Hospital finance leaders have a pivotal role to play in improving GAAP and quality compliance under pay-for-performance contracts, and to fulfill that role, they must have a sound work plan and make sure it is well executed.
As health plans continue to move away from strictly fee for service toward paying for outcomes, hospitals are challenged to address two important questions:
- How much revenue are our value-based contracts generating?
- How much compensation is unclaimed because we did not meet performance criteria under the present terms?
A recent survey disclosed that one of every three risk-contracting entities in the healthcare industry lacks financial projections for its contracts. a The survey also found that, among those that have developed financial projections, 92 percent could not update them regularly because of a lack of staffing or data availability, or because of limitations in data due to current operating processes.
Hospital leaders can gain valuable insights from a process that is currently being worked by a number of organizations across the country that will help safeguard the accuracy of financial statements, enhance revenue to the maximum under existing contract language, and improve patient care.
Compliance with GAAP generally requires accruing large contractual adjustments—sometimes amounting to 90 to 95 percent of gross revenues. The accounting department might apply historical collection rates or use an exact fee schedule to estimate the contractual. Both approaches are gradually becoming obsolete, as health plan contracts begin to include post-period settlements whose amounts are based on performance metrics that are outside the traditional finance arena. Unless the monthly financial statement accrual process can incorporate the financial impact of these performance metrics, the financial statements will at some point fail to reflect GAAP.
Currently, many organizations are seeing only 3 to 15 percent of their revenue streams affected. But the percentages are growing, and when one considers the contribution margin of these contracts relative to the bottom line, the significance becomes more readily apparent.
In 2012, one large health system participating as a Medicare accountable care organization (ACO)anticipated receiving a bonus of between $15 million and $20 million, but because the health system was unable to quantify the amount, it did not record that revenue in its financial statements. When the missing revenue was received, it proved to be about 4 percent of the health system’s total Medicare revenue.
Clearly, the financial impact of the risk contract for this ACO was highly positive. But the impact can be negative as well; the previously cited national survey found that, while 48 percent of organizations attest to having contracts with only upside risk (i.e., sharing in savings but not bearing financial responsibility for losses), as many as 52 percent say they have contracts with downside risk (i.e., assuming financial responsibility for the quality and cost of care).
Metrics and Domains
To understand how to measure the impact of risk contracts, it’s necessary to begin by looking at how the metrics affect provider compensation. There are generally two broad classes of metrics utilized: process metrics and outcomes metrics.
Process metrics. These metrics measure the frequency with which something has occurred, regardless of the outcome. An example is the number of patients presenting with chest pain who received an electrocardiogram test within 30 minutes of arriving in the emergency department (ED).
Outcomes metrics. These metrics measure the frequency with which a specific outcome was achieved. An example is the survival rate of patients experiencing an ST-elevation myocardial infarction (STEMI) who were admitted through the ED.
Metrics usually are logically grouped into domains, with a portion of the overall value-based payment incentive assigned to each domain. The exhibit below, for example, shows the FY17 domains and their relative impact on the overall bonus paid or penalty imposed by the Centers for Medicare & Medicaid Services (CMS) under its Value-Based Purchasing Program. Various decision rules and formulas are then applied to determine how each of the metrics assigned to the domain are factored in computing the domain incentive.
FY17 Value-Based Purchasing Program Domain Weighting
Although pay-for-performance contract language has been evolving, with increasing detail and complexity, there are few standard metric definitions and formulas. Many commercial contracts are based on the models established by CMS or currently used by professional medical associations, such as the Society of Thoracic Surgeons. Paraphrasing what can be long and complex paragraphs, the language is often reducible to the following simple formula:
Achieve X in this measure and get paid Y
In this formula, X might be one of the following:
- An absolute number—e.g., achieve 0 postsurgical infections in the contract year
- A rate of improvement from a base period—e.g., achieve a 50 percent reduction in postsurgical infections
- A peer quartile from a base period or the current reporting period—e.g., achieve at least the 75 percent quartile in the organization’s peer group, as defined by size or geographic location
The Y bonus could be a specific amount, like $50,000, or a percentage of past or future claims payments, paid as a lump sum or as an add-on to each claim. Or it also could be based on a pool of patient premiums (per member per month) or on internal savings achieved by the insurance company (presumably because of the healthcare organization’s superior contract performance).
Work Plan Steps
The use of metrics may present additional complexities, but the points outlined here provide a sufficient basis for understanding the next step for the organization’s finance leaders: to develop and implement a work plan for measuring and improving payment and, ultimately, producing GAAP-compliant financial statements. Such a plan should include the following steps.
