Your home health back office, expertly managed. From India.

Before PDGM, the Low Utilization Payment Adjustment was simple to remember: fewer than five visits in a 60-day episode and payment converted to a per-visit rate. One number, every episode. That is no longer how it works, and the change quietly costs agencies real money every month.

What actually changed

The Patient-Driven Groupings Model, effective January 2020, replaced the 60-day episode with two 30-day payment periods and replaced the flat LUPA threshold with a variable threshold that differs by case-mix group. Depending on the group a period falls into, the threshold sits somewhere in the range of two to six visits.

That means there is no single number your schedulers can memorise. The threshold is a property of the period, and it is knowable at start of care, but only if someone calculates it and puts it somewhere the scheduling team will actually see.

Where the case-mix group comes from

Each 30-day period is assigned to a group based on five factors. Together these produce the case-mix group that determines both the payment weight and the LUPA threshold:

Two of those five come straight from documentation quality rather than clinical reality. A primary diagnosis that does not map cleanly to the intended clinical grouping, or secondary diagnoses left uncaptured, will move the period into a different group, and with it, a different threshold and a different payment weight.

Why avoidable LUPAs happen

In our experience running back-office operations for home health agencies, most LUPAs are not clinical decisions. They are process failures, and they cluster around a short list of causes:

• Missed visits that never get rescheduled. A single no-access visit late in a period can tip a low-threshold period below the line.
• Patient refusals recorded but not escalated. The visit is documented as refused; nobody recalculates whether the period is now at risk.
• Hospitalisation near the period boundary. Visits stop, the period closes short, and the transfer paperwork obscures what happened to the count.
• Holiday and weekend scheduling gaps. Predictable, and therefore preventable, but only if someone is looking a week ahead.
• Late recertification. The new period opens without a schedule built against its threshold.
• Thresholds invisible to schedulers. The most common root cause of all: the number exists in the payment logic but not on the screen where visits are booked.

A monitoring cadence that works

You do not need new software to control this. You need the threshold surfaced at the right three moments:

1. At start of care or recertification. Determine the period's threshold and attach it to the schedule, not just the billing record. Build the visit plan with at least one visit of headroom above the threshold.
2. Around day 18 to 22. Compare completed plus scheduled visits against the threshold. Any period without headroom gets escalated to the clinical manager while there is still time to add a visit.
3. Two to three days before the period closes. Final sweep on anything still at or below the line. This is the last point at which the outcome can change.

What good looks like

Agencies that control LUPAs well tend to share three habits. They treat the threshold as a scheduling input rather than a billing output. They review at-risk periods on a fixed weekly rhythm instead of reacting when the remittance arrives. And they track avoidable versus unavoidable LUPAs separately, so the genuinely clinical ones do not hide the process failures.

None of that requires clinical judgement. It requires somebody with time, a checklist and access to the schedule, which is precisely the kind of work that does not need to sit with your nurses or your billing lead.

General operational guidance, not legal, coding or billing advice. CMS rules, thresholds and rates change, so verify against the current Home Health Prospective Payment System final rule and your MAC’s guidance before acting.

Almost every agency we work with can tell us their denial rate. Far fewer can tell us their top three denial reasons by dollar value. That gap is where the money is, because denials are rarely a billing department failure. They are usually a front-end failure that only becomes visible once the claim comes back.

Where home health denials actually originate

Sort your remittances by reason code and the same causes tend to surface:

• Face-to-face encounter documentation. Missing entirely, outside the required window, or present but not evidencing that the encounter related to the primary reason for home health.
• Physician orders and signatures. Verbal orders never reduced to writing, plans of care unsigned at the time of billing, signature dates that post-date the services.
• Eligibility and coverage. Homebound status not documented in a way a reviewer accepts, skilled need not established, or the patient covered by a Medicare Advantage plan nobody verified.
• Notice of Admission timing. The NOA replaced the RAP and must be filed promptly after start of care; late filing carries a payment reduction that scales with the delay.
• Medical necessity. Visit notes that record tasks performed but never articulate why skilled care was required.
• Authorization. Managed care periods delivered without auth in place, or auth exhausted mid-period without anyone noticing.
• Overlaps and duplicates. Another provider's episode overlapping yours, or inpatient stays not reconciled against your billing period.
• Coding and grouping. Primary diagnosis that does not support the clinical grouping, or a sequence that changes the payment group.

