Andrew Riley discusses the influential factors in using a programme budgeting and marginal analysis model to manage growth in prescribing costs


The NHS spends £8 billion a year on prescription drugs in primary care in England. Expenditure on drugs used in this setting has risen by 60% in real terms over the last decade, and the number of items dispensed has increased by 55%. The continued development of new drugs for use within the NHS, the identification of new applications for existing drugs, and the ageing population mean that further growth can be expected.1

This inflation in prescribing costs was previously contained within the annual increase in overall healthcare costs, which has been identified in past Comprehensive Spending Reviews (CSR), but the last CSR has set a deficit-reduction agenda and so an annual increase in the NHS budget will be less in future years. Therefore any inflationary pressure will need to be managed from within existing allocations.

The White Paper, Equity and excellence: liberating the NHS, places the responsibility for controlling and managing prescribing spend in the hands of GP commissioning consortia.2 Many GPs are familiar with the need to seek out efficiency savings and will have been actively engaged in initiatives linked to previous prescribing incentive schemes led by medicines management in their local PCT. The future challenge will be to create efficiency savings that:

  • contribute toward deficit reduction
  • enable new and effective, branded medicines to be funded from a fixed resource that covers the cost of all medicines.

The challenge for GP commissioning consortia will be to manage their prescribing budget across all of their constituent practices, but this will only be possible if individual practices are actively managing their own prescribing costs—to do this they will need to carefully focus on their own spend.

Programme budgeting and marginal analysis

Budgeting is often associated with loss of freedom and centrally controlled processes limiting what can and cannot be done. Therefore it is difficult to sell the benefits of a budget to those GPs who have developed an instinctive ability to manage the rapidly changing needs of their practice populations as 'firefighters', rather than careful budget managers.

Many large practices now have a prescribing budget in excess of £1 million and monitoring spend at this level is fraught with difficulty as it is too undifferentiated and efficiencies achieved in one area are quickly consumed by unforeseen drug spend in another.

Programme budgeting and marginal analysis (PBMA) is a process of choosing a subset of programmes that represent a significant proportion of overall spend and in which funding pressures have been seen to arise previously, pooling the spend in each of these as named 'programme budgets', and scrutinising historic spend to identify opportunities for efficiencies to be targeted and savings to be delivered.

When the PBMA process is applied to the management of prescribing budgets, it can help practices to work with local GP commissioning consortia support managers to maximise their use of healthcare resources by making prescribing budgets work harder to meet the assessed needs of their local populations. In the context of prescribing, this means managing-in new branded medicines appropriately and managing-out brands that have reached the end of their brand life cycle so that well-established 'cash-cow' brands deliver windfall savings when they become available to NHS prescribers as generics.

Programme budgeting—is a retrospective appraisal of resource allocation, broken down into meaningful programmes, with a view to tracking future resource allocation in those same programmes.3

Marginal analysis—is the appraisal of added benefits and higher costs when new investment is proposed (or lost benefits and lower costs when disinvestment is proposed), in an incremental way.3

Implementation of prescribing programme budgets

Prescribing budgets have evolved gradually over many years to meet the prescribing needs of the practice population so any significant adjustment needs to be phased in. The composition of the current prescribing budget and areas that have grown quickly in the past few years should be determined. There are six key prescribing budget areas corresponding to six of the chapters in the British National Formulary (BNF). These chapters represent 80%–90% of prescribing spend and so, programme budgets should be set for each of these areas:4

  • Cardiovascular
  • Central nervous system
  • Endocrine
  • Gastrointestinal
  • Nutrition and blood
  • Respiratory.

Budget controls and incentives

Two controls are used to provide a focus for GP practices once prescribing programme budgets are set for them:5

  • Financial target—a tolerance threshold of plus or minus 2.5% is set as a target for each budget
  • Clinical target—a target of 90% compliance with the prescribing formulary is set.

At NHS Stoke-on-Trent, the financial and clinical targets are monitored and progress is fed back to practices in a monthly report prepared by the Department of Medicines Management at Keele University, which has been commissioned to prepare the monthly practice and consortia reports.