Sample Contract-Summary Spreadsheet
Read and review all current managed care contracts looking for incentive and penalty clauses. During this review, it is helpful to create and maintain a spreadsheet that summarizes the key elements of each contract and tracks changes, as shown in the exhibit above. This high-level document can be shared with the organization’s managers to provide them with a simple at-a-glance overview that makes it unnecessary to go back and reread the contract.
Studying contracts can be tedious, and it is important to devote sufficient time to the effort, which could take a month or more to complete. As an example, the review process for one health system involved poring through a 5-inch-thick stack of contracts, about 20 of which were determined to include performance provisions, amounting to a catalog of about 200 measures. b Moreover, contract language is not always clear, and it may be necessary to contact the insurer for clarification, adding to the reading time. Regardless of the time commitment, it is essential that this part of the project be done thoroughly and accurately.
Once completed, the summary document can be used to drive the agenda for the year. For example, some analyst time could be allocated to review details on a contract that pays a lump sum 90 days after the contract period so that the insurer’s computation of the penalty or bonus can be reconciled with the healthcare organization’s internal records.
There may be up to four critical time periods to be considered in a contract, as represented in the exhibit: First, there is the familiar “Contract Term”—the period during which the current document is legally binding. New concepts include the “Base Period,” the baseline from which comparisons with metrics reported during the “Reporting Period” will be made to evaluate whether performance has improved or deteriorated. The “Payment Period” is the timeframe when any incentive or penalty payment earned during the contract term would be paid out. These dates are generally not aligned to the contract term to allow time to complete medical records documentation, coding, and claim settlement so that the reporting period data are static, final, and complete before settlement calculations are run.
The “Total Revenue” element included in the sample spreadsheet shown on page 60 refers to revenue subject to the incentive. It is important to be clear about which patients are subject to the incentive program. Contracts may exclude transfers, outpatient and prosthetic services, or separate incentive calculations for inpatient and outpatient services. And hospitals and physicians often have separate contracts and incentives.
The “plus or minus” in the “Maximum Penalty/Bonus” element identifies whether the contract has upside or downside impacts, or both. Many contracts with downside potential include a withhold percentage from each claim so the health system is not caught short of cash if a penalty occurs. However, withholds, unless properly recognized, can contribute to an overestimation of the contractual element. On the other hand, contracts with a risk of a penalty and no withhold may require the health system to establish a reserve.
Build dashboards to facilitate monthly status reporting on each metric. In this phase, the finance leader should develop a summary sheet that provides an at-a-glance understanding of each metric. The summary should identify the sources of values relative to each metric in the organization’s clinical systems and record the metrics’ value relative to different health plans. The simplified sample summary sheet shown in the exhibit below illustrates an effective way to approach this process. This sample summary sheet is for illustrative purposes only and provides detail for only the first metric listed, leaving space of two additional metrics.
Sample Metric Summary Sheet
The sample sheet uses a simple set of headings to describe each metric without the need for a lot of verbiage.
The first column refers to the metric’s “Domain.” Metrics are assigned to a domain by each health plan. It is possible that different health plans may assign the same metric to different domains, necessitating a second entry in the summary sheet.
The “Metric ID” (second column) is an arbitrary sequential number used to uniquely identify the measure. The ID could be used to encode additional meaning, if desired. When grouping metrics, care should be taken to differentiate between metrics that are similar but not exactly the same. For example, two health plans may appear to have identical mortality metrics that actually have significant differences apparent only on closer inspection, such as the inclusion or non-inclusion of neonatal patients. In such instances, it would be important to create a separate metric ID for each. Ideally, the summary sheet should provide sufficient detail to allow an analyst or programmer to assemble the supporting data set and replicate the computation for the dashboard when it is built.
Here are some other factors that may distinguish otherwise seemingly identical metrics across health plans:
- How the measure is adjusted if data are missing
- How many patients must be in the sample to make it large enough to be assigned weight in the performance formula
- The presence of outliers and how they are they factored in
- The effect of insured patients dropping in or out during the contract period
- How comparisons between current and performance years are affected by changes made to measurement formulas between years
In the column labeled “Metric Name,” each metric is assigned a meaningful name; the standardized name can be used if the metric originates from a nationally recognized measure.
