The Episode Design Space

The introductory course defined an episode as a trigger, a window, and a set of included services. That is the right starting frame. Working inside a real episode model means understanding that each of those is a separate design decision, and that the choices interact.

The four levers

The trigger determines which patients enter. TEAM triggers on five surgical categories: “lower extremity joint replacement, surgical hip femur fracture treatment, spinal fusion, coronary artery bypass graft, and major bowel procedures.” A trigger can be an inpatient admission, an outpatient procedure, or a diagnosis. CJR shows how this evolves: after knee and hip replacement came off the inpatient-only list, CMS “changed the definition of an ‘episode of care’, beginning in PY6, to include outpatient (OP) procedures for TKAs (OP TKAs) and THAs (OP THAs), in addition to inpatient procedures.” A trigger defined only by setting becomes obsolete when the setting shifts.

The window determines how long accountability lasts. CJR runs 90 days: “the episode of care continues for 90 days following discharge from the inpatient hospitalization or the date of the outpatient procedure.” TEAM runs 30 days from surgery through post-hospitalization.

The inclusion rules determine what counts. CJR includes, “with few exceptions, all related items and services paid under Medicare Part A and Part B,” with published lists of excluded MS-DRGs and diagnosis codes. Every exclusion is a judgment about what the accountable party could not reasonably have influenced.

The accountable party determines who bears the result. In CJR and TEAM it is the hospital. In other designs it has been a physician group or a convener organization, and the choice changes behavior because hospitals and surgeons control different decisions.

Why the window length is the sharpest lever

Window	What falls inside	What it drives
30 days	The surgery, discharge, the initial post-acute setting, early readmissions	Discharge planning and the first transition
90 days	All of the above plus extended rehabilitation, later readmissions, complications surfacing weeks out	Full post-acute management and longer follow-up

Lengthening the window adds savings opportunity and adds risk for events the hospital influences less and less as time passes. Shortening it does the reverse.

Worth remembering: a shorter window is not simply a weaker version of a longer one. It changes what the model is asking for. A 90-day window asks a hospital to manage a recovery. A 30-day window asks it to manage a discharge. Those require different capabilities, different partners, and different investments, and a hospital that built a 90-day post-acute network for CJR is not automatically well-positioned under a 30-day design. When a model’s window changes, the operating plan has to change with it.

The overlap problem

A patient in a bundled episode may also be attributed to an ACO, and both arrangements will count the same spending. Every episode model therefore needs precedence rules deciding whose result the spending lands in. This is not a technicality: it determines whether two organizations in the same market are pulling in the same direction or quietly competing for credit on the same dollars. Module 4 returns to it.

Key takeaways

• Trigger, window, inclusion rules, and accountable party are four independent design choices.
• Triggers defined by care setting go stale when procedures migrate, as CJR’s did.
• CJR uses 90 days post-discharge; TEAM uses 30 days from surgery through post-hospitalization.
• Overlap rules between episodes and population contracts determine who gets credit for the same spending.

Sources

• CMS, Transforming Episode Accountability Model (TEAM) (opens in a new tab)
• CMS, Comprehensive Care for Joint Replacement Model (opens in a new tab)