Fix it upstream, not downstream

1. Make intake do the qualifying

Every referral should clear a short, non-negotiable checklist before a clinician is scheduled: payer verified including Medicare Advantage plan specifics, benefit period confirmed, homebound rationale captured in the referrer's own words, face-to-face encounter identified or explicitly pending with an owner and a date.

The point is not paperwork. It is that a referral which cannot clear that list is a referral you are about to deliver care against and not get paid for.

2. Track orders like receivables

Unsigned orders are aged accounts receivable that have not been recognised yet. Maintain an aging report on outstanding signatures, with escalation at defined intervals rather than a monthly chase. Agencies that hold signature aging under a fortnight see materially fewer documentation denials.

3. Treat NOA timing as a hard operational deadline

This one is unusual because it is entirely within your control and entirely mechanical. Assign a named owner, monitor it daily rather than weekly, and measure the interval from start of care to NOA acceptance. Any drift here is pure avoidable loss.

4. Give reviewers the reasoning, not just the tasks

Documentation denials for medical necessity almost always come down to notes that describe what was done without establishing why a skilled clinician was required. Build that expectation into your QA review, and audit a sample against the standard a reviewer would apply rather than the standard your team is used to.

5. Work denials by root cause, in cohorts

Appealing denials one at a time recovers cash. Grouping them by reason code and fixing the originating process stops the next cohort. Both matter, but only the second one compounds.

The metrics worth watching

General operational guidance, not legal, coding or billing advice. CMS rules, thresholds and rates change, so verify against the current Home Health Prospective Payment System final rule and your MAC’s guidance before acting.

OASIS has always driven quality reporting. Under PDGM it also drives payment, through the functional impairment level and the comorbidity adjustment. That makes assessment accuracy a revenue function, not just a compliance one, and it means an assessment error is a payment error that nobody in billing can see.

The two places OASIS moves money

Of the five factors that determine a period's case-mix group, two come out of the assessment and the diagnoses reported alongside it:

• Functional impairment level: low, medium or high, derived from specific functional items on the assessment. Mis-scoring these shifts the payment group.
• Comorbidity adjustment: none, low or high, based on reported secondary diagnoses. Uncaptured comorbidities simply forfeit the adjustment.

Neither error produces a denial. The claim pays, just at a lower weight than the patient's actual condition supports. That is why these losses persist: nothing bounces back to tell you they happened.

Errors we see most often

Functional items scored on best performance

Functional scoring is meant to reflect the patient's usual status over the assessment timeframe, not their best single demonstration. Clinicians who observe a patient managing a task once, on a good day, often score to that. It is an honest mistake that consistently understates impairment.

Assessment that contradicts the visit notes

An assessment recording independence in an area where subsequent notes document assistance is the single most common finding in audit. It undermines both the payment group and the agency's credibility on review, because a reviewer reads them together.

Primary diagnosis that does not support the intended clinical grouping

The primary diagnosis reported on the claim assigns the clinical grouping. A diagnosis that is clinically defensible but not the reason home health was ordered can land the period in a grouping that does not reflect the care being delivered.

Secondary diagnoses left off

Comorbidity capture is the most commonly forfeited adjustment we encounter. The conditions are usually documented somewhere in the record, whether the referral, the hospital summary or the medication list, and simply never make it onto the claim.

Timing and date errors

Assessment completion outside the required window, or an incorrect assessment-completion date, creates problems that surface later and are awkward to correct. These are clerical and entirely preventable by review.

A QA review that is worth the time

A pre-submission review does not need to re-do the assessment. It needs to answer six questions:

1. Are the functional items internally consistent, and consistent with the visit notes and the referral?
2. Does the primary diagnosis map to the clinical grouping the care plan actually reflects?
3. Have all supportable secondary diagnoses been captured from the available record?
4. Are dates within the required windows, and does the completion date match the record?
5. Are physician orders present and consistent with the plan of care?
6. Would a reviewer reading only this documentation reach the same conclusions the assessment asserts?

That last question is the one that separates a compliance check from a useful review. Your team knows the patient; a reviewer knows only the file.

Make it a rhythm, not a rescue

Review before submission wherever possible, because correction after the fact is significantly more expensive and sometimes not available. Where volume makes full review impractical, stratify: review every start of care and resumption of care, plus a sample of recertifications, and widen the sample for any clinician whose error rate is drifting.

Track findings by category rather than just counting them. Six functional-scoring errors from one clinician is a training conversation. Six from six clinicians is a process or guidance problem.

General operational guidance, not legal, coding or billing advice. CMS rules, thresholds and rates change, so verify against the current Home Health Prospective Payment System final rule and your MAC’s guidance before acting.

Outsourcing your administrative backbone is a risk decision before it is a cost decision. The rate is the easiest thing to compare and the least likely to determine whether the engagement works. These are the questions we think you should ask any prospective partner, including us, and what a substantive answer looks like.

Compliance and data

1. Will you sign a Business Associate Agreement?

If the answer is anything other than an immediate yes, stop. Most US agencies cannot lawfully route protected health information to a partner without one. A capable partner treats the BAA and the security review as a normal part of onboarding, not an obstacle to be negotiated down.

2. Where does our data physically live, and what leaves your building?

You are looking for a specific architecture, not reassurance. Access through VPN into your systems with no local storage is a materially different risk profile from data being copied into the vendor's own environment. Ask which one it is, and ask how it is enforced rather than merely stated.

3. Who has access, and how is it removed?

Named individuals or a shared pool? How quickly is access revoked when someone leaves the account? Is multi-factor authentication enforced on every client-facing system, and can they evidence it?

4. What happens if you have a breach?

Notification timeline, who contacts whom, and what their incident response plan actually says. A partner who has never thought about this in writing has not thought about it.

Capability

5. Is home health a vertical for you, or the business?

A generalist BPO will learn PDGM on your account, at your expense, and their best people will be rotated to whichever vertical is growing fastest. Ask what proportion of their revenue comes from home health, and ask them to explain a LUPA threshold without notes.

6. Which EMRs have your team actually worked in?

Not "we can learn any system". Ask which platforms, for how long, and doing what. Homecare Homebase, WellSky, MatrixCare, Axxess and AlayaCare are different enough that real experience shows up immediately in ramp time.

7. Who owns quality, and how is it measured?

Ask whether QA sits inside the delivery team or independently of it, what sample rate is reviewed, and which metrics are reported to you. "We have a QA process" is not an answer. A sample rate and a reported error rate is.

8. What are the SLAs, and what happens when they are missed?

Turnaround times, accuracy targets, escalation response. Then the more revealing question: what is the consequence of a miss? Commitments with no consequence are aspirations.

Continuity

9. What is your turnover on a client account?

This is the question most likely to be deflected, and one of the most important. Every departure costs you re-training you did not budget for. Ask about average tenure on an account and how knowledge is retained when someone does leave.

10. How fast can you scale, in both directions?

Census moves. Ask how long it takes to add five people, and then the part nobody volunteers: what happens commercially if you need to reduce capacity for a quarter.

11. What does the pilot look like, and what does it cost us to stop?

A partner confident in their work will offer to run parallel to your existing team on live volume, with no long-term commitment, so you can compare output directly. If the only way to evaluate them is to commit first, that tells you something.

12. If this ends, how do we get our operation back?

Ask about the exit before you sign the entry. Documented processes, knowledge transfer, notice period, and what happens to work in progress. A partner who has a clean answer here is a partner who expects to be judged on performance rather than lock-in.

How we answer these

We sign a BAA with every client as a matter of course. Access is through VPN into your systems, with no patient data stored locally on any employee device and MFA enforced on every client-facing system. Home health is not one of our verticals. It is the whole business, and we already run live production operations for licensed California agencies. Every engagement opens with a two-to-four week parallel pilot at no long-term obligation, because we would rather be judged on output than on a proposal.

On the questions above where the honest answer is "it depends on scope", such as SLAs, ramp timelines and commercial terms, we would rather work through them with you on a call than publish a number that turns out not to apply to your situation.

General operational guidance, not legal, coding or billing advice. CMS rules, thresholds and rates change, so verify against the current Home Health Prospective Payment System final rule and your MAC’s guidance before acting.