Outcomes measurement in complex care is necessary, but it is also risky. Poorly designed targets can create perverse incentives: staff hesitate to escalate to EMS, restrictions increase to reduce incidents, or reporting becomes selective to protect performance metrics. None of these improve real outcomes, and all increase scrutiny when failures occur. A defensible approach is to build āsafety and rights guardrailsā into measurement so that stability and reduced crisis are pursued without suppressing escalation or drifting into unnecessary restriction. This article sets out practical controls aligned with complex care outcomes expectations and the governance principles embedded in complex care service design.
How perverse incentives show up in real services
Perverse incentives are rarely explicit. They emerge when teams feel judged on a narrow set of metrics without equal emphasis on safety, rights, and data integrity. Common patterns include: under-recording incidents to āimproveā performance, avoiding higher-acuity referrals to protect averages, and delaying escalation to keep ED/EMS counts down. These patterns often present as short-term āimprovementā followed by a serious event, safeguarding concern, or commissioner challenge.
Good measurement anticipates these failure modes and designs controls that make them visible early.
Two oversight expectations you should treat as non-negotiable
1) Reduction in emergency use must not be achieved by suppressing escalation
Commissioners and oversight bodies expect services to use emergency pathways appropriately. If reduced ED/EMS use coincides with deterioration incidents, delayed clinical response, or safeguarding concerns, the providerās credibility and safety rating can be harmed quickly.
2) Stability must be rights-based and defensible
Stability achieved through unnecessary restriction is not acceptable. Oversight scrutiny often focuses on whether restrictions were proportionate, time-limited, reviewed, and reduced where possible. Outcomes frameworks must therefore incorporate rights and restrictive practice governance, not treat them as separate topics.
Designing ābalanced measurementā to reduce gaming
A simple principle prevents many failures: never measure a desired outcome without also measuring the risks that appear when teams chase that outcome too narrowly. If you measure fewer ED visits, also measure timeliness of escalation actions and missed deterioration events. If you measure fewer incidents, also measure restrictive intervention use and quality-of-life indicators. Balanced measurement makes it harder to āwin the metricā by worsening practice.
Operational Example 1: Dual-metric escalation controls that detect under-escalation
What happens in day-to-day delivery
A provider pairs utilization metrics (ED/EMS contacts) with escalation quality metrics: time from first warning sign to clinical escalation, percentage of events where escalation pathway steps were followed, and ālate escalationā reviews (cases where deterioration was documented but escalation did not occur within agreed thresholds). Supervisors run brief reviews for any late escalation flags and record corrective actions: refresher coaching, pathway clarification, or clinical oversight changes. The dashboard shows both utilization and escalation quality so trends are interpreted safely.
Why the practice exists (failure mode it addresses)
The failure mode is under-escalation driven by performance pressure: staff try to avoid ED/EMS use and delay escalation, which increases risk and can lead to preventable harm. Dual-metric design ensures that reduced utilization is only celebrated when escalation quality remains strong.
What goes wrong if it is absent
A provider may appear to āimproveā on utilization while silently accumulating risk: missed deterioration, delayed response, and rising severity when crises finally occur. When a serious incident happens, scrutiny focuses on whether staff were discouraged from escalating, and the provider may face loss of commissioner confidence.
What observable outcome it produces
The service can evidence both reduced avoidable emergency use and safe escalation practice: timely clinical calls, documented pathway adherence, and fewer late escalation cases. This produces a defensible story: crises reduced through earlier intervention, not through suppressed access to emergency care.
Operational Example 2: Rights and restrictive practice guardrails embedded in outcomes reporting
What happens in day-to-day delivery
The provider tracks restrictive interventions (including environmental restrictions and access limitations) alongside incidents and stability indicators. Every restrictive intervention has a recorded rationale, review date, and de-escalation plan. A monthly rights review checks whether restrictions increased when incidents decreased, and whether quality-of-life indicators (community access, meaningful activity, relationship contact) remained stable or improved. Where restrictions rise, the service documents the decision pathway and tests alternatives through coaching, environmental changes, and clinical input.
Why the practice exists (failure mode it addresses)
The failure mode is āstability by restrictionā: incidents drop because autonomy is reduced, not because support improved. This can temporarily improve metrics while increasing long-term harm, distress, and safeguarding risk. Embedding rights guardrails keeps outcomes aligned with ethical and regulatory expectations.
What goes wrong if it is absent
Services may unintentionally drift into more restrictive practice because it feels safer and improves short-term indicators. Commissioners or oversight bodies then identify rights breaches, poor proportionality, or lack of review. The providerās outcome claims are discounted because they appear to be achieved through restriction rather than better care.
What observable outcome it produces
The service can demonstrate stable or reduced restriction alongside improved outcomes, supported by review records and measurable quality-of-life indicators. This produces stronger commissioning confidence because it shows stability is sustainable, rights-based, and governed.
Operational Example 3: Data integrity controls that prevent selective reporting
What happens in day-to-day delivery
A provider implements routine data integrity checks: incident log-to-progress note reconciliation, random sampling audits, and āmissing eventā prompts where narrative notes imply an incident or EMS contact but no formal record exists. Results are reviewed in governance, and corrective actions include retraining, simplifying reporting steps, and supervisor sign-off for high-risk events. The provider also maintains a clear āno blame for reportingā culture statement tied to safety, with leadership reinforcement when reporting volumes rise due to improved capture.
Why the practice exists (failure mode it addresses)
When metrics matter, teams can unconsciously under-report to avoid scrutiny. The failure mode is selective reporting: the dashboard looks better while real risk increases. Data integrity controls detect and correct this early, protecting both safety and credibility.
What goes wrong if it is absent
Commissioners eventually detect inconsistency (for example, fewer incidents but more safeguarding concerns, or fewer EMS calls but more serious events). Once credibility is lost, reporting burden increases and performance management becomes adversarial. Internally, learning is weakened because the service no longer āseesā its true risk profile.
What observable outcome it produces
Completeness improves and trends become trustworthy. The provider can demonstrate an audit trail: checks completed, gaps identified, actions taken, and improvement verified. This strengthens commissioner confidence and supports safer operational decision-making.
How to communicate outcomes without creating pressure that drives drift
Leaders should frame outcomes as learning and safety tools, not as ātargets to hit.ā Where contracts include outcome thresholds, providers can still set internal expectations that protect practice: escalate when needed, record consistently, and review restrictions proportionately. Commissioners generally respond well to transparent governance that shows how the provider prevents gaming and protects service users.
What āgoodā looks like to oversight bodies
Oversight bodies recognize mature systems by the presence of guardrails: dual-metric escalation controls, rights-based restrictive practice governance, and routine data integrity checks. These controls make outcome claims more credible because they demonstrate the provider is not improving numbers at the expense of safety or rights.