AHA/ASA Presidential Advisory:
Stroke Outcomes Measures Must Be Appropriately Risk-Adjusted To Ensure Patient Quality Care: A Presidential Advisory from the American Heart Association/American Stroke Association: G. Fonarow, et. al
Summary: This paper details concerns and issues with the new Centers for Medicare & Medicaid Services (CMS) 30-day stroke mortality and 30-day stroke readmission measures. Because these measures do not include a valid stroke severity measure such as the National Institutes of Health Stroke Scale (NIHSS), they may be prone to mischaracterizing the quality of stroke care being delivered by hospitals and may inadvertently cause negative care and systems of care issues for acute ischemic stroke patients.
This paper details 1) why the CMS acute ischemic stroke outcome measures in their present form may not provide adequate risk-adjustment; 2) why the measures as currently designed may lead to inaccurate representation of hospital performance and have the potential for serious unintended consequences; 3) what activities the AHA/ASA has engaged in to highlight these concerns to CMS and other interested parties; and 4) alternative approaches and opportunities that should be considered for more accurately risk-adjusting 30-day outcomes measures in patients with ischemic stroke.
- Risk adjustment is the statistical process of determining appropriate costs to healthcare plans based on risk profiles of patients sharing similar characteristics. Medical insurance plans use risk adjustment to determine how much they will reimburse hospitals for the care they provide. To accurately assess and report hospital level outcomes, adequate risk-adjustment for case-mix is essential.
- Using administrative claim data for ischemic stroke measures may not properly reflect severity of the stroke case, so it’s essential for acute ischemic stroke models to include adjustment for stroke severity. Several studies demonstrate that initial stroke severity as indexed by the NIHSS is the principal predictor of mortality in acute ischemic stroke. Adjustment for stroke severity is essential for optimal ranking of hospitals with respect to 30-day mortality.
- The NIHHS Stroke Scale, with scores from 0 to 42, is widely used in clinical practice and is also linked to medical practice by federal drug labeling for tPA. The NIHSS Stroke Scale is a well-validated, highly reliable, and an extremely strong predictor of both mortality and short and long term functional outcomes.
FLAWS IN THE 30-DAY MORTALITY AND READMISSION MEASURES
- Adequate case-mix adjustment for acute ischemic stroke by claims data may be particularly difficult as it is essential for acute ischemic stroke models to include adjustment for stroke severity. The CMS 30-day stroke mortality and 30-day stroke readmission measures were not endorsed by the National Quality Forum (NQF) nor approved by the NQF Neurology Committee. Several organizations argued against the CMS measure.
- Models that do not account for stroke severity may produce different rankings of hospital performance. Hospitals who admit the sickest of the sick stroke patients (most severe strokes) will be penalized if the patients dies or is readmitted within 30 days.
- There is growing evidence that suggests the primary drivers of variation in 30-day readmission rates involve factors not captured by these measures: (i.e., poor social supports, socioeconomic status, differential access to rehabilitation facilities, and inadequate community resources). Each of these factors is especially important to ischemic stroke patients who are often disabled and unable to care for themselves after their hospital stay.
AHA/ASA EFFORTS REGARDING STROKE MEASURES
- AHA/ASA stroke experts expressed some of these concerns in the development process of these additional 30-day outcome measures.
- In the initial 30-day public comment period, the AHA/ASA submitted comments to CMS on the proposed measures, suggesting that applying the standard cardiovascular risk-adjustment strategy developed for cardiac conditions to stroke outcomes assessment was inappropriate given that stroke is a distinct neurovascular disease, and reinforced the need to include stroke severity.
- The AHA/ASA specifically endorsed the NIHSS as a measure of severity in its comments to CMS.
- The AHA/ASA expressed its concerns that both the 30-day mortality and 30-day readmission measures were not well-validated, since the measures were not published in a peer-reviewed journal.
- During the NQF comment period, the AHA/ASA and other professional societies reiterated these concerns, and highlighted two recent papers that demonstrate that regarding the mortality measure, the NIHSS is highly correlated with outcome.
- In a public comment letter to CMS, the AHA/ASA reiterated prior concerns and gave observations as to why CMS should not adopt these measures into the Inpatient Prospective Payment System (IPPS).
POTENTIAL FOR NEGATIVE IMPACT/PUBLIC PERCEPTION
- Both stroke measures, as constructed, may be prone to mischaracterizing the quality of stroke care delivered by hospitals and have the potential to harm patient care by undermining stroke systems of care. Hospitals may consider turning away patients with more severe strokes or transferring them to other hospitals after emergency department assessment to avoid being misclassified as having worse risk-standardized outcomes.
- Safety-net hospitals that care for more poor and minority stroke patients are at risk of being disproportionately impacted by a measure that does not account for stroke severity.
- Regional stroke referral centers, which are the most qualified to treat patients with severe strokes and which often have the most severely disabled stroke patients, may also be negatively impacted because they may be assigned a low performance rating. This could result in significant financial penalties for these hospitals.
- CMS has stated that results for both of these measures will be reported publicly on the Hospital Compare Website, where the performance of hospitals will be publicly available. Consumers viewing this information may mistakenly believe that safety-net hospitals, teaching hospitals or stroke centers are poor performers and avoid these hospitals.
- While these measures are not currently included in pay for performance, it is foreseeable that these measures will eventually be incorporated into such a model of payment by CMS, and potentially by other payers, or in state initiatives.
- Require NIHSS data collection to determine the severity of the stroke. CMS could, at very little cost, require this for risk-adjustment in all eligible patients and hospitals with little or no missing data. These data are routinely and voluntarily reported by the 1,702 hospitals participating in the AHA’s Get With The Guidelines-Stroke.
- Require a stroke severity variable by creating an interim code set that does capture NIHSS in administrative claims data.
- Develop Electronic Health Record specifications of the 30-day stroke outcome measures that include an index of stroke severity.
ACTIVITIES UNDERWAY/PLANNED BY AHA/ASA
- AHA/ASA is planning to conduct additional research on data collected as part of its GWTG-Stroke program and other sources. Once the current CMS stroke measures are publicly reported, AHA/ASA will analyze whether Primary Stroke Center and Comprehensive Stroke Centers are more likely to be classified as having worse than national average 30-day outcomes using the CMS stroke measure. The AHA/AHA is collaborating with the Joint Commission on standardizing the collection of initial NIHSS for patients hospitalized with acute ischemic stroke. In addition, the AHA/ASA will continue offering a certification program in the NIHSS for U.S. physicians, nurses, and allied health personnel at no or nominal cost, a program so far completed by 562,088 providers.
The AHA/ASA is committed to partnering with others and providing CMS with additional data, expertise and support in developing revised versions of these measures to promote truly risk-adjusted outcome measures. We recognize that revising measures will take time; however, unintended consequences could occur from the use of these measures in their current form.
We are committed to work with CMS and other interested stakeholders to ensure that appropriate 30-day risk-adjusted stroke measure are created as this is fundamentally critical for care of stroke patients throughout the United States.
To that end, we urge all stroke stakeholders to share any concerns and issues directly to CMS. There will be an opportunity to provide public comment when the IPPS rule is announced. Be on the lookout when CMS updates the IPPS regulations in the next few months, with comment periods open prior to implementation of the final rule.
AHA/ASA Comments on 2014 IPPS Regulation: