Print

The PROMETHEUS Payment® Model Dividing The Pie for an Episode of Care

Karin Bierstein, JD, MPH
Vice President for Strategic Planning and Practice Affairs, ABC

When physicians, hospitals, home health agencies and other providers decide to create an Accountable Care Organization or other integrated delivery system, one major issue that will soon command attention is the distribution of patient care revenues. How will the various providers share the pie?

One model comes from the PROMETHEUS® Payment allocation system.i PROMETHEUS, a methodology developed beginning in 2004 by a team led by Alice G. Gosfield, Esq. and Franc?ois de Brantes, M.S., M.B.A. pays providers a single, risk-adjusted payment across inpatient and outpatient settings to care for a patient diagnosed with a specific condition. The payment is based on “evidence-informed case rates” (ECRs) and is theoretically equal to the resources required to provide care as recommended in well-accepted clinical guidelines. Thus the total payment for a typical episode of care, or the ECR, is equal to:

Types of services typically involved in treating the condition * Frequency * Price per service

A portion of the payment to each participating provider is withheld and, at the end of the measurement period, distributed based on provider performance on measures of clinical process, outcomes, and patient experience. A comprehensive scorecard measures those three variables (process, outcome, patient) at the level of the contracting provider, be it an individual physician, the group or the entire integrated delivery system. Seventy percent of the score is based on the performance of the contracting provider, while the other 30 percent reflects the performance of all the providers involved. The dependence on team performance for the 30 percent underlines the value of coordination of care.

Withholds to Cover Preventable Complications – Or to Distribute to the Providers If There Are No Complications

Since HHS, private payers and policy makers began to focus on “Potentially Avoidable Complications” (PACs) and the PAC subset, Hospital Acquired Conditions (HACs), PROMETHEUS has provided for the withholding of a certain percentage of the ECRs for the contingency of avoidable complications. A budgetary allowance for PACs is redistributed into each ECR and is adjusted for severity, so that the ECR for a sicker patient gets a higher PAC allowance. Currently, the PROMETHEUS system holds back roughly 50 percent of the costs of treating PACs, based on the crude estimate that 50 percent of complications are avoidable. Should complications occur, this portion of the budget serves to offset the actual costs of the corrective treatment. If the physicians and other providers can reduce or eliminate the PACs, however, they can keep the entire allowance as a bonus and significantly improve their margins per patient.ii Therein lies an important incentive to continue to bring down the number of complications.

Example. To illustrate how the payment and the contingency reserve might work, consider the example of the application of PROMETHEUS methodology to knee and hip arthroplasties.iii A group of researchers in Boston analyzed 2005-2006 claims from a database with a population of more than 4.5 million commercially insured persons. Each of the two arthroplasty Episodes of Care had (1) an inpatient facility claim and (2) an “other” grouping of claims including professional services, outpatient facility charges, pharmacy, laboratory, radiology and all other types of services. Pertinent claims from both categories were further classified either as “typical” care for the index condition or as PACs, depending on whether the claim bore a potentially avoidable complication code. All inpatient, professional and pharmacy claims for eligible cases within 30 days prior to surgery and 180 days following surgery were potentially included in the construction of the particular ECR. Eligible cases were defined by ICD-9 procedure and diagnoses codes (both for inclusion and exclusion), patient age and absence of defined conditions or major unrelated surgical procedures, as well as by continuous enrollment and complete data.

PACs for the arthroplasty analyses consisted of inpatient or outpatient claims in any of the diagnoses fields or and of claims for a procedure related to: adverse effects of drugs, overdose, poisoning, complications of implanted device, complications of surgical procedure or medical care, revision procedures, vascular catheter associated infection, septicemia, meningitis, hepatitis, fluid and electrolyte disturbances, blood incompatibility, perioperative hematoma, hemorrhage, stroke, coma, syncope, delirium, AMI, shock, cardiac arrest, air embolism, pneumonia, respiratory failure, lung complications, iatrogenic pneumothorax, tracheostomy, mechanical ventilation, acute renal failure, urinary tract infections, gastritis, ulcer, deep vein thrombosis, pulmonary embolism and decubitus ulcers.

The authors of the arthroplasty study were able to construct three different paradigm patients representing increasing levels of severity of illness and corresponding case rates and hold-backs, as shown in Table 3. The first component of the withhold is a flat 10 percent of the cost of typical care. This is repaid to the providers if they meet certain quality standards. The PAC allowance consists of a fixed fee that is the same across all levels of severity ($471 using the study claims data, or 25 percent of the overall PAC allowance divided by the 2076 cases) plus 7 percent of severity-adjusted costs for each level (7 percent is half of the actual total cost of PACS associated with hip arthroplasties, i.e., 14 percent).

