Reconcile PHP enrollment, planned attendance, actual participation, documented services, and billing review with clear definitions and exceptions.
A PHP daily census should distinguish enrolled people, scheduled participants, actual attendance, documented services, and billing review rather than presenting them as one count. A person can be enrolled in a program without attending every scheduled day, and an attended day still needs an understandable service record.
This guide describes operational reconciliation for partial hospitalization programs. It does not define a payable day, set clinical participation criteria, or establish coverage. Use the actual program and payer instructions with authorized clinical and billing review.
PHP daily census reconciliation starts with definitions
Write down what census means in the organization. It might refer to active episodes, expected attendance, people physically present, or another operational population. A dashboard label does not establish the definition.
The Medicare partial hospitalization overview describes PHP in a Medicare coverage context. Keep that context separate from the organization's internal enrollment and staffing views. An operational census and a coverage determination answer different questions.
Name the report owner, population, date, cutoff, and exclusions. If the organization changes its definition, annotate the change rather than presenting the new total as directly comparable with earlier periods.
Build the expected population from dated episodes
Identify active program relationships for the day using the organization's approved episode process. Preserve admission, transition, and discharge dates with their meaning. A current patient profile may not accurately represent the population for a historical day.
Use fictional cases for a new admission, established participant, planned absence, program transition, and discharged episode. Ask the report owner to state the expected treatment before running the software view.
Inspect effective-date behavior. If a transition is entered later, determine how historical census results change and how earlier saved reports remain explainable. A live current-state query should not be assumed to reproduce a prior day's operational snapshot.
Reconcile the schedule to actual participation
The planned schedule identifies expected activity, not delivered care. Record actual attendance and relevant changes through the organization's approved workflow. A shared group session can occur while individual participation differs.
Test late arrival, early departure, a canceled session, and a missed day using fictional data. Ask how staff document the actual facts and how the program's clinical review process receives relevant questions. Do not use a census discrepancy alone to make a clinical conclusion.
Keep the source of truth explicit. If staff maintain both a roster and a separate daily board, define which one supplies each report and how differences are resolved. Two manually maintained lists can drift even when each team believes its own view is current.
Trace attendance to individual documentation
Identify the expected relationship between the day's activity and individual records. A program-level attendance count does not explain each person's clinical work, response, or next step. Have the clinical lead define the relevant documentation review.
Inspect a fictional attendance correction after a note exists. Determine whether the note changes, remains unchanged, or needs an authorized amendment. The daily reconciliation should identify contradictory records rather than assuming one edit updates everything.
The CMS Medicare mental health coverage booklet provides broader program context. Use the applicable instructions and organizational policies when evaluating the actual documentation, rather than treating a complete calendar as sufficient evidence.
Keep billing status as a separate layer
Ask billing staff to define the review needed for the actual services and payer arrangements. A daily census does not establish a billable day, and a reviewed charge does not prove every clinical record element is complete.
Create a ledger linking the operational activity to relevant documentation and charge statuses. The relationship may not be one-to-one. Explain legitimate differences rather than forcing the same total across all views.
| Layer | Question it answers |
|---|---|
| Enrollment | Who has an active program relationship under the defined process? |
| Schedule | Who is expected and what activity is planned? |
| Attendance | What participation was actually recorded? |
| Documentation | Which clinical records exist and what remains unresolved? |
| Billing review | Which administrative items have been reviewed under applicable instructions? |
Use the ledger to identify exceptions. A missing note, canceled appointment, duplicate record, and held charge require different owners and should not all appear as a generic census mismatch.
Assign ownership to differences
Operations can investigate scheduling and roster issues. Clinical staff determine the facts and appropriate documentation action. Billing staff review charges and payer questions. Reporting staff explain definitions and calculations. Name those routes in the worklist.
Record the exception, source identifiers, owner, next action, and outcome. Keep sensitive detail in the authorized record. A broad operational list often needs only enough information to route the work and locate the evidence.
Review unresolved items at a sustainable cadence. Some questions need immediate handling under established clinical procedures; others are ordinary administrative cleanup. The worklist should make both ownership and escalation clear without using age alone as a clinical triage rule.
Test totals and individual cases together
Create a fictional day with known expected enrollment and varied activity. Include a duplicate, a canceled appointment, a corrected attendance entry, and a program transition. Write expected relationships before the vendor demonstration.
Compare totals, then trace each person through the layers. Equal totals can hide offsetting errors. Ask a reviewer who did not attend the demonstration to explain why each difference exists and which work remains.
Save report filters, date definitions, and observed limitations in the implementation record. If the proposed workflow requires a manual reconciliation, identify its owner and expected effort in the written scope rather than assuming it will be negligible.
Ease Health publishes this guide and offers a unified CRM, EHR, and RCM platform for behavioral health. Ask vendors to demonstrate the daily ledger and exception workflow for the proposed PHP configuration. Related material includes the RCM overview, PHP-to-IOP handoff guide, and IOP roster correction guide.
The useful output is an explained daily picture of enrollment and care, with unresolved items still accountable. That gives the program more insight than a single census number whose meaning varies by department.
Related reading
Frequently Asked Questions
Is the number of enrolled PHP patients the same as daily attendance?
Not necessarily. Enrollment, planned attendance, actual participation, and documented services are different states. Define which one each report counts.
Should daily census automatically determine billable days?
No. Billing review depends on the actual services and applicable program, coverage, documentation, coding, and payer instructions. A census total is an operational input, not a payment determination.
How should a mid-day program change appear?
Preserve the actual events and effective dates under the organization's approved process. Test how enrollment, attendance, documentation, and reports represent the transition without silently reclassifying earlier care.
What should a daily reconciliation include?
Compare the expected population with actual attendance, associated records, and relevant administrative work. Classify differences, assign owners, and retain explanations rather than forcing all totals to match.