Formulary
The formulary is jointly owned by the GP commissioning consortia on behalf of their constituent practices, the acute trust, mental health trust, and the community partnership trust. Its purpose is to provide guidance to GPs on which drugs offer the best value for the NHS without compromise on the quality of clinical outcomes. The formulary emphasises the availability of generic medicines for the majority of patients. It is designed to maximise the cost effectiveness of prescribing and thereby achieve budget control.

The formulary is web-based and updated after area prescribing committee meetings. Point of prescribing decision support software is loaded onto the GP clinical system to encourage high compliance. Compliance is set at 90%, which offers freedom to use an off-formulary drug for the small minority of patients whose clinical needs will be better met by a newer brand while providing a ceiling for prescribing spend to be contained within it.

Governance, oversight, and incentives
A medicines management steering group of GPs provides clinical leadership, financial oversight, and governance of any incentive linked to the active management and successful delivery of a balanced programme of investment across the six key prescribing budget areas (the BNF chapters as discussed above). Practices are also shown the progress of neighbouring practices in the monthly progress report as they have agreed to share comparative data down to practice level across the consortium.

The incentive for practices to make prescribing changes to achieve control of their prescribing programme budgets can take the form of payments from a prescribing incentive scheme, items of service fees in a medicines management local enhanced service (LES), or as part of the new prescribing indicators in the revised quality and outcomes framework for 2011/12.6

These incentives and reports are needed to provide:

  • a focus
  • early impetus and ongoing momentum for widespread acceptance
  • adoption of the controls that support the establishment of the new prescribing programme budgets.

Practice prescribing analyses

If the monthly practice-monitoring report shows that programme budgets are projected to exceed the agreed tolerance targets, medicines management prescribing analyses are used to identify prescribing changes that will realign budgets. Often this will simply involve focusing on prescribing more generic medicines and consideration of switching to generics from premium-priced brands when this is clinically appropriate.

Medication review

As well as raising awareness of the availability of low-cost generic medicines for treating new patients, it is sometimes necessary for practices to reappraise their use of premium-priced brands for existing patients. A medication review provides a basis for doing this in a systematic way and GPs can replace a brand with a generic when there is a rational, evidence-based clinical justification for making this change.

A systematic approach to medication review designed to reduce overall prescribing costs consists of four clinical interventions:

  • Step back—step back can be justified clinically when a step in a clinical pathway is skipped, leading to a branded medicine being used prematurely before a generic medicine that is next on the treatment algorithm. For example, if the following algorithm has been agreed: A – B – C – D. In this case, a medication review is needed if a patient's treatment history reveals that they were treated in the order A – B – D (i.e. step C was skipped). The review will involve stepping back down the algorithm from D to C. Drugs that are located first in treatment algorithms tend to be well-established cost-effective generics
  • Step down—step down can be justified when drug treatment has been intensified by either increasing drug dose and/or using more than
    one drug in an intensified combination to achieve a clinical target (e.g. clinical indicator in the quality and outcomes framework). Once this target has been achieved and maintained it may be clinically justified to step down in treatment dose or withdraw one or more drugs in an intensified regimen if there is a low risk of losing clinical control
  • Step off—stepping off a treatment pathway temporarily can be considered when the continued use of the drug ceases to be clinically justified. Treatment can be stopped temporarily for a 'treatment holiday' and if the original symptoms return, treatment can be resumed. Treatment holidays are a useful step-off device when the original reason for commencing therapy is unclear from the patient's medical records
  • Stop—in some circumstances stopping drug treatment can be justified if multiple drug therapies put the patient at excessive risk of iatrogenic disease. Adverse reactions have been linked with up to 17% of hospital admissions in older people.7 Stopping treatment is ethically acceptable when the risk of continuing treatment exceeds the likely benefit of continuing to prescribe it and the patient's GP is responsible for making the decision to stop a drug.

