In a prior edition of Insights InHealth, we introduced the ValueSCORE™ ratings system and the innovations it incorporates to deliver unique intelligence into hospital performance. As discussed in the prior article, the ValueSCORE™ integrates measures pertaining to five domains of value (reproduced in the figure below). In the last version of Insights InHealth, we detailed Safety and Cost, the first two domains of the ValueSCORE™. This installment of Insights InHealth provides a deeper dive into the two more of the five domains, Outcomes and Engagement. For each value domain, we will discuss: (i) why it was selected for inclusion, (ii) which metrics comprised the domain and why, (iii) the data sources referenced, and (iv) technical notes regarding the application of the metrics to the ValueSCORE™.
Since the passage of the Affordable Care Act in 2010, the term “Triple Aim” has become ubiquitous when discussing the value of healthcare. The Triple Aim describes the three primary objectives for value-based care: providing better care and experiences for patients, improving the health of populations and reducing the cost of healthcare. Over time, the industry recognized an essential element was missing from these aims: ensuring the sustainable engagement of healthcare providers. This understanding has led to the increasing reference to the “Quadruple Aim” (adapted from this article [1]), which expands the lens of the Triple Aim to include provider engagement as a necessary objective in value-based care.
The COVID-19 pandemic has aptly demonstrated the need to adopt the Quadruple Aim paradigm, given the extensive reports [2] of provider burnout. Using the Quadruple Aim as the foundation of the ValueSCORE™, we identified the five domains based on their documented contribution to furthering these objectives.
The Third Domain: Outcomes
The Agency for Healthcare Research and Quality (AHRQ) defines outcomes measures as those measures that “reflect the impact of the health care service or intervention on the health status of patients.” Historically, outcomes measures have served three primary purposes. First, they help inform evidence-based clinical decision making. Secondly, they help patients identify providers with a track record of successfully treating particular conditions. Finally, outcomes measures help highlight unnecessary costs in the delivery of care.
Because of this utility, outcomes measures have long been included in value-based care and quality programs. For example, the Centers for Medicare and Medicaid Services (CMS) operates the mandatory Hospital Readmissions Reduction Program (HRRP) as part of Medicare reimbursement for inpatient services. The HRRP evaluates the “excess readmission ratio” (ERR) to assess hospital performance. According to CMS, “the ERR measures a hospital’s relative performance and is a ratio of the predicted-to-expected readmissions rates.” CMS computes the weighted average of each participating hospital’s ERR for six conditions, and this aggregate measure is then used to adjust reimbursement from Medicare’s Inpatient Prospective Payment System (IPPS) to each hospital. Facilities with readmission rates in excess of the median rate for a peer group (based on the hospital’s percentage of dually eligible Medicare and Medicaid beneficiaries) may receive up to a three percent reduction to IPPS reimbursement. The HRRP measures are also used in the Hospital Value-Based Purchasing (VBP) Program as part of Medicare’s IPPS.
Given Medicare’s requirements to report HRRP data and their widespread availability, InHealth Advisors uses the active HRRP metrics in the outcomes domain in the ValueSCORE™. We supplement this data with measures reported as part of the Inpatient Quality Reporting Program pertaining to complications and deaths at hospitals. Many of these same measures are also included in the Hospital VBP Program. The active metrics for the 2022 rankings are summarized in the table below; all data utilized in the 2022 ValueSCORE™ rankings was drawn from reports published in calendar years 2021 and 2022.
ValueSCORE™ Outcomes Metrics |
|---|
Complications and Deaths – Hospital |
MORT 30 AMI - 30-day death rate for discharged heart attack patients |
MORT 30 CABG - 30-day death rate for patients discharged after undergoing coronary artery bypass graft surgery |
MORT 30 COPD - 30-day death rate for patients discharged with chronic obstructive pulmonary disease |
MORT 30 HF - 30-day death rate for patients discharged with heart failure |
MORT 30 PN - 30-day death rate for patients discharged with pneumonia |
MORT 30 STK - 30-day death rate for patients discharged after stroke |
Unplanned Hospital Visits - Hospital |
READM 30 AMI - 30-day readmission rate for discharged heart attack patients |
READM 30 CABG - 30-day readmission rate for patients discharged after undergoing coronary artery bypass graft surgery |
READM 30 COPD - 30-day readmission rate for patients discharged with chronic obstructive pulmonary disease |
READM 30 HF - 30-day death rate for patients discharged with heart failure |
READM 30 HIP/KNEE - 30-day readmission rate for patients discharged after undergoing hip or knee replacement |
READM 30 HOSP-WIDE - 30-day readmission rate for all patients discharged from the hospital |
READM 30 PN - 30-day readmission rate for patients discharged with pneumonia |
As described in the last edition of Insights InHealth, the ValueSCORE™ ratings system uses a relative rankings system that places each of the over 3,300 hospitals within a percentile from 0 to 100. In order to reward facilities for reporting metrics, we apply an adjustment to the final domain score to account for the number of metrics reported. That said, we also implement a cutoff whereby a hospital is excluded from the rankings if it reported less than 25 percent of the included metrics. Additionally, we adjust the final domain score based on the change in performance from the prior period (i.e., the change in ranking from 2021 to 2022); the final domain score could be adjusted upwards or downwards by 10 percentiles depending on the change from the prior period.
