Prioritisation of Prevention Services

Prioritisation of prevention services: determining the

applicability of research from the US to the English context

Final Report Commissioned for Health England by Department of Health Policy Research Programme

Executive Summary

This study examined whether the methods used and results obtained in the prioritisation of clinical preventive interventions, conducted by the “Partnership for Prevention” in the US, are applicable in England.

This report to Health England:

describes the methods and results of the US study;
examines the use of the US results in policy and practice;
considers the generalisability of the US methods and results to England;
considers whether the methods for clinical preventive interventions could be extended to community preventive interventions;
considers the options for adapting the US methods to determine priorities for public health in England.

The US prioritisation took as a starting point a ‘long list’ of interventions categorised as effective by the United States Preventive Services Task Force and the Advisory Committee on Immunization Practices. A systematic literature review on effectiveness, costs, cost-effectiveness and burden of disease was undertaken for each intervention using a pragmatic, hierarchical search strategy. From published data and studies, models for 25 interventions were constructed for cost-effectiveness (CE) and clinical preventable burden (CPB) based on the following definitions:

CPB = burden of disease multiplied by effectiveness;
CE = average net cost per QALY gained in a typical practice by offering the service at recommended intervals to a US birth cohort over the recommended age range compared with no service provision.

To ensure consistency across interventions and evaluations the US exercise:

used QALYs;
measured total potential health effects if delivered to 90% of the target population;
explicitly accounted for patient adherence;
related CPB and CE to a hypothetical birth cohort of 4 million;
assumed services were delivered over the life time expected;
standardised costs to US$ year 2000 and through a 'reference case' method.

Results were presented as quintile ranks for CPB and CE and summed to an overall score, so that 10 (the maximum score) reflected the most important priorities. Three interventions (discussing aspirin use with high-risk adults, childhood immunizations and tobacco-use screening and brief intervention) had scores of 10 and were cost saving. Eight others had combined scores of seven or more. For most Americans, the 25 interventions were already potentially available and were reimbursed by public or private insurance. The results implied that higher ranked interventions should be prioritised and uptake encouraged. The results have been used by employers, health planners, policy makers, consumers and the pharmaceutical industry.

To date, the US exercise has been restricted to clinical preventive services offered on a one-to-one basis in a health-care setting. The extension of the US exercise to community-based preventive services is currently under consideration.

The US exercise is of interest because there appears to be no formal, explicit and transparent process in England that prioritises across the range of clinical preventive services, let alone across community preventive interventions. Whilst evaluations of individual interventions or programmes by the National Institute for Health and Clinical Excellence (NICE), the Joint Committee for Vaccines and Immunisation (JCVI) and the National Screening Committee (NSC) are transparent, the processes of prioritisation are less clear. Although prioritisation by the Department of Health and by Ministers may be explicitly undertaken, it is not externally transparent.

A prioritisation process is needed across the range of clinical and community preventive interventions in England and it should account for the detailed evaluations already undertaken within English organisations. Its formalisation might encourage greater consistency between organisations and might also increase the usefulness of evaluations of public health interventions and programmes to a prioritisation process.

We conclude that an explicit and transparent prioritisation process is needed to set public health priorities in England. However, neither the methods nor the data used in the US exercise are sufficiently relevant to transfer to the English context. The main difficulties are:

differences in disease prevalence and incidence and risk factor distributions between the US and England;
differences in the detailed specificiation of interventions for the US exercise and services provided in England;
the need for resource use and cost data relevant to England;
differences in the perspective of the economic evaluation to that explicitly required by NICE;
the need to consider marginal rather than average cost-effectiveness ratios for services that might require expansion or contraction;
differences in, and the need for explicitness about, cost-effectiveness threshold values in England;
the need to include stakeholders, public, patients and service users in the evlauation process;
the absence of explicit consideration of equity in the US exercise.

If the results cannot be used directly, a variety of intermediate positions could still be adopted as part of a separate exercise for England. In the absence of an existing ‘long-list’ in England, the US list could form the starting point for an English exercise. We conclude that, whilst a long-list for clinical services for England would probably be similar to one devised for the US (for example including similar interventions but slightly different specifications), an English ‘long list’ for community-based prevention services may be very different, given differences in social and community structures and preferences between countries. However, where similarities in services exist, the US reviews of evidence (for clinical services, and possibly in future for community services) could provide a useful short cut or at least starting point.

The second possibility is that the methods of the US exercise could be used in England. However, England has more experience of the explicit use of cost-effectiveness to guide priority setting and, in the case of NICE, has developed and adopted a detailed specification of methods. Moreover, we question the validity of summing ranks for CE and CPB, which has the result that some cost-ineffective services are prioritised. CPB is not typically used as an adjunct to economic evaluation. However, we recognise that information on marginal CPB could inform decisions as to which cost-effective interventions should figure most prominently in a focussed national policy. For example, if an ‘effort constraint’ exists (e.g. gaining ministerial approval and stakeholder acceptance), and there is a limited capacity for adopting policy changes each year, then from a set of cost-effective interventions it may be appropriate to focus national policy change and implementation efforts on those diseases with the largest CPB. However, equity considerations would also be taken into account in English policy making.

The US prioritisation exercise offers useful lessons but a separate exercise would be needed for England. Because of the challenge of achieving consistency and the likelihood that there would be incomplete data on compliance and equity, for example, existing models would have to be adapted and new models developed, particularly for community preventive interventions.

Health England needs to decide what criteria it would use, and to consider the trade-off between a transparent, rigorous and inclusive process (such as undertaken by NICE) and a more informal expert-opinion-based process that could make judgements around the US evidence (and indeed from other such exercises elsewhere). The latter may be quicker and less costly, but perhaps with less subsequent acceptance by stakeholders.

We suggest that a two-stage process may be necessary. First, to identify and promote a small number of priorities for which both the evidence and the professional consensus is strong. For this stage, the US prioritisation exercise could provide useful pointers. Second, to begin a broader, more detailed and rigorous exercise that would form the basis of an ongoing, robust and transparent priority setting exercise that would not replicate the US methods although would build on their valuable experience.

The Final Report can be viewed here.

Click here to view appendices.

Appendix A: List of participants. Appendix B: List of people interviewed. Appendix C: Publications from the US 2001 prioritisation exercise. Appendix D: USPSTF, PfP service definitions & UK policy documents. Appendix E: Publications from the US 2006 prioritisation exercise. Appendix F: Search strategies used in US 2006 prioritisation exercise. Appendix G: Case studies of CE and CPB. Appendix H: Review of academic literature. Appendix I: Case studies of 3 preventive services in US and UK. Appendix J: Conceptual factors causing variability in economics studies.