Enhanced medication review specification
The three-tier medication review service allows all practices to provide the enhanced medication review service. All medication reviews are clinically managed by the GPs taking part in the service and the governance to ensure changes are clinically appropriate and specified clearly within the contract that all GPs providing the service are required to sign.

Tier 1: pre-specified review

This tier relies on the identification of patients for review whose treatment does not correspond with local prescribing guidelines, pathways, and policies. General practitioners will conduct a review if they believe that there is scope for changing medication in order to improve cost-effective prescribing. An example of a step-1 intervention is a reduction in the strength of combination inhalers used in asthma management; this encourages 'step down', which is clinically appropriate (and supported by British Thoracic Society/SIGN guideline on asthma) for patients with stable asthma.

Electronic medication review templates can be used for GP medication review and are populated with a dataset specific to the review in question. Some of the dataset will fall into 'mandated fields' to ensure that all necessary information has been considered during the medication review. This is essential to demonstrate to the payer that the review has been conducted safely. Review at this tier is targeted and support can be given to identify patients who are candidates for review and to accelerate the process so that changes can be quickly adopted and hence rapidly converted into efficiency savings.

Tier 2: targeted review
A practice-prescribing analysis is used to identify patients for medication review and a lead GP takes responsibility for carrying out the reviews to maximise the productivity of the service. Both Tier 1 and Tier 2 reviews have the advantage of being able to calculate the savings that can be achieved and as the workload can be predicted, the service more carefully specified.

Tier 3: free-form review
The review template for a tier 3 review can be programmed with a drop-down list of 'drugs moved from' and a drop-down list of 'drugs moved to'; this can be linked to an up-to-date drug tariff so that the cost reduction is automatically calculated and collated for reconciliation. If the review is incentivised on an item-of-service basis, these collated review templates can be submitted to the payer as invoices.

This tier of review is appropriate when a practice shows good compliance with local treatment pathways and the prescribing analysis fails to identify significant opportunities for targeted review. In this tier, GPs are able to identify their own targets for medication review in those programme budgets that show a projected overspend in order to bring the budget back within agreed tolerance limits.

Post-payment verification
Medication review is a face-to-face clinical intervention and it is necessary for the commissioner to ensure that there is a means of verifying that the review has been conducted properly before authorising payment. There is a critical need to improve the efficiency of post-payment verification for patients facing professional services such as enhanced medication review.

Post-payment verification is made much simpler by building an electronic claims form, which is held on a username- and password-enabled webpage. Details need to be completed on the claims form by the GP conducting the review; mandatory fields are included on the form so that the nature of any changes made to medication can be easily linked to the patient concerned.

The electronic claims form uses a drop-down search function to select the drug that is the subject of the review, and once the drug change (step back, step down, step off, or stop) has been chosen, the template is programmed to automatically calculate the monthly efficiency saving.

Reimbursement
Cash release that accrues from medication review can be used to reimburse the GP intervention(s) required to achieve the change that generates the savings in the first place. An example of a locally agreed reimbursement based on this payment-by-results model is that payment is made from the equivalent of 2 month's worth of savings. This payment can also be paid quarterly as efficiency savings accrue and this will add momentum to the service and drive further efficiency savings.

Simultaneous active management of several programme budgets will drive significant in-year cash release, which is essential if programme budgets are to be stretched to meet the needs of the responsible population of a GP commissioning consortium with a tightly controlled budget.

Local enhanced medication review service

Medication review is included in the GMS contract6 as a means of reappraising the need to continue a drug and this usually involves reauthorisation of repeat treatment. Locally enhancing the specification for medication review provides an opportunity to refocus this professional activity toward rationalising previous prescribing. A local enhanced medication review service provides the momentum for marginal analysis and rationalisation of prescribing spend to enable programme budgets to be actively managed and brought back into balance. For this reason, and in order to identify local targets for review, item-of-service payments for conducting a locally enhanced medication review are negotiated between a GP consortium and local practices; and the medication review is specified in an LES contract.