To derive the ranking for the Safety domain, we employ the following process:
1. Determine the percentile rank for each of the 13 metrics above. We note that some hospitals may not report all metrics.
2. Determine the average percentile rank for the reported metrics.
3. Adjust the average ranking based on the number of metrics reported. For every metric that was not reported, the final score was adjusted downwards by 0.6 percent.
4. Apply an adjustment to the average ranking based on the change in performance from the prior period. Depending on the increase or decrease in percentile ranking (e.g., from 2020 to 2021 in the 2022 ValueSCORE™ rankings), the final score can be adjusted by up to 10 percentiles. The reward from performance improvement is higher if a facility increased from a lower percentile ranking starting point in the prior year; conversely, the penalty is higher if the facility exhibited a decrease in performance from a higher percentile ranking starting point.
5. Derive a final percentile ranking of the scores that incorporated the prior two adjustments.
We utilize the average percentile score as the starting point in recognition that some facilities may not have a sufficient volume of patients to report all 13 metrics. Nevertheless, to avoid penalizing facilities for reporting data, we apply the score reduction related to the number of reported measures. Both the reporting adjustment and adjustment for change in performance are capped to prevent biasing the scores too heavily away from the average across all reported metrics.
The Fourth Domain: Engagement
Patient satisfaction emerged as a common measure of healthcare value in the 1990’s with the development and proliferation of the ubiquitous Press Ganey survey. In 2002, AHRQ and CMS partnered to develop the CAHPS Hospital Survey (HCAHPS), the “first national, standardized, and publicly reported survey of patients' experiences with hospital care.” HCAHPS was endorsed by the National Quality Forum in 2005, implemented by CMS since 2006, and public reporting of results began in 2008. The Institute for Healthcare Improvement (IHI) introduced its Triple Aim framework in 2007, which integrated patient satisfaction with quality improvement and cost reduction as the major thrusts of healthcare transformation.
As the Triple Aim has expanded into the Quadruple Aim, the concept of provider satisfaction now receives a level of scrutiny comparable to that of patient satisfaction. With increasing physician shortages and growing provider burnout following the COVID-19 pandemic, healthcare employers now compete vigorously to recruit and retain provider talent. Certain measures from HCAHPS relate directly to provider engagement, and thus provide useful insight into this “fourth leg of the stool.”
To assess patient and provider satisfaction in the Engagement domain, we reference numerous measures reported within HCAHPS and the Outpatient and Ambulatory Surgery CAHPS. Many of these values are also included within the Hospital VBP Program. The table below details the survey items we include in the ValueSCORE™:
ValueSCORE™ Engagement Metrics |
|---|
Outpatient and Ambulatory Surgery CAHPS Survey Items |
Communication – Procedure |
Facilities and Staff |
Patient Recommendation |
Patient Rating |
Hospital CAHPS Survey Items |
Care Transition |
Cleanliness |
Communication - Medications |
Discharge Information |
Doctor Communication |
Nurse Communication |
Overall Hospital Rating |
Quietness |
Recommend Hospital |
Staff Responsiveness |
As with the technical discussion of the Outcomes metrics, the Engagement domain relies on a percentile ranking system. Below, we highlight the technical differences between the methodology applied to the Engagement domain relative to the Outcomes domain:
1. The Engagement scores were based on publicly reported data from 2021 and 2022 reports.
2. As with the Outcomes domain, we require 5 or more metrics to be reported in order to rank a hospital.
3. Given the slightly higher number of Engagement metrics relative to the Outcomes metrics, the adjustment based on the number of metrics reported is decreased to 0.5 percent per metric excluded.
A future edition of Insights InHealth will detail the formulation of the Risk domain adjustment in the ValueSCORE™.
Comments