The four CMS levels of hospice care — Routine Home Care, Continuous Home Care, General Inpatient Care, and Respite — and the clinical criteria that govern moving between them
Routine Home Care (RHC) is the level of care under which the overwhelming majority of hospice patients receive services for the overwhelming majority of their hospice days. It is intermittent, not continuous — the interdisciplinary team visits on a schedule set by the plan of care, and the patient or family manages symptoms between visits with medications, equipment, and 24/7 telephone access to an on-call nurse. Understanding RHC is the foundation for understanding when and why a patient should ever leave it.
RHC is the default, presumed level of hospice care under the Medicare Hospice Benefit's Conditions of Participation (42 CFR §418.204). "Home" is defined broadly — it can be a private residence, an assisted living facility, or a nursing facility (with room-and-board billed separately to the patient or Medicaid). What defines RHC is not the location but the intensity: care is intermittent, delivered by a visiting interdisciplinary team (RN case manager, hospice aide, social worker, chaplain, volunteer, and physician oversight), with the patient or an informal caregiver responsible for day-to-day symptom management between visits.
A typical RHC plan of care might include an RN visit one to three times per week (more often in the final days of life, less often when stable), a hospice aide for bathing and personal care two to three times per week, a social worker visit monthly or as needed, and as-needed (PRN) visits triggered by a phone call to the 24-hour triage line. The hospice supplies durable medical equipment (hospital bed, wheelchair, bedside commode), all medications related to the terminal diagnosis, and incontinence and wound-care supplies — all bundled into the RHC per-diem rate regardless of how many visits actually occur that day.
Before any level-of-care discussion is possible, the patient must first be hospice-eligible: two physicians (the hospice medical director and the attending physician, if there is one) must certify that the patient has a life expectancy of six months or less if the disease runs its normal course, based on clinical indicators of decline (functional status, weight loss, disease-specific markers such as ejection fraction, FEV1, or Palliative Performance Scale score). Recertification occurs at defined intervals — two 90-day periods followed by unlimited 60-day periods — and after the third certification a face-to-face encounter with a hospice physician or nurse practitioner is required to document continued eligibility.
Once enrolled, RHC is the presumed level unless the patient meets specific criteria for one of the other three levels. The clinical logic is simple: as long as symptoms are controlled with the current medication regimen and visit schedule, and the informal caregiving structure at home is intact, there is no clinical justification — and no reimbursement justification — for a higher, more expensive level of care. Roughly 98% of hospice days nationally are billed as RHC, which is precisely why the other three levels draw intense programmatic and audit scrutiny: they are the exception, and payers expect exception-level documentation to support them.
Continuous Home Care (CHC) exists for exactly one purpose: to manage a brief, acute symptom crisis in the patient's own home with intensive skilled nursing, so that the patient does not have to be transferred to an inpatient setting. It is the least-used of the four levels nationally and the most operationally demanding — a hospice must be able to mobilize eight or more hours of predominantly nursing care within hours of the crisis being identified, something few community hospices can staff reliably around the clock.
Under 42 CFR §418.204, CHC is furnished only during a period of crisis to maintain the patient at home, and it must consist predominantly of skilled nursing care furnished by an RN or LPN, with hospice aide or homemaker services allowed to supplement but never to substitute for the nursing component. The federal minimum is at least 8 hours of care in a 24-hour period, and more than half of those hours must be nursing care (not aide time) — a threshold CMS enforces strictly because CHC is billed hourly and is dramatically more expensive than RHC.
A "period of crisis" is not defined by diagnosis but by symptom trajectory: uncontrolled pain, air hunger and severe dyspnea, terminal agitation or delirium, intractable nausea and vomiting, uncontrolled seizures, or a hemorrhage — anything that represents an acute departure from the patient's baseline and that the IDG judges can plausibly be brought back under control at home with intensive nursing rather than requiring transfer to an inpatient facility. If eight hours cannot be safely staffed, or if the crisis clearly requires interventions unavailable at home (IV medication titration under close monitoring, procedures, imaging), the correct level is GIP, not a partial or under-resourced CHC episode.
Because CHC requires same-day mobilization of continuous nursing coverage, most hospices maintain a dedicated on-call CHC nursing pool or contract with a per-diem staffing agency. The triggering event is typically an after-hours or urgent phone call from the caregiver reporting a crisis; the triage RN performs a telephone or in-person assessment, and if criteria are met, the hospice dispatches a nurse to begin continuous or near-continuous shifts (commonly split into two 12-hour or three 8-hour blocks to satisfy labor rules while maintaining coverage).
