How to analyze referral-source concentration

Define what a referral source means in this report

Begin with the counting rule. A family's web inquiry, a discharge planner's service request, a physician's order and a payer authorization may all appear in the same file. They answer different questions. State which event sets the source field in this report. Keep other relevant events in separate fields.

The referral-source definition explains that distinction. Choose a consistent method for the population being reviewed, including repeat inquiries, returning clients, transfers and multiple contacts. Show where the source label came from: an intake field, a referral record or someone's memory. A filled-in field is not always a verified fact.

State the service line, locations, payer population and dates. A home-care inquiry report should not silently absorb home-health clinical referrals. A month's starts may also include inquiries from earlier months. Use labels that describe the actual records. Do not give unlike events one convenient name.

Normalize names and examine shared relationships

Create a source map that retains the original entry, normalized name and proposed group. Review spelling differences, branch labels and individual contact names. Merging duplicate entries can reveal dependence. Combining unrelated sources can make the business look more dependent than it is.

Support a group view with evidence relevant to the risk being studied. Common ownership may matter, but so may a shared decision process or one key relationship holder. Keep those reasons distinct. If the facts are unclear, show the ungrouped result and the proposed grouping as separate views until the team resolves the issue.

The goal is a clear trail back to the records. Another reviewer should be able to repeat the result and undo a grouping that lacks support. A polished chart without that history can hide gaps in the raw list.

Follow the same cohort from inquiry to service

Choose a defined group of inquiries. State how long you follow that group for resulting starts. Count each inquiry once under that rule. Separate unresolved or still-open inquiries from confirmed outcomes. A late start does not belong in the wrong cohort merely because it appears in a later month's activity total.

The move-in matching FAQ adds a fictional record check before any channel comparison. Four entries include the same person twice and two plans without a completed admission at the review date. Use the exercise to identify unique prospects and completed events. Then review the referral list’s support for future activity.

Use a fictional cohort of 120 unique inquiries, followed until the stated review cutoff. Sixty result in service starts. Two branches have documented membership in one network; the other sources and unknown entries remain separate. These are invented counts, not an industry conversion benchmark.

The combined network contributes 45 of 120 inquiries, or 37.5%, and 30 of 60 starts, or 50%. The same relationship has different shares because the denominators differ. Looking only at the branch names would also hide the combined relationship in this hypothetical case.

If the analyst removes the 20 unknown-source inquiries, the network becomes 45 of 100 known-source inquiries, or 45%. That is a known-source measure, not the original all-inquiry measure. Nothing about the underlying network activity changed. Keep unknown attribution visible and label any restricted denominator.

Connect starts to revenue without assuming equal value

A share of starts does not establish a share of revenue, profit or cash. Service duration, payer terms, service intensity and costs may differ. Define the group and period for the financial review. Then explain how its service records link to the invoices and receipts.

Suppose a separate fictional monthly financial schedule attributes $90,000 of $300,000 in revenue to the network. Its revenue share is 30%. You cannot get that figure by multiplying the 50% start share by total revenue. It comes from a separate set of financial records. Check those records on their own terms.

Trace totals to the books and supporting documents. IRS guidance identifies invoices, receipts, deposit information and other records as support for business entries. A referral spreadsheet is not a substitute for that reconciliation. (Source: IRS supporting records)

Use the census and payer reconciliation guide for the broader connection between activity, billing and collections. Keep payer mix separate from referral attribution: who pays is not necessarily who first connected the person with the business.

Assess the relationship behind the percentage

Ask why the source chooses this provider and who supports that decision. The answer may involve service availability, response time, communication, location, clinical capability or a long-standing personal contact. Test the explanation against records and the team's actual process rather than accepting a general assurance that the relationship is strong.

Identify owner dependence. Does the seller handle every introduction, or can the team explain the intake process and resolve service issues? Review how contacts are maintained, how missed requests are tracked and how the agency responds when it lacks capacity. A contact list does not show how the work gets done.

Also examine changes over time using comparable cohorts. A lower share might mean diversification, a decline in the source's activity or growth elsewhere. A higher share might reflect better service, fewer alternative sources or a data-capture change. Check the total being counted and any changes to the records before choosing an explanation.

Keep patient choice and referral-law review separate

Do not assume that a relationship can be purchased as a guaranteed stream of people. For hospitals subject to 42 CFR 482.43, the current post-acute provisions address appropriate provider lists, network information and patient choice. Paragraph (d)(2) requires hospitals to inform patients about their freedom to choose among participating providers and not limit the qualified options available. These are hospital discharge duties, not a guarantee that any agency receives future business. (Source: 42 CFR 482.43(d))

Have qualified counsel review actual financial and contractual relationships. OIG describes the federal Anti-Kickback Statute's prohibition on knowing and willful remuneration to induce or reward referrals or business payable by federal healthcare programs. Its safe-harbor discussion requires the applicable conditions to be met. A commercially useful arrangement does not establish legal compliance, and this guide does not approve a fee or contract. (Source: HHS OIG Anti-Kickback Statute overview)

Keep the review specific to the parties, services, payers and jurisdiction. Do not use the hospital rule as a universal description of private placement arrangements, or assume a different payer removes every legal concern. Record the question for the appropriate adviser rather than converting a general source into a transaction-specific answer.

Test an interruption with explicit cost assumptions

A concentration share shows dependence. It does not decide how much money would be lost. Build a limited scenario from revenue and cost records you can support. State the size and duration of the assumed interruption, whether replacement activity is assumed, and which costs would actually change. Staff commitments and other duties may prevent a quick cost reduction.

The following fictional monthly example begins with $300,000 of revenue, $180,000 of variable costs and $80,000 of retained costs. The illustrative remainder is $40,000. Assume a $30,000 revenue interruption with no replacement revenue during the modeled month. The two cases differ only in how much cost is assumed to go away.

The first interruption case reduces the remainder by $12,000; the second reduces it by $24,000. Neither is a valuation discount, cash forecast or standardized earnings measure. Debt, taxes, capital needs and collection timing are not modeled. Cost assumptions do not authorize staffing cuts or changes to patient care; actual service obligations and qualified review must govern the plan.

Control access and document the transition questions

Begin with the information necessary for the review and follow the agreed confidentiality process. Source names can identify the seller, and small groups or detailed dates can identify patients. HHS explains that de-identification requires more than deleting names. Coded workpapers and an NDA do not independently establish permission to disclose protected information. (Source: HHS de-identification guidance)

Record unresolved attribution, contract, privacy and continuity questions in the diligence checklist. Assign the next action and identify which forecast assumptions depend on the answer. Coordinate any source contact with the seller and the authorized advisers; this article does not authorize outreach.

A useful closing review shows the measured dependence and the team behind each relationship. It also records the remaining limits and downside assumptions. Carry the operational items into the first 90 days plan. The goal is a supportable transition plan, not a promise that historical referral patterns will repeat.