Utilisation Review

How utilisation review evaluates healthcare use

Utilisation review assesses whether proposed, ongoing or completed healthcare services meet defined requirements for clinical appropriateness, setting, intensity, duration and coverage. This page explains how review operates before, during and after care, how criteria are applied, and how the effects on spending, access, quality and administrative burden should be evaluated.

The review does not determine clinical need in isolation. It applies payer, programme or organisational rules to a specific request or claim, often using clinical evidence, benefit design and information supplied by patients and providers.

What utilisation review is intended to achieve

Utilisation review is used to influence the amount, type and location of healthcare delivered. Its intended objectives can include reducing low-value care, directing patients to appropriate settings and protecting a healthcare budget from unnecessary expenditure.

Objectives may include:

• Confirming that a service is clinically appropriate.
• Applying coverage and benefit rules consistently.
• Preventing avoidable or duplicative care.
• Encouraging use of lower-cost but clinically appropriate alternatives.
• Managing treatment duration or frequency.
• Directing care to an appropriate level or place of service.
• Identifying cases requiring coordination or specialist review.
• Detecting billing errors, waste, abuse or fraud.
• Monitoring patterns of service use.

These objectives are not interchangeable. A process designed to detect incorrect billing may not be suitable for judging clinical appropriateness, and a mechanism that reduces expenditure may not improve care.

Prospective, concurrent and retrospective review

The timing of review determines which decisions can be changed and which risks arise. Review systems commonly operate at three stages.

Prospective review

Prospective review occurs before a service is delivered. Prior authorisation or precertification requires approval before the payer will confirm coverage for a medicine, procedure, admission, test or other service.

Prospective review can prevent spending on services judged inappropriate, but it can also delay care, create administrative work and discourage use of services that would have benefited the patient.

Concurrent review

Concurrent review occurs while care is being delivered. It may assess continued hospital stay, treatment extension, additional therapy or transfer to another setting.

Concurrent review can respond to changes in a patient's condition but requires timely clinical information. Delayed decisions can disrupt discharge, treatment planning or continuity of care.

Retrospective review

Retrospective review occurs after a service has been delivered. It may examine claims, medical records, coding, adherence to coverage rules or patterns of use.

Retrospective review can identify inappropriate payment or inform future policy, but it cannot prevent the original service. Denial after care has been delivered may shift financial risk to patients or providers, especially when requirements were unclear.

Services selected for review

Review is often targeted at services with high cost, rapid growth, wide variation, safety concerns or uncertain value. Requiring review for every routine service can create more administrative cost than financial or clinical benefit.

Selection may consider:

• High unit price or aggregate expenditure.
• Risk of inappropriate or non-evidence-based use.
• Availability of effective lower-cost alternatives.
• Variation between providers or regions.
• Potential patient-safety risk.
• Use outside approved indications or coverage conditions.
• New technologies with uncertain evidence.
• Services vulnerable to billing error or fraud.

Targeting rules should be reviewed as evidence, prices and practice change. A service that once required close review may no longer justify the burden, while new patterns of low-value use may emerge.

Applying clinical and coverage criteria

Review criteria translate evidence, coverage policy and benefit rules into operational decisions. Criteria should be specific enough for consistent use while allowing justified consideration of individual circumstances.

Criteria may address:

• Diagnosis and disease severity.
• Previous treatment and response.
• Contraindications or safety requirements.
• Age or risk group.
• Treatment dose, frequency or duration.
• Provider qualifications.
• Place of service.
• Availability of alternatives.
• Documentation requirements.
• Coverage exclusions and limitations.

Clinical appropriateness and contractual coverage should be distinguished. A service can be clinically reasonable but excluded from a particular benefit, while a covered category may still require evidence that its use is appropriate for the individual patient.

The review pathway

A transparent review process should tell patients and providers what information is required, who decides and how quickly a response will be issued. The pathway should reflect the urgency of the clinical situation.

A typical process includes:

1. Identify the service subject to review before or during the ordering process whenever possible.
2. Submit the request with the required clinical and administrative information.
3. Check completeness and eligibility under the applicable benefit or programme.
4. Apply the approved criteria using qualified reviewers.
5. Request additional information only when it is necessary for the decision.
6. Approve, modify or deny the request within the required timeframe.
7. Communicate the decision and reasons to the patient and provider.
8. Offer reconsideration or appeal when the decision is adverse.
9. Record the decision and outcome for audit and policy evaluation.

