.jpg){kind=link}
- Published
- Updated
Different dates for the same site visit across systems can cause a missed or duplicate payment. Reconciliation is not merely an exercise in matching totals. Its purpose is an understandable account of the service paid for, the applicable period and the exceptions still awaiting resolution.
Decision framework
Identify the authoritative payment source and data-cut date. A visit occurring, its data being entered and its acceptance for payment may be separate events. Connect these distinctions to the agreement instead of treating delayed entry automatically as evidence that a visit did not happen.
Worked example
For example, a visit takes place at month-end but is entered the following week. The payment report includes it in the next cycle while the site queries the earlier month. The team compares visit date, payment cut and previous payments. It explains the timing without changing the clinical record simply to make financial reporting align.
Implementation checks
- Check consistency of trial, site, activity unit and period identifiers on each payment line under review.
- Track duplicate, pending and corrected items using separate statuses and link them to previous payments where relevant.
- Keep site-query evidence, the response owner and the resolution outcome together in one traceable history.
Acceptance and boundaries
Accept reconciliation when a sampled payment history can be reconstructed independently. This exercise does not replace clinical data verification; it clarifies boundaries between teams using related information for different purposes. Communicate a changed payment-cut method through the applicable agreement and site communication processes.
Frequently asked questions
- Are matching system totals sufficient?
- No. Incorrect items can offset one another; sampled checks at activity level are still useful.
- Does a financial correction change the clinical record?
- Do not alter source records merely for financial alignment; assess genuine errors through the authorised data process.