Potentially Avoidable Complications as Measures of Quality

The hip and knee replacement surgery study showed that “[d]istinguishing between typical care and potentially avoidable complications (PAC) creates an opportunity to hold the system accountable for the latter while holding it harmless for the former.” Avoidance of complications as a quality target with an economic incentive makes good sense. Its financial value can be measured objectively (albeit sometimes with proxy measures). It is of high dollar value: according to the Agency for Healthcare Research and Quality, employers spend about $1.5 billion annually for potentially preventable medical errors occurring during or within 90 days following surgery. A single catastrophic, preventable complication can cost an individual hospital amounts in the six or even seven figures in uncompensated care and malpractice settlements or awards.

Avoiding negative outcomes is a major quality marker in surgical anesthesia practice. All three of the anesthesia measures available for reporting through the Physician Quality Reporting System (timely antibiotic prophylaxis, protocol for prevention of catheter-related bloodstream infection and maintenance of postoperative normothermia) are aimed at preventing surgical infections.

Many of the 26 adverse perioperative events and outcomes defined in the ASA Committee on Performance and Outcomes Management’s August 2009 Annual Report”iv (Figure 1) potentially have a measurable cost that could also be used in establishing a reserve or withhold for PACs. Caution: until satisfactory methods for risk adjustment, data analysis and trimming and other statistical techniques, and a host of other technical considerations have been addressed, these events and outcomes are not ready for use in any system that would base compensation on quality. The Committee’s list is a valuable starting point for groups assessing potential areas for clinical and improvement and cost savings in their own practices, however.

How else might we start thinking about not just the total amounts, but also the individual providers’ respective shares of reserve funds not spent on treating complications or readmission? A simple method might be to assume that physicians account for very roughly 20 percent of spending on medical care. You might substitute the proportion in your own hospital. In Table 3 (previous page), the 20 percent of the $4925 (=$985) combined total margin and PAC allowance for Patient 2 would be shared among the physicians involved in the hip arthroplasty episode—orthopedic surgeon, anesthesiologist, and perhaps the patient’s internist and other doctors who provided care during the index episode. The physicians could decide to allocate the $985 by consensus, or by a formal method such as comparing total Relative Value Units (RVUs for the anesthesiology service could be computed through a ratio of conversion factors or some other mathematical process—this is a topic for a future article).

The PROMETHEUS payment model is just one possibility, albeit a well developed method. It does have the virtue of not needing to go through a provider organization or ACO. A health plan could make a single global payment to the organization for distribution, but the PROMETHEUS model also permits each provider or physician to be compensated directly by the participating payer based on that provider’s own quality scorecard. The model can also be used within an ACO or other integrated delivery system. Although it is now more than six years old, it remains highly flexible. It is currently the focus of several pilot studies underwritten by the Robert Wood Johnson Foundation.

Quality-based payment for anesthesia services within a group, an ACO, or other more or less integrated organization is not circumscribed by any established methodologies. One alternative to the model presented above, for example, would be to start with an allocation method based on the proportion of net revenues from professional anesthesia services as compared to other physicians’ services and inpatient/medications/ supplies/OR time and other OR charges/ procedures/anesthesia.

The requirements for participation in Medicare’s future Shared Savings program as an ACO are very vague (anticipated federal regulations giving more shape to the above requirements of the Affordable Care Act had not been published as of the date that this issue of the Communique? went to print). To be eligible, an ACO must:

The future regulations will be another tool in our growing understanding of how anesthesiologists might steer and thrive in ACOs and other organizations that reward coordinated care and measurable quality achievements. We already have the PROMETHEUS payment model and the resources on the PROMETHEUS website (www.prometheuspayment.org); the data that many anesthesia groups’ and hospitals’ information systems contain; practical experience that you may already have with private sector integrated health care systems, and your creativity—as well as ours. Comments on the ideas in this article are most welcome. We hope to be working with you on ACO and other shared savings strategies in the near future.

Karin Bierstein, JD, MPH, serves as Vice President of Strategic Planning and Practice Affairs for ABC. Ms. Bierstein came to ABC from the American Society of Anesthesiologists in 2007. She concentrates on ABC’s partnerships including those with ePREOP and Surgical Directions and serves as a Medicare and healthcare reform expert. She can be reached at Karin.Bierstein@AnesthesiaLLC.com.

[i] de Brantes F, Rosenthal M., Painter M. Building a Bridge from Fragmentation to Accountability – the PROMETHEUS Pay- ment Model. N. Engl. J. Med. 2009; 261:1033-1036 (September 10, 2009).

[ii] Robert Wood Johnson Foundation, What Is PROMETHEUS Payment®? An Evidence-Informed Model For Payment Reform. Available at http://www.rwjf.org/files/research/prometheusmodeljune09.pdf . iii Rastogi A, Mohr B, Williams JO, Soobader MJ, de Brantes F. PROMETHEUS Payment Model: Application to Hip and Knee Replacement Surgery. Clin Orthop Relat Res. 467(10): 2587-2597.

[iv] American Society of Anesthesiologists, Annual August Report of Committee on Performance and Outcomes Measurement, August 23, 2009. http://aqihq.org/CPOM%20Registry%20Data%20Set.pdf.