Emergency and urgent care require expedited procedures or exceptions. A standard review timetable can be clinically inappropriate when delay could cause harm.

Automated and human review

Automated systems can compare requests with structured criteria, check documentation and approve routine cases. Human clinical review is needed when information is incomplete, circumstances are unusual or judgement is required.

Automation should not turn missing data into an automatic clinical conclusion. Systems should distinguish between a service that fails a criterion and a request that lacks enough information to evaluate it.

Automated rules require:

• Transparent logic and current criteria.
• Validated data inputs.
• Monitoring for systematic error.
• A route to timely human review.
• Protection against discriminatory effects.
• Version control and audit records.

Artificial intelligence may assist prioritisation or document review, but responsibility for adverse coverage decisions should remain accountable and reviewable.

Prior authorisation

Prior authorisation is one of the most visible forms of utilisation review. It makes coverage conditional on approval before the service is delivered, except where an emergency or another defined exception applies.

Its effectiveness depends on:

• Whether the targeted service is frequently used inappropriately.
• The clarity and clinical validity of the criteria.
• The time required to submit and decide a request.
• The availability of appropriate alternatives.
• The rate of approval and reversal on appeal.
• The cost of administering the process.
• The clinical consequences of delay or abandonment.

A high approval rate may indicate that the requirement deters inappropriate requests before submission, but it may also indicate that the review adds burden to services that are usually appropriate. The result cannot be interpreted without information about requests that were never submitted or were abandoned.

Step therapy and treatment sequencing

Step therapy requires use of one treatment before coverage is provided for another. It may encourage lower-cost or established treatment but can be harmful when the required first option is unsuitable for an individual patient.

A defensible policy should specify:

• The evidence supporting the treatment sequence.
• Clinical exceptions and contraindications.
• How previous treatment history is recognised.
• The duration of the required trial.
• The outcome that establishes failure or intolerance.
• The process and timetable for requesting an exception.

The administrative sequence should not override urgent clinical need or force repetition of a treatment already shown to be ineffective.

Concurrent review and length of stay

Concurrent review may assess whether hospital-level care remains necessary or whether care can continue safely in another setting. It should consider the patient's clinical status and the availability of an appropriate alternative.

Shortening a hospital stay does not produce a genuine efficiency gain when community, rehabilitation or social-care services are unavailable. Premature transition may lead to readmission, harm or greater burden on patients and caregivers.

Evaluation should therefore examine total care pathways rather than counting hospital days in isolation.

Retrospective claims review

Claims review can identify coding errors, duplicate billing, services outside coverage and unusual patterns of use. Statistical outliers can support investigation but do not establish inappropriate care on their own.

Retrospective review should distinguish:

• Administrative or coding error.
• Lack of required documentation.
• Service outside the benefit.
• Clinically inappropriate use.
• Fraud or deliberate misrepresentation.

Each conclusion requires different evidence and response. Labelling every unusual claim as fraud can create unfair sanctions and distort future reporting.

Appeals and independent review

An appeal allows the patient or provider to challenge an adverse decision. The process should be accessible, timely and proportionate to the urgency and consequence of the denied service.

An effective system provides:

• A clear explanation of the decision and criterion used.
• Instructions and support for requesting review.
• Expedited review for urgent cases.
• Review by a clinician with relevant expertise.
• Consideration of individual circumstances and new evidence.
• Independent external review where required.
• Protection from loss of access while an urgent appeal is pending when clinically appropriate.

Appeal reversal rates are important but incomplete. Patients and providers may lack the time, knowledge or resources to challenge an incorrect decision.

Measuring the performance of utilisation review

Performance should be assessed through clinical, financial, access and administrative outcomes. Counting denials or savings alone can reward barriers rather than appropriate care.

Measures may include:

• Request volume.
• Approval, partial approval and denial rates.
• Decision time.
• Requests for additional information.
• Appeal and reversal rates.
• Abandoned or delayed care.
• Changes in service use.
• Substitution to alternative services.
• Patient outcomes and adverse events.
• Total healthcare expenditure.
• Provider and payer administrative costs.
• Patient and clinician burden.
• Differences across population groups.

Rates should be stratified by service, urgency and outcome. A single overall approval rate can conceal severe problems in a small but clinically important category.

False approvals and false denials

Review criteria can produce errors in both directions. A false approval permits care that does not meet the intended rule, while a false denial blocks care that should have been covered.

