Fig. 1 Orthopaedic Clinical Episodes in BPCI Advanced Source: Centers for Medicare and Medicaid Services (CMS)


Published 5/1/2018
Shreyasi Deb, PhD, MBA

A Closer Look at BPCI Advanced

On Jan. 9, 2018, the U.S. Centers for Medicare & Medicaid Services (CMS) Center for Medicare and Medicaid Innovation (Innovation Center) announced the Bundled Payments for Care Improvement Advanced (BPCI Advanced) model, a new voluntary episode-based payment model that will test bundled payments for 32 clinical episodes (29 inpatient and three outpatient episodes), many of which are relevant for orthopaedic surgeons (Fig. 1).

The model, which begins on Oct. 1, 2018, will run through Dec. 31, 2023. CMS will provide a second application opportunity in January 2020. The application portal closed on March 12, 2018, and CMS expects to provide target prices to participants by May 2018 and participation agreements by August 2018.

BPCI Advanced is a voluntary model with a single retrospective bundled payment and one risk track, with a 90-day episode duration. At the outset, this model qualifies as an Advanced Alternative Payment Model (APM) in the Quality Payment Program. However, the first Qualified Participant (QP) determination will be on March 31, 2019. BPCI Advanced will operate under a total-cost-of-care concept, in which the total Medicare fee-for-service spending on all items and services furnished to a beneficiary during a clinical episode, including outlier payments, will be part of the clinical episode expenditures for purposes of the target price and reconciliation calculations, unless specifically excluded.

BPCI Advanced is a nationwide model, and a participant is defined as an entity that enters into a Participation Agreement with CMS to participate in the model. BPCI Advanced will require downside financial risk of all participants from the outset. A convener participant is a participant that brings together multiple downstream entities, referred to as Episode Initiators (EIs), and facilitates coordination among its EIs and bears and apportions financial risk under the BPCI Advanced model. A non-convener participant is a participant that is an EI that does not bear risk on behalf of other downstream EIs.

Acute care hospitals (ACHs) and physician group practices (PGPs) can participate as convener or nonconvener participants. Eligible entities that are either Medicare-enrolled or not Medicare-enrolled providers or suppliers can all participate as conveners. However, critical access hospitals (CAHs), prospective payment system (PPS)-exempt cancer hospitals, inpatient psychiatric facilities, hospitals in Maryland, hospitals in the Rural Community Hospital Demonstration, and hospitals in the Pennsylvania Rural Health model are excluded from the model.

BPCI Advanced will not use time-based precedence rules, and the hierarchy for attribution of a clinical episode among different types of EIs is pre-determined at this announcement stage (Fig. 2).

Quality measures
CMS has selected seven quality measures for the BPCI Advanced Model. Two of them, All-cause Hospital Readmission Measure (National Quality Forum [NQF] #1789) and Advance Care Plan (NQF #0326), will be required for all clinical episodes. The remaining five quality measures will only apply to select clinical episodes; they are as follows:

  • Perioperative Care: Selection of Prophylactic Antibiotic: First or Second-Generation Cephalosporin (NQF #0268)
  • Hospital-Level Risk-Standardized Complication Rate (RSCR) Following Elective Primary Total Hip Arthroplasty (THA) and/or Total Knee Arthroplasty (TKA) (NQF #1550)
  • Hospital 30-Day, All-Cause, Risk-Standardized Mortality Rate (RSMR) Following Coronary Artery Bypass Graft Surgery (NQF #2558)
  • Excess Days in Acute Care after Hospitalization for Acute Myocardial Infarction (NQF #2881)
  • AHRQ Patient Safety Indicators (PSI 90)
Fig. 1 Orthopaedic Clinical Episodes in BPCI Advanced Source: Centers for Medicare and Medicaid Services (CMS)
Fig. 2 Precedence Map for Clinical Episode Attribution in BPCI Advanced Source: Centers for Medicare and Medicaid Services (CMS)

AAOS’ concerns
The Academy has long requested a voluntary payment model that will qualify as an Advanced Alternative Payment Model (APM). We are therefore pleased that BPCI Advanced is a voluntary model that is available nationally, allows for physician leadership, and is an Advanced APM. BPCI Advanced includes many orthopaedic procedures, enabling surgeons and their patients to participate in musculoskeletal care redesign and improved outcomes. However, we have serious concerns with the model’s design and application requirements.

