Comparing Hospice Annual vs. Quarterly Claims Files

12/17/18 at 06:35 PM by Cordt Kassner

Comparing Hospice Annual vs. Quarterly Claims Files

Cordt T. Kassner, PhD, CEO Hospice Analytics, 12/14/18

 

Medicare claims files are the primary data source for describing hospice care. Medicare beneficiaries account for ~80% of all deaths, and ~50% of Medicare deaths occur in hospice. Medicare has released annual (i.e., calendar year) hospice claims files dating back to 1999 and through 2017. Starting 2016, Medicare also began releasing quarterly hospice claims files dating back to 2015Q3 through 2018Q2. Is there a difference in information included in the annual versus the quarterly hospice claims files?

 

To answer this question, let's take a brief look at how the claims files are created.

  1. Annual claims files are created ~6-months post calendar year end (LDS files; or 12-months post calendar year end for RIF), with no further changes, edits, or updates to the file. Quarterly claims files are created ~4 months post quarter end. (Historically, annual files were created based on finalized claims the date a request is approved. This means if I request the 2010 Annual Hospice Claims file and it is approved 9/12, it will be a slightly more complete file than if it is requested and approved 9/11.)
  2. Annual files are generally released in the fall, allowing ~8 months for claims to process and be finalized. Quarterly files are generally released after ~3 months. The time between claim submission and finalization is called the maturity period. Per a 2017 paper on Medicare Claims Maturity reviewing 2010 claims, below is a table summarizing the percentage of claims finalized for selected time periods after the service ends:

 

Table 1: Maturity Rates for 2010 Medicare Claims

Provider

1 Month

2 Months

5 Months

12 Months

Hospice

9%

84%

97%

99%

Home Health

29%

79%

96%

99%

Inpatient Hospital

48%

91%

98%

99%

 

As you can see in Table 1, claims by different provider types mature at different rates.

 

However, the paper above examined 2010 claims, and I wanted to compare the 2016 Annual Hospice file to the four aggregated 2016 Quarterly Hospice files. I also wanted to compare results from these files to results reported in the current Medicare Payment Advisory Commission (MedPAC) March 2018 Medicare Payment Policy: Report to the Congress. It is important to note that methodically MedPAC looks at hospice use over a beneficiary’s lifetime, while using the Annual and Quarterly files I look at hospice use by beneficiaries during a calendar year. Understanding claims maturity and this methodological difference explains most of the variance in results. Table 2 below compares various Medicare hospice claims metrics for 2016 Medicare from MedPAC, Annual claims, and Quarterly claims:

 

Table 2: Comparing 2016 Medicare Hospice Claims From Difference Sources

 

2016 MedPAC Report

2016 Annual

2016 Quarterly

Annual File % Difference From MedPAC

Quarterly File % Difference From MedPAC

Hospice Beneficiaries

1,400,000

1,439,528

1,413,180

2.8%

0.9%

Hospice Deaths

~1,099,000

1,048,448

1,046,194

-4.6%

-4.8%

Hospice Utilization

49.7%

47.4%

45.9%

-4.6%

-7.6%

Hospice Days

101,200,000

101,446,045

99,103,864

0.2%

-2.1%

Hospice Days Mean

88

70

70

-19.7%

-20.1%

Hospice Days Median

18

24

24

33.3%

33.3%

Hospice Payments

$16,800,000,000

$16,851,221,650

$16,458,795,117

0.3%

-2.0%

Hospice Payments Mean / Beneficiary

~$12,000

$11,706

$11,647

-2.4%

-2.9%

 

There are subtle, although interesting, differences when comparing the 2016 Annual vs. Quarterly hospice files nationally. As expected, there is more variation by state (e.g., hospice utilization rates between Annual vs. Quarterly files differ from 0.4% - 3.2%). I plan to explore similar differences in the Medicare Home Health Agency and Inpatient Hospital files.

 

Applications

This project highlights differences between the 2016 Hospice Annual vs. Quarterly Claims files, in comparison to MedPAC results.

  1. Are these differences enough to explain minor reporting differences between MedPAC and others working with Medicare claims files?
  2. Are these differences important enough to influence who uses which kind of file, and when?
  3. As a provider, it is important to better understand the data, where it comes from, and how to best use it.

 

Cordt T. Kassner, PhD
CEO, Hospice Analytics
v. 12/17/18

Back to Healthcare Policy