“Type” indicates whether the metric involves a percentage or a point value. And “Source” identifies where the data can be found—possibly internally, in the electronic health record, or in the patient billing system. For measures that come from external sources such as Premier, the Society of Thoracic Surgeons, CMS, or the Centers for Disease Control and Prevention, arrangements should be made either to intercept the hospital’s data set during submission or to receive digital copy of the data set after it has been appropriately massaged by the external agency.
“Computation” provides a space to record the exact formula or queries used to generate the measure. This information could include a report name with relevant field/column references, or an actual SQL Query.
“Current Value” reflects the most recent data point available. This data field would be subdivided into numerator and denominator in the event the “Type” is a percentage.
“Target Info” provides a space to record a human-friendly description of significant computation aspects of the measure in contrast to “Computation,” which is more technical in nature. Sometimes the target is a single point, and sometimes it consists of tiered thresholds with different incentive values for each. Again, the objective should be to capture as much information about how the metric works as possible to avoid having to go back into the contract to relocate all the relevant terms, which can be a time-consuming and tedious process.
Each health plan also is assigned a column, in which a facility-specific coding system can be used to describe how the health plan uses the measure. In the example, for Health Plan 1, the number 5 indicates the metric is number 5 within the Experience domain, and the “>64” indicates that the target threshold is an absolute value of 64 (i.e., scoring 64 on this HCAHPS survey question). Numbering the metrics according to health plan sequencing facilitates reporting and may also be helpful if the health plan only counts the best-performing metric within a domain.
By comparison, for Health Plan 2, the metric is the first value in the domain and the target threshold has a value of “>1Q” indicating the target threshold is to be achieved in the first quartile. Quartiles refer to a ranking of the organization’s score within a peer group—and because an organization cannot know the scores of its peer group until after the reporting period ends and its peers have submitted their values, the organization cannot attach an actual numerical value for modeling. Possible interim solutions are to use a known value from a historical period or to forecast a proxy value for the current reporting period based on extrapolations. Substitutions for actual values should be footnoted as such so that users of these summary sheets dashboards are not misled.
The finance leader should conclude this step by meeting with his or her organization’s IT representative and other stakeholders to determine how best to consolidate all the metrics into a centralized digital warehouse. IT staff may recommend a client-server database, such as MySQL or Oracle, and some generic web-based query and reporting tools, or they may offer a dedicated dashboard tool. The lowest-cost option would be to develop an Excel- or Access-based dashboard. The Visual Basic (VBA) language included with all Microsoft Office applications also could be used to create sophisticated macros to automate the more complex or time-consuming data manipulations and replicate the functionality of more costly tools.
Another important decision requiring IT input is the frequency with which the metric data set values will be updated. A best practice is to collect the data monthly, which coincides with the monthly financial reporting cycle used in most hospitals. However, budgetary constraints and the cost of extracting these measures from the hospital information system or of manually computing them and populating the reporting database may make it necessary to limit the update frequency to quarterly.
Identify metrics that present the greatest improvement opportunities. With a clear idea of the organization’s contract terms and metrics, and with a database in place that can provide monthly, quarterly, and annual values for each measure, the finance leader will be ready to begin the data analysis. It should be possible at this point to calculate the maximum ranges of the incentives and penalties that are possible under each contract. Using hospital-specific metrics, the finance leader can further estimate the organization’s position on the incentive-penalty continuum and determine the dollar value associated with making marginal improvements to each measure.
The tool should not only allow data to be filtered by health plan, bonus versus penalty, time period, and other factors, but also enable an overall report to be produced that includes all payers and all metrics and that highlights those metrics that present the greatest opportunities. In the sample report shown in the exhibit below, total dollar amounts attached to each measure are summed across all health plans, the amounts earned are estimated based on the organization’s current level of performance, and the opportunity amount is computed by taking the difference between the organization’s current amount earned and the maximum available. This type of report, sorted from highest to lowest opportunity, should be presented to management.
Sample Report Showing Estimated Total Dollars Associated With Measures
Identify “leverage points” within network processes. Having determined which metrics present the greatest opportunity, the finance leader next can solicit help from operations personnel on how best to improve performance on the metric. This effort involves identifying which hospital processes need to be tweaked to make the greatest impact on the metric and then meeting with the person or persons responsible for that process to jointly develop a strategy for moving the value closer to achieving the maximum.
The sample in the exhibit on page 63 discloses that the hospital has a maximum incentive of $185,000 across all payers for the metric “HCAHPS: Would Recommend.” The report further reveals that only $85,000 of that incentive is being earned, leaving $100,000 unclaimed. A detailed review of which health plans compensate for this metric and how they compute that compensation shows that improving the value from 64 to 74 would allow the remaining $100,000 to be earned.