CHC is deliberately short-lived. Every shift, the assigned nurse documents the specific interventions performed, the patient's response, and whether the 8-hour predominantly-nursing threshold was actually met that day — because CMS pays CHC by the hour and will deny or recoup days that fall short of the threshold or that are not clearly crisis-driven. As soon as the symptom is controlled — often within 24 to 48 hours — the IDG downgrades the patient back to RHC. If instead the crisis intensifies or proves unmanageable at home, the correct move is to transfer to GIP rather than attempting to sustain CHC indefinitely; CHC is not meant to be a long-term private-duty nursing benefit, and prolonged CHC episodes are a recurring audit finding.
General Inpatient Care (GIP) is hospice's highest-intensity level: acute, hospital-level symptom management delivered in a Medicare-certified hospital, skilled nursing facility, or freestanding hospice inpatient unit, for pain or symptom crises that genuinely cannot be brought under control in the home environment — even with CHC support. GIP is also the level under the heaviest compliance spotlight, because its per-diem rate is several times the RHC rate and because "custodial" or prolonged GIP stays unsupported by an active symptom crisis are among the most common findings in OIG and MAC audits.
GIP is justified when the level of symptom control required exceeds what can be safely and effectively delivered at home — even with continuous nursing. Classic indications include pain crises requiring rapid IV opioid titration with close monitoring, uncontrolled nausea and vomiting causing dehydration, respiratory distress requiring more intensive intervention or monitoring than the home allows, acute delirium or terminal agitation posing a safety risk, wound care or procedures beyond what home nursing can perform, or a caregiver-imposed limitation where no safe home crisis-management option exists despite CHC being clinically appropriate.
Critically, GIP is a symptom-management admission, not a general hospitalization and not a placement solution. A patient admitted to GIP for a pain crisis must have daily documentation showing the crisis is still active and the level of care still necessary; once symptoms stabilize to a level manageable at home or in a lower level of care, continued GIP is no longer covered and the patient must be discharged to RHC (or Respite, if the barrier is now caregiver capacity rather than symptom control) — even if the family would prefer the patient remain inpatient.
Because GIP reimburses at four to six times the RHC rate, it has been a recurring target of OIG reports and Medicare Administrative Contractor (MAC) audits going back over a decade. Common findings include GIP stays that continue well past symptom stabilization (sometimes explained by the finding of a nurse's note reading "patient comfortable" for several consecutive days while GIP billing continued), GIP used as a substitute for nursing-facility placement or for family respite rather than genuine symptom crisis, and inadequate physician documentation connecting the GIP level to a specific, ongoing, unresolved symptom.
Best-practice hospices now require daily IDG or physician review of every GIP day with an explicit clinical justification note, target length-of-stay flags (commonly 5 days) that trigger a formal utilization review, and a discharge-planning conversation that begins on day one of the GIP stay rather than after the crisis has already resolved. National median GIP length of stay runs around three to five days; stays extending well beyond a week without a clearly escalating or unresolved crisis are the pattern auditors look for first.
A useful audit heuristic used by hospice compliance teams: if you removed the GIP billing code and just read the nursing and physician notes, would a reasonable reviewer conclude a symptom crisis serious enough to require inpatient management was actively occurring that day? If the notes instead describe a stable, comfortable patient, GIP is very likely no longer justified — regardless of how unstable the situation looked on the day of admission.
Respite care stands apart from the other three levels of hospice care because its trigger is not the patient's symptom burden at all — it is the exhaustion, illness, or temporary unavailability of the informal caregiver upon whom home-based hospice depends. Up to five consecutive days of inpatient care are covered per respite episode, in a Medicare-certified facility with sufficient staff to provide 24-hour nursing care, giving the family a planned, bounded break before returning the patient home.
Respite care is available to any hospice patient whose caregiver needs planned relief — most commonly because of caregiver fatigue, a caregiver's own medical procedure or illness, a family event the caregiver must attend, or simply the cumulative burden of round-the-clock caregiving for a patient with a terminal illness. Unlike GIP, there is no requirement that the patient be in symptom crisis; the patient can be entirely stable. What must be documented is the caregiving need: why the informal support system cannot continue uninterrupted at this moment.
