Calculating KPIs for Value-Based Payment Models
Organizations may have to measure new key performance indicators such as patient satisfaction, patient access to care, and gaps in care.
The new generation of key performance indicators (KPIs) under value-based payment models are constantly evolving. The question is, are organizations ready for this new wave of metrics, and do they have the data to succeed? John Kelly, principal business advisor at Edifecs, weighs in on how to calculate some of these KPIs and what healthcare providers need to do now to ensure long-term financial success.
Based on your experience, are organizations starting to move beyond traditional finance-related metrics now that so many payers are initiating value-based contracts?
Kelly: Many organizations—and especially those that have entered risk-based contracts with payers—are moving beyond traditional KPIs to measure metrics such as readmissions and ED [emergency department] rates. Even if organizations aren’t measuring these KPIs yet, they probably will be in the near future as more payers adopt alternative payment models that have already proven their ability to save significant money. For example, according to the Centers for Medicare & Medicaid Services (CMS), hospitals participating in the Comprehensive Care for Joint Replacement Model (CJR) saved more than $37.5 million during the program’s first performance year alone.
You mentioned readmissions and ED rates. What do organizations need to consider from a data perspective when measuring these and other value-based KPIs?
Kelly: Here are my thoughts on some ways to measure these new KPIs.
Readmissions. Most organizations are already looking at readmissions within 30 days of discharge thanks to CMS’s Readmissions Reduction Program that has reduced payments to inpatient hospitals with excess 30-day readmissions since 2012. What’s changed is the way in which health plans measure this metric. They aren’t looking at whether a single patient is readmitted within 30 days of discharge (and reducing payment for that one admission). Instead, they’re looking at total readmission rates for an entire population.
For example, a healthcare organization serving 1,000 patients for a particular health plan enters a value-based contract that puts it at risk for 30-day readmissions. Being at risk means the organization must prevent readmissions for a subset of these patients who are attributed to the organization. This idea of attribution is new, and it’s highly complicated. A patient could have Aetna insurance for three months, change jobs, and then have Blue Cross Blue Shield. So for three months of the year, the hospital could be at one level of risk for this patient and then a completely different level of risk for the remaining nine months.
To measure readmissions effectively, organizations must identify the following information for each payer with which it has a risk-based contract:
- Specific patients for whom the provider is at risk if a 30-day readmission occurs
- Risk that the provider incurs (i.e., If you exceed a specific readmission rate, you’ll incur this specific financial penalty.)
- Period of time during which the provider is evaluated against benchmarks for the at-risk populations
ED rate. Payers using anED rate KPI hold organizations accountable for staying below a certain number of ED visits per year. This target number is based on patient severity as evidenced by previous claims data. If organizations exceed this number, they’re penalized. If they stay below this number, they receive financial incentives. But again, it’s all about attribution—knowing the specific patients for whom you’re at risk and for how long.
Patient leakage. Patient leakage is another KPI that’s helpful to measure under value-based payment models. What services do patients tend to seek elsewhere, and how often does this occur? If an organization is at risk for a particular population, it needs to capture every dollar that’s spent on patients within that population. Doing so creates accurate cost benchmarks against which organizations are held accountable.
Patient leakage also calls into question an organization’s ability to control outcomes and prevent readmissions. For example, a patient undergoes an inpatient procedure at a local hospital but receives post-discharge care at a physician practice not affiliated with that hospital. What happens if that patient is readmitted, and the payer has attributed them to the local hospital? In some cases, the hospital might be at risk for penalties or reductions to incentive payments.
Total cost of care. Total cost of care is another KPI to measure. In a total cost of care contract, a per member/per month payment is established. Providers are successful when they identify at-risk populations, manage care for those populations within an integrated network, and deliver (and measure) efficient, high-quality care. Providers must be able to maintain an accurate core set of KPIs that dynamically capture performance and quality data and present it in real time so effective interventions can be initiated.
Value-based contracts may also require organizations to measure a whole host of other KPIs such as patient satisfaction, patient access to care, and gaps in care. Keep in mind that these KPIs are a moving target as health plan contracts evolve.
What steps should organizations take now to prepare for these newer KPIs?
Kelly: There are three steps they can take now to be successful under value-based payment models:
Know your limitations. Each payer contract drives the specific KPIs organizations should measure. Are you collecting the right type of data, and what’s the quality of that data? Are you capturing severity and risk of mortality as much as possible? Given the complexity of attribution, it may make sense to work with an outsourced vendor. Even though it can be costly to outsource, the idea is that you’re hopefully offsetting those costs—and earning even bigger bonuses—through value-based payment incentives.
Generate real-time dashboards. Being able to examine your performance under these payment models in real time allows for immediate action and prevents future revenue loss. These dashboards should include payer-specific KPIs, targets/benchmarks, and real-time performance relative to the benchmarks.
Start somewhere. Even if none of your payer contracts are risk-based, it’s a good idea to start thinking about ways that you can build risk into your contracting strategy using data that is advantageous to your organization. For example, have you recently implemented early invention programs with a large network of physicians? If so, admissions per thousand may be a good metric to include.
Have you already built the infrastructure to support public reporting, CJR, Merit-Based Incentive Payment System, or other bundled payment programs? If so, can you integrate some of this data into your value-based contracts? For example, your organization may be required to report opioid use and abuse to your state department of health. Can you include ”potential opioid users identified per day” in a payer contract—especially if you’ve implemented community outreach and prescription management programs?
Sometimes measures a hospital should look to produce are not objective targets but improvements over time. For providers who specialize in joint replacement, a good quality measure is the rate at which you shorten the time for patients to return to normal daily activity. Another is to target a percentage for the reduction in the rate of complications and readmissions. It’s about knowing what your organization already does well, implementing process improvements—and then making the data work for you.
See related article: 4 KPIs to Increasing Profitablity Under Value-Based Payment Models
Interviewed for this article:
John Kelly is a principal business advisor at Edifecs.