHA (Home Health Agency)?
Up until now, we have looked at home health cost report data for all free standing home health agencies combined. This is important for analysis regarding the proposed and final rules because this is the only point of view used by CMS for these regulations. Using this perspective, the industry consists of one large combined home health entity serving all Medicare beneficiaries everywhere, as well as all patients with Medicaid and Other insurance.
The flaws in this point of view emerge when you scale this down to other definitions of HHAs. For example, as I described in a previous blog post, when CMS estimates future HHA expenses for the proposed rule, they make the flawed assumption that any reduction in visits lowers HHA expenses for all HHAs equally and by the calculated cost of the visit. This error is directly related to their perspective of HHAs as being a single large entity.
The data reveals that smaller agencies with less staff can’t take advantage of these incremental visit cost reductions because, for them, they do not result in reductions in staff and consequently, visit costs. There are other significant differences in HHA subsets that we will explore in the next post, but first, we need to agree on what an HHA is and how to measure the KPIs.
Most of us know HHAs by name. When we refer to specific ones, that is how we describe them to others. When it comes to BI, names are a poor indicator of the identity of an agency. When developing data insights, we have to be able to connect HHAs with their data without errors or omissions. For this purpose, we need a coded value that will be present in the data that is unique for each agency. There are two provider identification systems used in the industry. The first is the CCN (CMS Certification Number) and the second is the NPI (National Provider Identifier). The CCN is used for all published Medicare data. When Medicare publishes the cost report data, this is the value it uses to identify the agency and connect the data to the HHA provider.
The CCN is assigned to the HHA by the MAC (Medicare Administrative Contractor) when the agency is ready to begin submitting claims. These MACs also process the cost reports for CMS when they are submitted by the HHAs. In this environment, from the perspective of the MACs and CMS, the CCN is the definition of an HHA. Under this definition, the HHA is the unique entity that submits home health claims to CMS.
When HIPAA was introduced in 1996, one of the requirements was to create a NPI (National Provider Identifier) to replace all the unique values used by all the different health plans to identify providers when processing claims. The HIPAA legislation also required identifiers for health plans and for individuals, but the NPI was the only one of these national IDs that was implemented.
The NPI is considered the most valid identifier for providers. The NPI registry includes not only all provider organizations by physical location, but all individual healthcare providers as well. The organizational NPI value is required for all insurance claims submitted to any health plan, the CCN only applies to Medicare Part A.
It would seem that if you would use any ID to represent home health providers, you would use the NPI. CMS includes the NPI as a value in the cost report data in the report record, it also includes the CCN. The CCN is unique and included with every cost report. CMS includes in the cost report data a list of all HHAs by CCN including name, address and other demographic data. However, in the cost reports, all the NPIs are blank. Here is the reason why.
The CCN has been around since before the processing of claims electronically began in the 1980s. The CCN predates the NPI because of this. When the NPI registry was created, all providers were required to create organizational records in the NPI registry in order to submit claims under the new ANSI 837i electronic claim format, implemented in 2001. An organizational NPI had to be created for each distinct physical location.
When the CCN was created, it was defined as each individual entity of a specific provider type, home health in our situation, submitting claims to Medicare. As it turned out, many of these CCNs submitted claims for multiple physical locations (NPIs) that operated as a single entity. Based on an analysis I did a few years ago, about 5% of the CCNs include multiple NPIs. This is why the NPI is blank on the cost report.
Sometime after CMS decided to include the NPI in the cost report data, CMS discovered this issue. There is not a one to one relationship between CCN and NPI. Another problem is that many providers identified by the NPI might have multiple CCNs if they submitted claims under multiple provider types to Medicare. This is the case when an HHA, defined by the NPI, is also a hospice.
Because of this, these provider identifiers represent two slightly different definitions of HHAs and perspectives when viewing the data. If we are trying to describe HHAs as they define themselves, the CCN is more valuable because those CCNs with multiple NPIs are operated and managed collectively as a single entity. If our objective is to develop a way to match HHAs managed collectively as a single unit, I have discovered another way to identify these HHAs with unique CCN values that are managed collectively at a level above the CCN.
When you take the list of all cost reports for a given year and sort them alphabetically, you will see many that match or nearly match at the name level. Some of these matching values represent organizations we all recognize as entities that manage many different HHAs.
However, the name itself can’t be used in many cases to establish this link reliably. There may be many “1st Choice Home Health Care” simply because this appears at the beginning of any list of agencies by name. Other common names used by unrelated agencies for this purpose include “ABC” or “A1”. Other HHA names use common adjectives intended to project an image of quality like “Best”, “Better”, “Blessed”, “Care”, “Choice”, “Comfort”, “Community”, ”Complete”, “Dependable”, etc. Many include “Home Health” or Home Care” in their names that are unrelated to each other. There are many Christian based HHAs that use similar names like St. Mary or St. Luke. Because of this, we can’t make definitive connections between these agencies based on name alone.