Respite is delivered in a Medicare-certified hospital, skilled nursing facility, or hospice inpatient unit with sufficient round-the-clock staff, and Medicare limits each respite stay to five consecutive days; on the sixth day the patient reverts to whatever level of care (usually RHC) is otherwise appropriate, and the hospice — not Medicare — becomes financially responsible for continued inpatient days beyond the five-day cap unless a new, separate justification (such as an intervening symptom crisis that now independently qualifies for GIP) applies. There is no statutory limit on how many times a patient can use respite over the course of a hospice election, but each episode must be independently justified and IDG-documented, and clusters of frequent, closely-spaced respite stays for the same patient tend to draw the same audit attention as prolonged GIP stays.
Caregiver burden is one of the strongest predictors of early hospice disenrollment and of caregiver bereavement complications, and respite exists explicitly to protect the sustainability of the home-care model that RHC depends upon. A caregiver who has gone days without uninterrupted sleep, who is managing their own chronic illness, or who is approaching a breaking point is a legitimate and expected reason to invoke respite — the IDG social worker and RN case manager are typically the ones who first flag rising caregiver strain during routine visits, using informal caregiver-burden screening (sleep disruption, mood, physical symptoms, expressed overwhelm) well before a crisis forces the issue.
Because respite is caregiver-driven rather than patient-driven, it is sometimes confused with GIP by families and even by less-experienced clinicians ("the patient needs to go to the hospital for a few days"), but the documentation, billing code, and clinical justification are entirely different, and using GIP to accomplish what is really a respite need (or vice versa) is a compliance error. Distinguishing the two accurately — is this an unresolved symptom crisis, or is this a caregiver in need of relief while the patient remains stable — is one of the most common judgment calls IDG teams make in practice.
The clinical skill that ties all four levels together is not knowing the definitions in isolation but continuously and accurately matching the real-time situation — symptom severity, setting, and caregiver capacity — to the least intensive level of care that can safely meet the need, and then promptly stepping the patient back down as soon as the crisis resolves. Getting this wrong in either direction has consequences: under-escalating risks unmanaged suffering and unwanted hospitalization; over-escalating (or failing to de-escalate) risks improper payment, audit exposure, and unnecessary disruption for the patient and family.
When a patient's situation changes, the IDG works through a consistent sequence of questions. First: is there an acute, unresolved symptom crisis (pain, dyspnea, agitation, nausea, seizure, hemorrhage) that represents a clear departure from baseline? If no, and the issue is caregiver capacity rather than the patient's symptoms, respite is the relevant option, assuming a bed is available and the family agrees to a planned admission. If yes, the next question is whether the crisis can plausibly be brought under control at home with intensive nursing support within a short window — if so, and 8+ hours of predominantly skilled nursing can actually be staffed that day, CHC is appropriate. If the crisis requires interventions, monitoring, or an environment beyond what CHC can provide — or if adequate CHC staffing simply is not available — GIP is the correct level.
Throughout, the default assumption is RHC: every escalation must be affirmatively justified by a specific, current clinical (or, for respite, caregiving) need, and every day at an elevated level of care must be re-justified, not simply carried forward from the day before. This is why hospices build daily or near-daily IDG check-ins into any CHC or GIP episode — the clinical picture that justified escalation yesterday may no longer justify it today.
Every level-of-care determination needs a paper trail that would make sense to someone who was not in the room: what specific symptom or caregiver circumstance triggered the change, what interventions were attempted at the prior (lower) level and why they were insufficient, what the plan is for returning to a lower level, and — for CHC and GIP specifically — contemporaneous, shift-by-shift or day-by-day notes showing the crisis is ongoing, not historical. Physician orders should accompany every level-of-care change, and the level should be reflected consistently across the plan of care, the visit notes, and the billing record; discrepancies between what is billed and what the clinical notes describe are the single most common driver of hospice payment denials and program integrity investigations.
The transition back down to RHC deserves the same discipline as the escalation itself: as soon as pain is controlled, agitation resolves, the caregiver has had their planned rest, or the acute event stabilizes, the IDG should document the resolution and revert the level of care — ideally the same day. Hospices with strong utilization review processes typically achieve this by pairing every CHC or GIP admission with an automatic reassessment trigger (for example, at 24, 48, and 72 hours) rather than waiting for someone to notice the crisis has passed, which both improves patient and family experience and keeps utilization patterns defensible under audit.
The four levels of care are not a ladder of "worse" to "better" hospice — they are four different tools for four different situations, and the mark of high-quality hospice practice is using each one for exactly as long as it is needed and no longer. A hospice whose GIP and CHC utilization looks similar to national benchmarks (roughly 1–2% and well under 1% of days, respectively) with short, well-justified stays is generally a strong signal of disciplined, patient-centered level-of-care management — not under-treatment.