Our primary concern is the precedence of Comprehensive Care for Joint Replacement (CJR) over BPCI Advanced for lower extremity joint replacement (LEJR) episodes. This will limit the ability of independent physicians to participate in BPCI Advanced and thereby their ability to participate in an Advanced APM. We have reviewed data from a PGP convener of orthopaedic practices in BPCI Classic undertaking LEJR cases at CJR hospitals from Q3 2016 to Q2 2017. The comparison data are from BPCI and CJR cases limited to Medicare Severity Diagnosis Related Groups (MS-DRG) 469 and 470 performed in CJR hospitals. These data consist of 15 PGPs performing 6,340 LEJR episodes in 40 CJR hospitals. Collectively, these PGPs reduced costs by $12.6 million over the study period.

To examine the impact of the BPCI Advanced precedence policy, we reviewed these same PGPs and only the CJR hospitals with positive gains in the current CJR program (12 PGPs and 20 CJR hospitals). We did this to compare savings between the two groups based on current performance in their respective programs. In this hypothetical example, the PGPs would save $4.9 million on 2,635 episodes. Under the current policy for BPCI Advanced, all these cases would move to the 20 CJR hospitals, and based on their current case rate, we estimate the CJR hospitals would save $2.8 million. It appears that given their current performance, these PGPs would outperform the 20 CJR hospitals by a margin of approximately 2:1.

Given the excellent performance of PGPs in BPCI, many orthopaedic surgeons are questioning why CMS found it necessary to structure BPCI Advanced in a way that severely limits a PGP physician’s ability to manage all their Medicare patients’ care for LEJR episodes. Further, this policy of precedence disrupts the physician-patient relationship in the affected geographical areas, as the coordination of the Medicare beneficiary’s LEJR care will now fall to the CJR hospital, regardless of the care the beneficiary’s physician may have provided previously in BPCI. This is likely to leave beneficiaries that experienced BPCI previously confused as to why their current physician cannot manage their next LEJR. Also, pushing these well-prepared BPCI-experienced participants into the CJR model may ultimately contaminate CJR evaluation results. Moreover, there is legal precedence in U.S. Supreme Court’s interpretation of the Administrative Procedure Act that CMS may have to institute this change in precedence policy via rule-making and not via the current application process.

Another concern is with the impact of CMS’ decision to take TKA out of the Medicare Inpatient Only (IPO) list. While some hospitals understand the intent of CMS’ rulemaking on this topic, a number of hospitals are apparently directing TKA patients to a default outpatient status that is clearly in contradiction to CMS’ stated positions. Contrary to CMS’ expectation that most TKA cases will not be performed in the outpatient setting, more and more TKA cases are being pushed to outpatient facilities, thereby creating a situation in which most medically complex, high-comorbidity patients will remain inpatient. This change in patient mix has significant implications for BPCI Classic, BPCI Advanced, and CJR models. One solution for this issue is for CMS to allow both inpatient and outpatient LEJR bundles in BPCI Advanced. Moreover, in light of this precedence policy, BPCI Advanced and CJR must have updated target pricing from CMS.

Wrongly defaulting TKA cases to the outpatient setting is especially concerning for surgeons and patients in Medicare Advantage plans across the country. The Academy has heard from several surgeons across teaching hospitals, community hospitals, and urban and rural hospitals that Medicare Advantage plans are denying claims for TKA procedures not performed in the outpatient setting.

The AAOS Office of Government Relations has participated in several discussions with CMS staff who oversee the varied aspects of the issues enumerated above. Although the CJR precedence requirements were established via rule making and will have to be changed by the same process, education of Medicare providers and Medicare Advantage plans can happen immediately. CMS staff across these programs are aware of the problems and will address them concurrently.

Shreyasi Deb, PhD, MBA, is senior manager, health policy in the AAOS Office of Government Relations.