Payment-by-results service model

Incoming electronic claims are collated by medicine management and payments are activated 3 months from the date of the claim. This ensures that any changes made to medication are sustained and the associated efficiency saving is available to fund the intervention that created it. This payment-by-results service model eliminates the risk of commissioning a service that requires a significant capital investment by the GP commissioning consortium, which would diminish the size of the net efficiency saving that can be achieved by the service.

Payments are made to practices each quarter to stimulate active participation in the service and to create a momentum behind it. Practices are also given feedback about the cumulative savings that they have achieved, and the overall efficiency saving in a locality is shared between all the practices in the locality in an update report.

Critical success factors

Previous attempts to use an evidence-based PBMA model for healthcare resource allocation have identified four critical success factors:8

  • Specification of the budget constraint
  • Scope of the programme budget
  • Composition and role of the advisory group
  • Incentives for/against contributing to a 'shift list' (medicines for which there is a like-for-like cheaper equivalent that has been included in the formulary) of options for disinvestment and resource release.

Specification of the budget constraint
In the PBMA model, the general practice prescribing allocation is broken down into the six BNF chapters as described previously, and a tolerance threshold of ±2.5% is set by the medicines management subgroup as a target for each programme budget. The aim is to establish how the budget is spent by using a forecast projection as the prescribing data comes through each month and for individual practices to relate their spend in each of the six BNF chapters with the prescribing allocation that has been made at the start of a financial year and their forecast year end out-turn.

Committing to managing primary care prescribing spend down at this level introduces a new 'affordability' context to requests from secondary care specialists asking GPs to fund high-cost drugs, the cost of which has not properly been accounted for at the start of the year.

Scope of the programme budget
The formulary should provide a list of drugs for GP prescribing, the use of which has been accounted for in each programme budget. The availability of generic drugs is fully exploited in the formulary across all programme budgets and the entire cost of any new brands that have been approved for inclusion in the formulary needs to managed in. This is carried out with a planned package of medication reviews that will offset the

cost of any new medicines. The cost of these new medicines is modelled, anticipated, and factored into the practice programme budget area in which they belong to, in advance, at the start of the financial year.

If the budget is going to be put at risk by prescribers beginning to use newly approved drugs, a cost–offset plan needs to be agreed by identifying equivalent efficiency savings in a prescribing analysis. These savings need to be actively pursued to avoid an overspend in the programme budgets in which the new drugs are going to be prescribed.

Composition of the advisory group
General practitioners need to be convinced that the budgetary constraints introduced by the PBMA model are not too restrictive and allow them to not only respond effectively to the needs of their patients, but to also provide preventive treatment as part of local public health initiatives designed to reduce health inequalities. Therefore the advisory group should consist of GPs who have been appointed as prescribing and clinical leads within the consortium. They should be supported in their decision making by data and analyses provided by medicines management and finance experts who have the experience to strike the right balance between financial restraint and high-quality evidence-based prescribing choices. Public health and specialist clinician expertise will also need to be co-opted when needed.

General practitioners with a special interest may have prescribing patterns skewed by their interest; for example, doctors with a special interest in diabetes may use more insulin than other practices in the consortium, and their choice of insulin should be guided by the local diabetes pathway in which prescribing choices are influenced by the principles of cost effectiveness and evidence-based medicine.

Incentives for disinvestment

The feasibility of moving towards medicines with higher cost effectiveness needs to be carefully considered by the steering committee. Only those changes that are within reach of a GP in primary care and are generally applicable across all practices should be proposed for inclusion in the local medicines management strategy for improving cost effectiveness of prescribing across the consortium. Changes that carry excessive risk should not be attempted in primary care unless specialist supervision and support is available for patients (e.g. treatments that need careful individual monitoring, such as changing insulin). An LES that self-funds a higher specification medication review provides an impetus for GP-led marginal analysis of the prescribing budget; this allows for identification of opportunities to appropriately reduce prescribing spend without diminishing the quality of clinical outcomes that are achievable in primary care.

Incentives are necessary to maintain momentum behind a drive toward improved cost effectiveness especially when the prescribing that is being targeted for disinvestment is not necessarily wrong, but is disproportionately expensive compared to the generic alternative.