At this point, finance can share this information with nursing and a multidisciplinary process improvement team could be established. At the first meeting, nursing might suggest that the reason scores are arbitrarily low is that routine maintenance work orders are being completed by the evening shift maintenance personnel, which is upsetting to patients. The team might then negotiate with maintenance to perform their routine work during the daytime for a 60-day period to see whether the scores improve. Assuming scores do improve, the team next would recommend to administration to permanently refrain from performing routine work orders in the evening. A 5-point improvement in the metric would lead to an additional incentive of $50,000. The multidisciplinary team then would continue to look for ways to improve the scores to eventually gain all $100,000 of the initially lost incentive. Of course, in the real world, such changes are not achieved so quickly and easily, which is why the following steps also are recommended.
Use change management principles to create incentives for desired behavior. Improving performance on metrics has the potential to be one of the more difficult challenges in this work plan. Hospital cultures have a history of resisting change and operating like independent fiefdoms. But it helps if strategies are based on a sound understanding of human psychology and the use of a well-tested model. One such model is the “Switch” model described by Chip and Dan Heath in their book Switch: How to Change Things When Change Is Hard.
In short, strategies should be tailored to the sub-culture of the specific individuals involved. For example, different approaches would be required for physicians, floor nurses, and ED nurses. Short-term improvements can be managed using a Plan-Do-Act-Check process, with self-assessment and reporting by local staff to the management team. Sustained improvements should be reflected in the dashboard.
Track and reconcile health plan incentive and penalty actions. Insurers’ systems also are evolving, but they should not be given free rein to set the providers’ compensation. It is important for finance leaders to double check the math for any health plan incentive or penalty assessments prior to accepting the settlement. The dashboard and the underlying database provide the foundation to facilitate this step.
Establish a multidisciplinary contract review team. Responding effectively to the challenge of pay-for-performance contracts requires a high degree of integration across departments and functions. All too often, the hospital finance team is charged with contract review in a process that is far removed from the operations team, which must live with the impact of the contract language. The remedy is to form a multidisciplinary contract review team that includes the following representatives and role:
- Financial analysis and decision support
- Case management, utilization review (UR), and quality
- Hospital admissions
- Strategic planning and budgeting
Each of these areas has an important role to play in the overall process. Financial analysts and decision support should be charged with setting up and managing contract models. IT staff can assist with software required to support the models and with providing access to data, which often will be captured in the clinical system and not in accounting or revenue cycle systems. Meanwhile, case management, UR, and quality staff will be the ones charged with performing real-time monitoring of patient processes, and nursing and other clinicians often may be responsible for collecting data for improving performance on metrics.
Hospital admissions department representatives should be included because admissions will need to know how to tag the patient records during registration so that these data can be extracted into the model. And billing personnel should be included because billers will be receiving payments and applying contractual adjustments to accounts that may not accurately reflect the final value of the claim.
Accounting staff play an important role because this area must accrue the monthly contractual adjustment, and strategic planning and budgeting staff will need to update their models to incorporate the impact of these new measures—starting with net revenues, but also considering changes in costs and process design. The strategic planning representative also can help align opportunities with overall organizational plans and make sure that necessary complex or systemwide changes are incorporated into the strategic plan, where they will have executive visibility.
Contract language should be reviewed by all members of the team with an eye to understanding the likely impact of proposed terms on their areas of operation. The team should encourage standardization among health plans in metric selection and computational formulas.
Incorporate contract risk assessment and valuation in monthly financials, budgets, strategy, and planning activities. This is the final stage of the project, requiring the ability to produce dashboards that show the current status and trends for each health plan and metric.
A Leading Role for Finance
The addition of pay for performance in hospital contracts has impacts beyond revenue cycle, and offers an unprecedented opportunity for financial analysts and decision support to apply their skills to optimize revenue, maintain GAAP compliance, and support clinical personnel in making positive changes to patient outcomes. In every hospital and health system embarking on a risk contract, the finance department, guided by the finance leader and backed by IT, can play a pivotal role in ensuring the organization’s long-term growth and success by applying the tools and skills to track and manage contract performance.
Paul Selivanoff, CPA, is principal and lead consultant for Simply Better Outcomes, Lincoln, Neb.
a. Kerns, C., Managing Financial Performance Under Risk , Advisory Board Company, 2013.
b. The author estimates that there currently more than 1,200 metrics in active use within the industry.