Performance-based contracting is often discussed as a payment concept, but it succeeds or fails as an operational control system. If performance measures are not grounded in real workflows, they become either uncollectable (staff ignore them) or gameable (numbers look good while outcomes drift). In mature systems, the performance model sits inside contract management and provider performance arrangements and is designed to match the realities of intake, eligibility, and triage operating models so only eligible demand enters, services are authorized cleanly, and performance data reflects what was actually delivered.
What “Good” Looks Like: Measures That Control Delivery, Not Just Report It
The practical aim is simple: create measures that (1) guide staff decisions, (2) trigger escalation early when delivery is drifting, and (3) align financial payment to verifiable service activity and quality safeguards. In Medicaid and county systems, this typically means combining utilization integrity (units/visits/contacts), timeliness (response/assessment), quality and safety signals (incidents/restrictive interventions), and outcome proxy measures that are credible for the service type.
Operational Example 1: A Timeliness Measure That Prevents “Queue Drift”
What happens in day-to-day delivery
The contract sets a timeliness standard (for example: referral acknowledged within 1 business day; initial assessment within X days; service start within Y days). Intake staff record referral timestamps in the case system; supervisors run a daily “aging” report; and a weekly huddle assigns named actions (call-backs, home visits, expedited authorizations). Contract managers validate the report against a sample of case notes so the measure reflects reality, not just status changes.
Why the practice exists (failure mode it addresses)
Timeliness controls exist because queues silently drift when demand rises, staffing changes, or authorizations stall. Without a hard operational standard, services appear “open” while people wait, risks escalate, and the system loses visibility on who is deteriorating or disengaging.
What goes wrong if it is absent
If timeliness is not measured and acted on, the backlog becomes normalized. Providers may “start” cases administratively without meaningful contact to protect reported performance. Commissioners experience rising complaints, higher ED use for preventable issues, and late crisis presentations that could have been mitigated with earlier engagement.
What observable outcome it produces
A real timeliness control produces visible backlog reduction, fewer late starts, and a stronger audit trail showing who made contact and when. It also supports better capacity planning because the organization can demonstrate demand volume, aging profiles, and where the flow is breaking.
Operational Example 2: Verification Controls That Stop “Phantom Performance”
What happens in day-to-day delivery
The provider uses visit/service verification appropriate to the service (EVV for in-home supports, encounter capture for community-based programs, signed attendance where relevant). Finance and operations reconcile weekly: delivered encounters must match case documentation, staff schedules, and authorization rules. A small sample is audited monthly by an internal QA lead who checks documentation quality (what was delivered, risks addressed, outcomes observed) rather than treating the visit as a tick-box.
Why the practice exists (failure mode it addresses)
Verification exists to prevent paying for activity that cannot be substantiated, and to prevent “numbers-first” behavior where staff record encounters without the quality content needed to manage risk. It also protects staff by ensuring records accurately reflect what happened if concerns or incidents arise later.
What goes wrong if it is absent
Without verification, services can be over-claimed, duplicated, or miscoded, creating exposure during Medicaid audits and recoupment actions. Operationally, leadership loses the ability to trust performance dashboards. A provider can appear compliant while clients report missed visits, poor follow-up, or unresolved safeguarding concerns.
What observable outcome it produces
Strong verification reduces claim denials, minimizes retrospective corrections, and creates defensible documentation trails. Commissioners see fewer invoice disputes and a clearer link between funded units, actual delivery, and outcomes monitoring.
Operational Example 3: A Quality Measure That Drives Supervision (Not Paperwork)
What happens in day-to-day delivery
The contract requires supervision-linked quality checks (for example: a monthly case review rate; a “contact after incident” standard; or a medication reconciliation check where relevant). Supervisors complete structured case review templates that capture risk status, restrictive practice use, incident trends, and follow-up timeliness. Findings are logged, themes are fed into training, and repeat issues are escalated to leadership with deadlines and owners.
Why the practice exists (failure mode it addresses)
This exists because quality failures rarely begin as major incidents; they start as missed follow-ups, weak documentation, unmanaged risks, and inconsistent practice across staff. A supervision-driven quality measure forces routine visibility on whether care planning, safeguarding, and escalation pathways are functioning.
What goes wrong if it is absent
Without a practical quality control, providers may rely on annual training and reactive investigations. Problems repeat: restrictive interventions drift upward, incident themes recur, or medication risks accumulate. Commissioners then see late discovery of harm, inconsistent provider narratives, and poor learning cycles after adverse events.
What observable outcome it produces
When done well, the organization can evidence reduced repeat incidents, improved timeliness of follow-up, and stronger consistency in practice. The “proof” is visible in audit logs, supervision records, improved documentation completeness, and trend charts that show risks stabilizing rather than escalating.
Oversight Expectations to Build In From Day One
Two expectations show up repeatedly in funder and system oversight: first, measures must be auditable (definitions, data lineage, sampling, and variance handling are documented); second, performance must trigger action (named owners, deadlines, escalation thresholds, and commissioner visibility). If a performance model does not produce operational decisions, it is usually treated as non-credible during monitoring and renewal discussions.