Commissioners and system leaders routinely face a credibility problem: provider-reported performance packs don’t match payer views, Medicaid claims, encounter submissions, or utilization extracts. If that mismatch is handled informally, oversight becomes either passive (“we can’t trust anything”) or punitive (“prove everything”), neither of which supports safety, continuity, or value. A mature approach to using data for commissioning and oversight starts with reconciliation discipline—so outcomes frameworks and indicators rest on verified signals rather than competing spreadsheets.
This article describes a practical reconciliation method commissioners can run without building a parallel analytics bureaucracy: define what each dataset can and cannot prove, triage mismatches using a standard workflow, and require corrections that create an audit-ready trail.
What oversight bodies are expected to demonstrate when “the numbers don’t match”
Expectation 1: Defensible decision-making. When oversight escalates—additional monitoring, corrective action plans, payment holds, network interventions—commissioners are expected to show decisions were based on validated information, not assumptions or untested provider narratives. A reconciliation method creates that defensible basis.
Expectation 2: Proportionate, minimum-burden validation. Oversight should request evidence that is targeted and timeboxed, not open-ended document dumps. Reconciliation works when commissioners define what “good enough validation” looks like and use sampling and exception lists rather than asking providers to re-prove the entire service.
Start by being honest about what each dataset can prove
Claims and encounter data are not “truth.” They reflect billed/accepted events and payer rules. Provider operational systems reflect what staff recorded and what the organization believes it delivered. Reconciliation begins by mapping: which metrics should align tightly (e.g., paid units by service code), which will differ by design (e.g., non-billable coordination contacts), and which require a bridging rule (e.g., encounters submitted but not adjudicated yet).
The reconciliation workflow: triage, prove, correct, prevent
1) Triage mismatches into three buckets
Timing: lags between delivery, documentation, submission, and adjudication. Definition: mismatched service codes, eligibility windows, or denominator rules. Integrity: missing submissions, duplicates, or invalid member/service combinations.
2) Request bounded evidence
Commissioners should request an exception list and a defined sample, not a narrative. The sample should show the minimum fields needed to reconcile: member ID, service date, service code, units, location/setting indicator, and documentation timestamp.
3) Require a correction log and version control
Any corrected figure should come with: what changed, why, which records were affected, and whether prior commissioner decisions or reports were impacted. This is what prevents repeated “moving targets” and protects both sides during audits.
Operational example 1: Units of service don’t match between provider dashboards and claims extracts
What happens in day-to-day delivery
A commissioner sees a provider dashboard showing stable weekly units, but the claims extract shows a sharp drop. The commissioner issues a structured query requesting: (1) a weekly units-by-code report from the provider system, (2) the payer units-by-code extract for the same dates, and (3) an exception list of delivered services not appearing in claims/encounters. The provider’s billing lead runs a “delivered-not-submitted” report and tags each exception with a reason code (documentation incomplete, authorization missing, member eligibility gap, submission rejected). A small sample of exceptions is provided with the supporting artifacts needed to verify the reason code.
Why the practice exists (failure mode it addresses)
This practice prevents commissioners from confusing billing pipeline failure with service delivery failure—or vice versa. Without reconciliation, commissioners may assume a provider is not delivering contracted volume when the real issue is authorization or submission breakdown. Alternatively, they may assume services were delivered because dashboards say so when there is no evidence of submission or adjudication.
What goes wrong if it is absent
If mismatches are not triaged and proven, the commissioner may escalate contract remedies unfairly (damaging relationships and destabilizing delivery) or may miss genuine under-delivery (leading to access gaps and unplanned crisis use). Providers can also fall into “late scramble” billing behavior, creating spikes near deadlines that distort oversight and increase audit exposure.
What observable outcome it produces
A disciplined reconciliation produces a clear, auditable conclusion: timing lag, definition mismatch, or integrity issue. Outcomes include faster correction of rejected encounters, fewer authorization-related denials, a reduction in “delivered-not-submitted” backlogs, and commissioner confidence that utilization signals used for oversight reflect verified operational reality.
Operational example 2: Encounter data shows duplicates while the provider reports clean activity
What happens in day-to-day delivery
A payer flags duplicate encounters for the same member, same date, and overlapping times. The commissioner requests a duplicate exceptions file and asks the provider to map each duplicate to the operational record: visit note, scheduling entry, and submission transaction log. The provider’s data lead runs a de-duplication routine in the source system to identify the pathway that creates duplicates (mobile capture retries, template auto-save behavior, or staff logging both a phone contact and a visit as the same billable code). The provider submits: the duplicate categories, the corrected encounter file, and the prevention change (system rule, staff prompt, or billing edit).
Why the practice exists (failure mode it addresses)
Duplicate encounters inflate utilization, distort oversight thresholds, and create fraud/waste concerns even when the underlying cause is a system workflow bug. Reconciliation ensures commissioners can separate billing integrity risk from operational process failure and target the right corrective action.
What goes wrong if it is absent
Without a structured reconciliation process, commissioners may treat duplication as intentional overbilling and respond punitively before understanding the failure mode. Providers respond defensively, corrections are delayed, and the system loses a reliable view of true service volume—making access, capacity planning, and safety monitoring weaker.
What observable outcome it produces
Observable outcomes include a documented duplicate-prevention control (billing edits, validation rules, or capture workflow changes), reduced duplicate rates over subsequent cycles, and a correction log showing which records were amended and why. Oversight decisions become defensible because commissioners can demonstrate the signal was validated and addressed proportionately.
Operational example 3: Outcomes appear strong in provider reports but aren’t reflected in payer indicators
What happens in day-to-day delivery
A provider reports high rates of “successful community tenure” or “stability” outcomes, while payer indicators show high ED use or inpatient admissions for the same cohort. The commissioner initiates a reconciliation exercise that aligns cohorts, time windows, and attribution rules (which members, which months, which services). The provider supplies a defined sample of “successful” cases with the operational evidence used to classify success (risk reviews, care plan updates, crisis contacts, follow-up completion). The commissioner compares those cases to payer events and asks the provider to explain attribution differences (admissions outside program window, admissions unrelated to target condition, or misclassification due to missing documentation).
Why the practice exists (failure mode it addresses)
This practice prevents outcomes frameworks from becoming self-referential. Providers can legitimately deliver strong practice while payer indicators lag due to timing, attribution, or cohort mismatch. Equally, providers can overstate outcomes when success definitions are loose or documentation is late. Reconciliation forces shared, testable definitions.
What goes wrong if it is absent
Commissioners may either dismiss provider outcomes entirely (“we only trust claims”)—missing meaningful improvement work—or accept provider claims without challenge, embedding weak definitions into renewals and rate discussions. In both scenarios, oversight loses the ability to connect investment to verified impact and to detect risk early.
What observable outcome it produces
Outcomes include aligned attribution rules, clearer success definitions, and a repeatable validation method using samples and exception logic. Over time, commissioners see fewer unexplained gaps between operational outcomes and payer indicators, and providers build more reliable documentation and evidence chains that withstand scrutiny.
Make reconciliation routine, not a crisis response
Reconciliation works best as a scheduled control: monthly unit alignment checks, quarterly duplicate and denial reviews, and periodic outcomes attribution validation. Commissioners should publish a short reconciliation playbook so providers know what will be asked, how samples are selected, and what “resolved” means. Providers should maintain correction logs and submit version-controlled amendments rather than silently overwriting prior reports.
Done well, reconciliation reduces conflict and burden while strengthening oversight. It creates a shared reality: commissioners can act earlier and more proportionately, and providers can demonstrate delivery and impact with evidence that survives payer, audit, and public scrutiny.