Hospice Quality Reporting Program Update
New Report on HQRP Data
September 16, 2014 09:17 AM
Earlier this year hospices submitted data to CMS on two quality measures that impacted the 2015 payment year. Specifically, hospices had to submit data on two measures – a structural measure and the NQF #209 pain measure – by April 1, 2014. The data was collected by the hospices during calendar year 2013. On August 18, 2014 CMS released the analysis of this data, Hospice Quality Reporting Program Fiscal Year 2015 Reporting Cycle Data Analysis. Approximately 3500 hospices submitted and attested to the data for both measures.
Nearly 100% of hospices answered “yes” to the only structural measure question: “Does your QAPI program contain 3 or more patient care related quality indicators?” This is consistent with the prior year’s response.
The NQF #0209 measures the percentage of patients who report being uncomfortable because of pain at the initial assessment and who report that pain was brought to a comfortable level within 48 hours.
The main purpose of the FY 2015 analysis for NQF #0209 is to answer the following questions about the NQF #0209 measure:
How many hospices can correctly report the (quality measure) QM, and what is the extent of data errors?
What is the extent of patient exclusion from the QM? How does the number of patients included in the measure denominator compare to the number of patients admitted to hospice?
Were there any changes in findings related to errors and patient exclusions between FY 2014 and FY 2015?
Were there any changes in the QM scores from FY 2014 to FY 2015?
NQF #0209 has been removed from the HQRP for FY 2016 and subsequent years because analysis of the FY 2014 reporting cycle data revealed two critical concerns with the measure. The first is that the measure does not easily correspond with the clinical processes for pain management, resulting in variance in what hospices collect, aggregate, and report. The second (and more important) concern is the high rate of patient exclusion due to patient ineligibility for the measure and patients reporting they are not uncomfortable due to pain at the initial assessment. This high rate of patient exclusion from the measure results in a small and non-representative denominator and creates validity concerns.
As with FY 2014, the majority of errors (2,270) for FY 2015 were due to providers having “missing” patient-level data that cannot be accounted for. Forty-one hospices (1.1%) had one or more errors that resulted in exclusion of their data from the NQF #0209 analysis, a slight decrease compared to FY 2014. The mean QM score was 64.8% (S.D. 23.9), indicating that on average nearly two-thirds of patients who were admitted with discomfort because of pain report that their pain was brought to a comfortable level within 48 hours of the start of hospice care. Our data shows that 9.1% of hospices had a “perfect score”; all of their patients that reported being uncomfortable because of pain on initial assessment reported that their pain was brought to a comfortable level within 48 hours. The findings for FY 2015 confirmed those of the FY 2014 data and reaffirm CMS’s decision to no longer require the NQF #0209 measure as part of the HQRP. Data errors affected 43.4% of all hospices’ data submissions, a 12.3 percentage point increase from the prior year, suggesting that hospices are not improving with time.
The findings from the FY 2015 closely mirror those from FY 2014 for both the structural measure and the NQF #0209 measure. This is true despite the fact that FY 2014 was based on one quarter of data and FY 2015 was based on four quarters of data. These findings support CMS’s decision to discontinue both measures for future reporting cycles in favor of implementing patient-level data collection using the Hospice Item Set (HIS) in the HQRP to collect standardized data to calculate seven QMs (six NQF-endorsed QMs and a modification of one NQF-endorsed QM).