Summary

In an environment of cost constraint, new service development will rely on reinvestment of savings generated from disinvestment. The model of prescribing programme budgeting presented in this article provides an opportunity for primary care to manage prescribing costs by:

  • breaking down the allocated budget into six named prescribing programme areas budgets
  • introducing the appropriate and proportionate incentives of a clinical LES and high-quality clinical support from medicines management
  • driving appropriate clinical change, which will improve the cost effectiveness of prescribing overall.

The success of a GP commissioning consortium implementing a PBMA model prescribing strategy will rely on dogged medical leadership, supported by medicines management, and financial expertise to effect the change from a resource-led, clinically informed discussion to a clinically led, resource-informed discussion.

Acknowledgments

The views expressed in this article are those of the author and not those of his employing authority.

Key Points
  • Use of new effective branded medicines needs to be planned into the programme budget in which they belong and if necessary, cost efficiencies need to be identified to offset this new investment
  • The prescribing programme budgets for six of the chapters in the British National Formulary cover at least 80% of the overall prescribing budget and include all the areas of financial risk
  • Financial constraints and clinical targets are set and agreed within consortia and if necessary these are phased in over several years to maintain clinician engagement and involvement
  • Incentives are necessary to drive the adoption of programme budgets and to compensate GPs for the work required to provide medication reviews that will help to deliver a balanced budget and address observed variation in clinical practice
  • For some practices it may be necessary to adopt strategies for more rapid adoption of prescribing policies, which in some cases will include the switching of like-for-like, brand to generic medicines
  • The purpose of medication review is to reduce prescribing costs when it is safe and appropriate to do so; this will enable more patients to be treated and new drugs to be used when formulary generic drugs have already been tried or are considered inappropriate
  • The three-tier medication review service allows all practices to participate in the service, with practices earning increased autonomy once they have developed the skills, competency, and capacity to provide such a service
  • The efficiency of post-payment verification needs improving for patients facing professional services (e.g. enhanced medication review)
  • A local enhanced medication review service provides the momentum for marginal analysis and rationalisation of prescribing spend to enable programme budgets to be actively managed
  • There are four critical success factors in the use of an evidence-based programme budgeting and marginal analysis model for healthcare resource allocation:
    • Specification of the budget constraint
    • Scope of the programme budget
    • Composition and role of the advisory group
    • Incentives for/against to a 'shift list' of options for disinvestment and resource release.

References

  1. National Audit Office. Prescribing costs in primary care. London: The Stationery Office, 2007. Available at: www.nao.org.uk/publications/0607/prescribing_costs_in_primary_c.aspx
  2. Department of Health. Equity and excellence: liberating the NHS. London: DH, 2010. Available at: www.dh.gov.uk/en/Publicationsandstatistics/Publications/PublicationsPolicyAndGuidance/DH_117353
  3. Brambley P, Fordham R. What is PBMA? Hayward Medical Communications, 2003. 4 (2). Available at: www.medicine.ox.ac.uk/bandolier/painres/download/whatis/pbma.pdf
  4. British National Formulary. BNF 61. March 2011. London: British Medical Association and the Royal Pharmaceutical Society of Great Britain, 2011.
  5. NHS Stoke on Trent. Local enhanced service report 2010/2011. NHS Stoke on Trent, 2010.
  6. British Medical Association. NHS Employers. Quality and outcomes framework guidance for GMS contract 2011/12. London: BMA, NHS Employers, 2011. Available at: www.bma.org.uk/employmentandcontracts/independent_
    contractors/quality_outcomes_framework/qofguidance2011.jsp
  7. Audit Commission. A spoonful of sugar: medicines management in NHS hospitals. London: Audit Commission, 2001. www.auditcommission.gov.uk/nationalstudies/health/other/Pages/aspoonfulofsugar.aspx
  8. Mortimer D. Reorienting programme budgeting and marginal analysis (PBMA) towards disinvestment. BMC Health Services Research 2010; 10: 288. Available at: www.biomedcentral.com/1472-6963/10/288 G