Value-Based Purchasing

Value-Based Purchasing

Value-based purchasing uses a purchaser's contract, provider selection or payment terms to encourage better health-care quality, outcomes or efficiency for the resources spent. Its details depend on the purchaser and programme; there is no single universal score or payment formula. This page explains the purchasing decision, the link between measured performance and payment, and the safeguards needed to avoid rewarding attractive scores instead of better care.

What a purchaser is trying to change

A public programme, insurer or other purchaser can decide which providers to contract with and how their payments respond to performance. A scheme may adjust a base payment, award a bonus, withhold a portion for redistribution, or include quality requirements in a contract. The mechanism should say who bears financial risk, which patients count and whether incentives reward improvement, attainment or both.

This is distinct from value-based pricing of a medicine, which concerns the price paid for a product in relation to its assessed value. It also differs from a general aspiration to provide high-value care: purchasing requires an actual rule or contract that changes selection, payment or accountability. Fee-for-service, capitation or bundled payment may be the underlying payment method on which a performance adjustment is layered.

Design choice	Example	Risk to assess
Measure set	Outcomes, safety, patient experience and resource use.	A narrow score can divert effort from unmeasured care.
Comparison	Absolute benchmark or improvement from baseline.	Baselines and case mix can make comparisons unfair.
Payment lever	Bonus, penalty, withhold or contract renewal.	The incentive may be too weak or create harmful pressure.
Accountable unit	Clinician, facility, network or programme.	Attribution may not match control over outcomes.
Review period	Performance year and payment year.	Delayed data can weaken feedback and obscure changes.

The purchaser should make the payment rule and data definitions auditable before the period begins. A provider cannot respond fairly to measures it cannot observe or control, and retrospective rule changes can undermine credibility.

Choose measures that reflect patient value

Good measures should be meaningful to patients, reliably recorded and sensitive enough to detect important differences. Outcome measures may be closer to the ultimate aim than process measures, but can require risk adjustment and long follow-up. A process measure can be useful when it represents evidence-based care and is less easily confounded by case mix.

Risk adjustment may account for differences in patient health that affect measured outcomes independently of care quality. It should not automatically erase disparities attributable to poor access or care, and it cannot fully resolve unmeasured severity or social barriers. Stratified reporting can reveal whether an average improvement bypasses disadvantaged patients.

Pair a headline metric with balancing measures. For example, an incentive to reduce admissions should also track patient harm, avoidable emergency use, follow-up access and experience so that necessary admissions are not discouraged. Audit coding changes, denominators and exclusions to limit gaming and selection of easier-to-treat patients.

Follow a payment adjustment through the accounts

Consider a fictional purchaser contract in which a provider's otherwise payable annual base amount is £10 million. The purchaser withholds 2% (£200,000) and later returns 110% of that withheld amount (£220,000) based on a defined performance score. Final payment is $£10{,}000{,}000-£200{,}000+£220{,}000=£10{,}020{,}000$, which is £20,000 above the original base.

This is a teaching contract, not a description of any current programme's formula. The 110% applies to the withheld £200,000, not to the full £10 million; applying it to the base would create an entirely different payment. Whether the overall programme is budget neutral depends on all providers' adjustments and the purchaser's funding rules, not this provider's arithmetic.

Spreadsheet item	Illustrative formula	Result
Original base payment	=10000000	£10,000,000.
Withheld amount	=10000000*0.02	£200,000.
Performance return	=200000*1.10	£220,000.
Final payment	=10000000-200000+220000	£10,020,000.
Change relative to base	=10020000-10000000	£20,000 increase.

A £20,000 payment increase is a transfer under the contract, not evidence of a £20,000 health gain or saving in care resources. The purchaser and provider may also incur reporting and improvement costs. An evaluation of the purchasing programme must compare patient outcomes, spending and administrative burden with a credible alternative policy.

Determine whether the scheme caused improvement

Higher scores after a scheme begins do not establish causal impact: secular improvement, changing patients and documentation practices may also raise measured performance. An evaluation should specify the counterfactual, measure trends before implementation and compare appropriately exposed providers or periods. When possible, use designs that address selection, spillovers and changes in coding.

Report both intended and unintended outcomes, including quality, safety, access, spending and distribution across providers and patients. A payment change can occur without improved care if the measure is insensitive or easy to game. Conversely, worthwhile quality gains may be missed if the chosen outcome takes longer to appear than the payment cycle.

Equity and practical design safeguards

Providers serving people with complex needs may face different baseline risks and fewer resources for quality reporting. A poorly designed comparison can withdraw resources from precisely those services that need improvement. Consider whether to fund support, use appropriate adjustment, report group-level performance and protect access while maintaining accountability.

• Specify the objective: A score should correspond to quality or outcomes the purchaser actually values.
• Align accountability: Attribute performance to a provider only when it can reasonably influence the measured result.
• Publish the rule: State denominators, weights, thresholds, timing and payment calculations in advance.
• Watch for gaming: Coding, exclusions and service avoidance can change scores without helping patients.
• Measure distribution: Examine effects on underserved groups and providers with different case mixes.
• Evaluate the policy: Distinguish a provider's bonus from the programme's net health and resource consequences.

Sources and further reading

The US Centers for Medicare & Medicaid Services Hospital Value-Based Purchasing overview is one concrete programme linking quality and payment; its rules should not be generalised to every purchaser. The OECD report on innovative provider payment models discusses payment design across health systems, and the AHRQ patient-safety discussion of unintended quality-reporting effects illustrates why narrow measures need balancing checks. The payment figures above are original fictional terms.