Getting cost-effective technologies into practice: the value of implementation. An application to novel anticoagulants in the prevention of stroke and systemic embolism

Key findings

There appears to be value in additional implementation efforts directed towards encouraging the utilisation of novel anticoagulants for the prevention of stroke and systemic embolism in people with nonvalvular atrial fibrillation. This additional value can be represented both in terms of health and monetary benefits to the NHS. At a cost-effectiveness threshold of £20,000 per QALY gained, additional investment in an education outreach activity that increases utilisation by 5% (absolute increase in uptake rates) would generate an additional 71 QALYs (£1.42 million in terms of a monetary equivalent) across England and Wales compared to the use of these resources in other (health generating) NHS activities. In an average clinical commissioning group (CCG) population, the equivalent gains would be 0.16 QALYs (£3,161).

There remain issues about the heterogeneity and quality of the existing evidence-base surrounding the effectiveness of alternative implementation strategies and their generalisability to the specific case study. Consequently, it has not been possible to compare a range of alternative implementation activities, and educational outreach was selected as a specific example. However, the framework can also be applied more generically to indicate the level of uptake that would be required per £ spent by the NHS in order for an implementation activity to provide additional value to the NHS. At a threshold of £20,000 per QALY gained, an activity costing one million pounds across England and Wales (or £4,095 for the average CCG) would need to increase utilisation by at least 1% to be considered of value to the NHS. These can provide indicative estimates which could be applied to a broader range of implementation activities.

The findings also indicate that the value of implementation appears highest in targeting efforts to increase utilisation of novel anticoagulants in patients with average to poor warfarin control. For example, an increase of 5% in utilisation in such patients could potentially generate an additional 790 QALYs (£15,798,276) across England and Wales and 3.06 QALYs (£61,187) in an average CCG population. This is more than ten times the value of increasing utilisation in the overall population on warfarin.

Most importantly, greater (absolute) value to the NHS would potentially be achieved with higher uptake of anticoagulation more generally (i.e. including warfarin) given the high proportion of patients with atrial fibrillation who are currently receiving no treatment or antiplatelet therapy only. Switching 5% of patients potentially eligible for anticoagulation but currently on no treatment or on antiplatelet therapy to warfarin would generate an additional 7,550 QALYs (£151,004,965) across England and Wales and 30 QALYs (£606,866) in an average CCG population.

Additional investment on implementation at a CCG level also needs to take into account local circumstances and particularly the local cost of warfarin clinics. The results presented here are based on an annual per- patient cost of £242 as applied in the original NICE appraisal. Lower warfarin monitoring costs reduce the value of implementation of novel anticoagulants; hence, there is less scope for investment in increasing their utilisation. Conversely, higher value would be realised in local settings where the current annual per-patient costs of providing warfarin clinics exceed these estimates.

1.4. Implications for NICE technology appraisals

The framework and its application to novel anticoagulants also highlight issues which may have wider implications for the NICE technology appraisal and implementation processes. Specifically:

(i) The importance of clearly documenting the cost-effectiveness (incremental cost and incremental QALY) estimates based on the Committee’s preferred assumptions both for the entire reimbursed population and also relevant subgroups. This would increase the efficiency of subsequent implementation activities by ensuring that their type and intensity can be tailored to the areas with the greatest value to the NHS.

(ii) Difficulty in defining the optimum (target) utilisation rate for interventions, particularly when a intervention is appraised through the single technology appraisal process and recommended as an ‘option’. When several alternative interventions are listed as options, it may be difficult for local commissioners to determine where additional implementation activities should be targeted and determining the appropriate level of investment. In situations where there exist high levels of uncertainty regarding the optimal treatment pathway and/or position of specific treatments and this is perceived as a significant barrier to uptake, referral to the NICE multiple technology appraisal process or clinical guidelines may be a potential route to encouraging wider (and more efficient) implementation.

(iii) Recognition that the separation of budgets over primary and secondary care may present additional challenges to the NHS and local commissioners in implementing particular interventions (and/or impose additional transaction costs which should be considered). More routine reporting of total costs (and longer-term cost offsets) disaggregated by setting (e.g. primary, secondary care) in NICE Technology Appraisals (TAs) would provide an indication of where additional transaction costs may be incurred and/or where financial incentives maybe required to encourage further uptake.

(iv) Recognising that the long-term time horizon underpinning the cost-effectiveness estimates reported in many TAs maybe inconsistent with the local constraints faced by commissioners (e.g. shorter budgetary cycles operated by the NHS). Encouraging the time profile of the cost-effectiveness results to be more routinely reported (and particularly the time horizon over which a particular technology is expected to ‘break even’ and start to confer additional value to the NHS) would provide further information to local commissioners in planning implementation activities and in identifying which specific investments provide clear long-term value but may be difficult to implement within existing constraints (i.e. the requirement for CCGs to breakeven year on year). At both a national and local 7 level, this could inform whether re-profiling of long-term investment prospects may be required to ensure greater consistency between national and local constraints.