Tightening criteria may reduce inappropriate approvals but increase inappropriate denials. The balance should reflect the clinical and financial consequences of each error.

Evaluation should consider:

• Harm caused by delayed or forgone appropriate care.
• Cost and harm caused by inappropriate care.
• Frequency of each error.
• Availability of rapid correction.
• Whether errors are concentrated in particular groups or services.

Appeals identify only some false denials because not every affected person challenges the decision.

Administrative burden and opportunity cost

Utilisation review consumes time for clinicians, administrative staff, patients and payer reviewers. These costs can offset part of the expenditure the programme is intended to save.

Net financial impact can be represented as:

$$ Net\ savings = Avoided\ healthcare\ spending - Review\ administration\ cost - Additional\ downstream\ cost $$

Administrative cost may include staff time, technology, documentation, appeals and delays. Additional downstream cost can arise when a denied service is replaced by a more expensive alternative or when delayed treatment leads to deterioration.

The relevant question is not whether the process reduces spending on the reviewed service, but whether it improves total value across the care pathway.

Effects on access and equity

Review requirements can affect patients differently. People with limited language support, digital access, time, clinical advocacy or ability to pay while waiting may face greater barriers.

Equity analysis should examine:

• Approval and denial rates across relevant groups.
• Time to decision and treatment.
• Access to specialists who can submit required documentation.
• Ability to complete appeals.
• Geographic variation.
• Effects on people with complex or rare conditions.
• Continuity during insurance or provider changes.

Uniform criteria do not guarantee equitable access when patients have different capacity to navigate the process.

Clinical autonomy and shared decision making

Utilisation review can support evidence-based care but may also constrain clinician and patient choices. Criteria should allow exceptions when individual circumstances make the standard pathway inappropriate.

The review process should not imply that payer approval proves a service is clinically necessary or that denial proves it is clinically inappropriate. Clinical decision making and coverage determination are connected but distinct responsibilities.

Transparent reasons and timely dialogue can reduce conflict and help distinguish a coverage rule from a clinical recommendation.

A simplified example

Suppose a payer requires prior authorisation for an imaging test because use varies widely and lower-cost assessment is usually recommended first. During one year, 10,000 requests are submitted, 8,800 are approved initially, 700 require additional information and 500 are denied.

Of the denied requests, 200 are appealed and 120 are approved after review. The observed appeal reversal rate is:

$$ Appeal\ reversal\ rate = \frac{120}{200} = 0.60 = 60% $$

The high reversal rate suggests that the original decisions, documentation rules or criteria require investigation. It does not show what happened to the 300 denied requests that were not appealed or to patients whose clinicians abandoned a request before submission.

Distinguishing utilisation review from related activities

Utilisation review overlaps with other management and oversight functions but should not be treated as identical to them.

• Utilisation management is a broader programme that may include review, case management, provider feedback and benefit design.
• Quality review evaluates whether care meets quality and safety standards rather than primarily determining coverage or payment.
• Case management coordinates care for people with complex needs.
• Medical audit examines clinical practice or records against defined standards.
• Claims adjudication applies contractual and administrative rules to payment.
• Fraud investigation examines possible intentional deception and requires evidence beyond unusual utilisation.

The same organisation may perform several functions, but each should have a clear purpose, evidence standard and accountability process.

Common misunderstandings

Utilisation review is not automatically evidence of inappropriate care and does not create savings simply by reducing approvals. Its effects depend on the services targeted, criteria, implementation and downstream consequences.

Common misunderstandings include:

• Prior authorisation and retrospective claims review are not the same process.
• A denial is not necessarily a clinical judgement that the service would not help.
• A high approval rate does not by itself prove that review is unnecessary or effective.
• A low appeal rate does not prove that denials are correct.
• Reduced use does not automatically represent improved efficiency.
• Administrative costs should not be treated as zero.
• Automated review does not remove the need for accountable clinical judgement.
• Consistent rules do not necessarily produce equitable access.
• Savings on one service may be offset by costs elsewhere.

Interpreting utilisation-review evidence

Evidence should be interpreted using the review type, targeted service, criteria, timing, comparison and complete care pathway. Changes in use should be accompanied by information about outcomes, substitution, delays and administrative cost.

A well-designed utilisation-review programme focuses scrutiny where it can improve value, applies transparent evidence-based criteria and provides rapid correction when standard rules do not fit the patient. Its success is measured by appropriate care and net system value rather than by denials alone.