Cash Application Controls

Cash application is an application decision, not transcription: money has arrived — by electronic funds transfer, by a mailed check, by a card or cash at the desk — and the task is to match it to the account it belongs to and apply it there. Most of the cluster is about doing that accurately, so a payment lands on the right claim and the right patient. This article is about doing it with integrity: so that money cannot be diverted or lost along the way, and so the practice can demonstrate it wasn't. Those are different guarantees. Accuracy is threatened by carelessness; integrity is threatened by carelessness and by the rare bad actor, and the second threat is the reason cash application needs controls that ordinary posting does not.

A control is not “post more carefully.” In internal-control terms, cash is a vulnerable asset, and the standard response to a vulnerable asset is a system of designed controls rather than a resolution to pay attention. That system is the five-component framework most practices know by name — the one the private-sector COSO framework defined and the federal Green Book adapts for public bodies — and segregation of duties is one of its control activities. The defining property of a control is that it does not rely on any one person being careful or honest: it is built so that a mistake or a diversion is stopped, or caught by someone else, regardless of who was at the keyboard. That is exactly why reconciliation is a control rather than a second look — its power comes from comparing the record against something the person who posted cannot influence.

The single most important control over cash application is that the work is split among different people. The internal-control standards state it plainly: management divides key duties among different people so that no one individual controls all the key aspects of a transaction. Four responsibilities in particular are kept apart — authorizing a transaction, recording it, handling the asset itself, and reviewing the result. In cash application that maps to four different hands: whoever has custody of the money (opening the mail and preparing the deposit, or holding access to the bank and the posting system) is not whoever records it against accounts; neither of them is whoever authorizes a write-off or a refund; and none of them is whoever reconciles the day's postings back to the bank.

The reason this is the spine, and not just one control among many, is that the danger in cash application is specifically that the money and the record of it pass through the same hands. If one person receives a payment, posts it, and reconciles it, then every check performed on that money is a check that person runs on their own work — the same self-reference reconciliation exists to break, now applied to a person instead of a process. A diversion is not exposed by a ledger that balances, because the person who balanced it is the person who moved the money. Split the chain and that stops being possible: the poster cannot quietly divert a payment because the person who reconciles will see a deposit with no matching posting, and the person who reconciles cannot invent a posting because they never touched the cash. Independence is not a nicety here; it is the mechanism.

No single control catches everything, so the money is protected by layers — the same reason a posting error needs several standing checks, applied here as deliberate design rather than luck. The layers fall into three kinds, arranged along the path a payment travels from received to applied to closed.

Money enters an account as a payment, and it leaves by exactly two doors that do not require anyone outside the practice to receive it: a write-off, which erases a balance, and a refund, which sends money back out. Those two doors are precisely where a diversion disguises itself — write off a balance that was actually paid and keep the payment, or route the refund of a real credit to the wrong hands. So both doors get the same lock: authorization by someone other than the person who posts, granted before the entry is made rather than reviewed after it.

This control sits on top of two posting decisions the cluster already owns, and it is orthogonal to both. Whether an adjustment is a required contractual adjustment or a discretionary write-off is a category question that adjustment vs. write-off answers; whether a credit balance is real and who may receive the refund is credit balance refund's subject. The control here is neither of those. A group code names whether an amount is the patient's or the provider's to absorb; it does not say that erasing the balance was authorized — that is a decision, and a decision that makes money disappear from an account gets an approver who is not the person who handled the money. The point of segregation is exactly this: the person recording the cash is not the person empowered to make it vanish.

Every control above produces evidence, and the evidence is the point — a control you cannot show ran is indistinguishable from one that never did. Reversing rather than deleting leaves the correction on the record; the reconciliation carries the name of whoever performed it and whoever reviewed it; the write-off carries its approver; the suspense review carries a date. “We are careful” is a claim about people. “Here is the reconciliation, signed by someone who does not post, showing the deposit matched the postings” is proof about the process — and only the second kind survives a staff turnover, a system conversion, or an audit.

Proof matters here, and not just care, because a balanced total is not evidence of integrity. A compliant remittance advice has to balance — the total paid equals the total charged plus or minus the adjustments, checked at the line, the claim, and the provider level — so the cash reconciles around a payment applied to the wrong patient, or a real balance written off and pocketed, exactly the way it reconciles around an honest posting error. The numbers add up either way; what separates “applied correctly by an honest process” from “moved, and the books made to agree” is an independent trail, nothing else. That trail is also what protects the staff: when money is missing and everyone had access to everything, everyone is a suspect; when duties were segregated and the record is intact, the trail clears the people who did their part and points at the actual gap.