Fortunately, the cost report gives us another indicator of collectively managed HHAs. In the balance sheet portion of the cost reports, HHAs must submit total assets. Cost reports must be submitted to CMS for each CCN individually. When an entity that considers itself to be a collection of HHAs by CCN submits these reports, they provide unique revenue, expenses, and statistics like census and visits for each of the CCNs. However, they do not separately book assets for these agencies in their own financials so these entities enter the asset value for all the CCNs combined repeated in each of the individual cost reports. Using this along with the names, we can definitively establish that these agencies are connected at this organizational level.
Here is an example using 2022 cost report data:
I discovered this indicator when viewing all financial data from cost reports in a list of HHAs by CCN sorted alphabetically. This is not a coincidence, but a valuable way to create a new level of organization of HHAs to improve our perspective of them in a way that matches how they see themselves.
For this post and future ones, I will refer to this level of organization as the “Enterprise”. This entity is a collection of cost reports for HHAs that are joined by the same management structure.
In addition to this indicator, I have also used the HHA name (CMS descriptions by CCN) in circumstances where the assets provided are unique, but the name included in the cost report data is complex and identical or structured identically for all CCNs with that name. This includes enterprises like “SUNCREST HOME HEALTH” or “CENTERWELL HOME HEALTH”.
This organizational level is constantly changing as these enterprises acquire new HHAs by CCN, rebrand and otherwise change at this level. Since cost reports are submitted annually, we can identify enterprise organizations annually, but these definitions are not valid for prior years or in future cost reports.
I went through this exercise for 2022 manually, linking all of these cost reports by a common name that I created for each enterprise. The criteria was these shared asset values and complex name matches in the cost report data. In some cases, it was not completely clear if an individual agency cost report was part of an enterprise. When I ran into this situation, I excluded them from the enterprise. I felt it was more valuable to make sure every agency identified as an enterprise agency was connected than to try to include all possibly connected HHAs by CCN.
Based on this limitation, we can assume that there are many HHAs by CCN that are actually connected to other non-enterprise HHAs or existing enterprises that have not been included as enterprise agencies. The end result was 249 enterprise organizations that exist in the cost report data for 2022. These enterprise organizations represent 30% of all HHAs by CCN and 44% of the total census.
In order to further explore the data associated with the enterprise agencies and to compare financial performance, I created two groups within the enterprise organizations, one is the top 10 enterprise organizations measured by their collective census and the other is the remaining enterprise organizations.
If we take the total census for all patients in 2022 and classify it by these enterprise groups and the remaining HHAs, this is what we see:
In future blog posts, we will use this new designation (enterprise) to analyze the performance of this organizational level in comparison to non-enterprise HHAs and large enterprises (by census) compared to smaller enterprise organizations. This will be in addition to other KPIs we have used so far like profit margins, visits, census and others.
How can we measure the relative financial performance of HHAs?
If we are going to measure financial performance of subsets of all HHAs and compare it to the entire industry or other HHA groups, we have to have KPIs that accurately reflect the relative performance of these groups when compared to each other. The most obvious KPI that we might use is dollars, the common unit of most revenue cycle KPIs. Although this is a valuable measure of financial performance for a given HHA over time, it does not work well when comparing HHAs. The reason is that the dollar has a different relative value depending on the location of the HHA services performed and where clinicians are paid.
Under PDGM, the wage index is used as a component in the payment formula in order to compensate for these differences. This index value is intended to match the difference in wages throughout the country centered on a value of 1. These values range from 0.65 to 1.50. This means that dollar values used to compare an HHA to others would mean very little due to this large variance.
CMS uses this value to mirror payments to costs throughout the country in an effort to neutralize the impact of these wage variances. This is not just with HHA payments, but all Medicare healthcare provider payments.
When CMS applies the wage index to these payments, they do so at the claim level. The claim data and the Medicare beneficiary database determines the county where the patient lives or is receiving treatment. This is what determines the wage index for the claim. The definition of a Medicare billable home health encounter includes at least one therapy visit to the patient’s home. We can assume that the clinician providing this treatment belongs to the same wage index area as the patient. This is CMS’s assumption they use regarding the wage index.
If this is true, then although revenue and expenses can fluctuate, they should mirror each other and maintain a consistent ratio, or profit margin, as the value of the dollar fluctuates for individual HHAs. Furthermore, we should be able to measure other ratios of dollar values between these HHAs to measure relative performance. For example, we could compare the total of clinician wages to total operating expenses. The remaining fraction would be operating expenses in addition to clinician wages. This could possibly be a KPI that could measure operating efficiency or the degree that HHAs could limit their non-clinician expenses as a ratio of total revenue or profits.
Other metrics can be used as KPIs that do not involve wages/revenue like census and visits. Visits per census (CMS uses 30 day periods in PDGM) is a key metric used in the proposed and final rules and is the basis for the behavioral adjustment. We will see that this KPI, like profit margins in the previous blog posts, looks very different when you look at HHAs with different shared characteristics.
Comments