Skip to main content


Technology Guidance - Drug

Novel oral anticoagulants (NOACs) for the prevention of stroke and systemic embolism in non-valvular atrial fibrillation First published on 3 May 2017

Guidance Recommendations
The Ministry of Health’s Drug Advisory Committee has recommended:
  • Rivaroxaban 15 mg and 20 mg tablets, and apixaban 2.5 mg and 5 mg tablets for preventing stroke and systemic embolism in patients with non-valvular atrial fibrillation (NVAF) and:
    • CHA2DS2-VASc score of 1 or more for men; and
    • CHA2DS2-VASc score of 2 or more for women.

Rivaroxaban or apixaban should not be used in patients with valvular AF (especially rheumatic mitral stenosis), or patients with prosthetic heart valves.

Subsidy status

Rivaroxaban 15 mg and 20 mg tablets, and apixaban 2.5 mg and 5 mg tablets are recommended for inclusion on the Medication Assistance Fund (MAF) for the abovementioned indication.

MAF assistance does not apply to dabigatran.

Get Acrobat Readeropens in a new window

Factors considered to inform the recommendations for subsidy

Technology evaluation

Point Item
1.1 The MOH Drug Advisory Committee (“the Committee”) considered the evidence presented for the technology evaluation of non-vitamin K antagonist oral anti-coagulation agents (NOACs – apixaban, rivaroxaban, and dabigatran) for preventing stroke and systemic embolism in NVAF in 2016. The Agency for Care Effectiveness (ACE) conducted the evaluation in consultation with the Ministry of Health NOACs Working Group members.
1.2 The evidence was used to inform the Committee’s deliberations around four core decision-making criteria:
  • Clinical need of patients and nature of the condition
  • Clinical effectiveness and safety of the technology
  • Cost-effectiveness (value for money) – the incremental benefit and cost of the technology compared to existing alternatives
  • Estimated annual technology cost and the number of patients likely to benefit from the technology.
1.3 Additional factors, including social and value judgments, may also inform the Committee’s subsidy considerations.
1.4 Manufacturers of NOACS (apixaban and dabigatran) that were not recommended for subsidy in 2016 on the basis of unacceptable cost-effectiveness or budget impact were invited to submit revised price proposals, which the Committee considered in April 2018.

Clinical need

Point Item
2.1 The Committee recognised that:
  • Prevalence of AF and the risk of stroke related to AF both increase as the population ages; and
  • NOACs and warfarin were considered first-line treatment for preventing stroke and systemic embolism in people with NVAF.

Clinical effectiveness and safety

Point Item
3.1 The Committee agreed that warfarin was the appropriate comparator for NOACs for people with NVAF who required anticoagulation.
3.2 The Committee acknowledged that warfarin, an effective treatment to prevent stroke, was associated with frequent drug-drug interactions, dietary restrictions, and the need for regular monitoring.
3.3 The Committee considered the clinical evidence from pivotal trials of the NOACs (ARISTOTLE [apixaban], RE-LY [dabigatran] and ROCKET-AF [rivaroxaban]) versus warfarin. It noted that apixaban (5 mg twice daily, 2.5 mg twice daily in some patients), dabigatran (150 mg twice daily and 110 mg twice daily), and rivaroxaban (20 mg daily, 15 mg daily in some patients) were as effective as warfarin in preventing stroke and systemic embolism.
3.4 In particular, the Committee noted better safety experienced with NOACs compared to warfarin with respect to reducing intracranial hemorrhage (ICH). Although the absolute risk reductions in ICH were small (ranging from 0.2% to 0.5% per year, or two to five ICH events avoided for every 1,000 patients treated per year), the Committee concurred with the clinical experts that this benefit was clinically significant because of high morbidity and mortality associated with ICH.
3.5 The Committee noted the lack of head-to-head trials comparing all three NOACs. It noted that:
  • the population in the study comparing rivaroxaban with warfarin (ROCKET-AF) had a higher mean baseline CHADS2 score, and a higher proportion of patients had comorbidities (heart failure, diabetes, and hypertension) than the population in RE-LY or ARISTOTLE; and
  • a lower proportion of patients in the apixaban study (ARISTOTLE) took concomitant aspirin compared to those in RE-LY or ROCKET-AF.
3.6 The Committee considered that the differences in baseline characteristics between study populations could lead to difficulties in interpreting the results of any indirect treatment comparison.
3.7 The Committee concluded that the NOACs could be considered comparable with no clinically important differences in outcomes.


Point Item
4.1 Cost-minimisation among the NOACs
Given all three NOACs were considered comparable, the Committee agreed a cost-minimisation approach was appropriate for selecting the lowest-priced NOAC for subsidy consideration. It noted at the 2016 meeting that rivaroxaban, which had the lowest cost, was the most cost-effective option.
4.2 In April 2018, following a revised price proposal for apixaban, the Committee agreed the cost of apixaban was reasonable and could be considered an acceptable use of healthcare resources. Dabigatran remained at a higher cost compared with rivaroxaban and apixaban and was the least cost-effective option.
4.3 Cost-effectiveness of NOACs versus warfarin
The cost-effectiveness model compared the NOACs to warfarin for stroke prevention in NVAF over a lifetime period. The Committee noted that at a selling price of [●] *, the base case incremental cost-effectiveness ratio (ICER) for NOACs compared with warfarin would fall in the range of less than $15,000 per quality-adjusted life-year (QALY) gained. It agreed that the ICERs were within an acceptable range of cost-effectiveness in sensitivity analyses. The Committee accepted that NOACs were a cost-effective treatment option compared with warfarin for stroke prevention in Singapore.
* Information redacted

Estimated annual technology cost

Point Item
5.1 The Committee estimated around 4,800 people in Singapore would benefit from government assistance for rivaroxaban and apixaban. The annual cost impact was estimated to fall in the range of $3 to $5 million at the prices proposed by the manufacturers. It was noted that, given the ageing population, the prevalence of AF, and the risk for AF-associated stroke increasing with age, patient numbers would increase in time.
5.2 The Committee also noted cost estimates were based on various assumptions. The exact rate of people switching from warfarin, aspirin, or no treatment to a NOAC can be difficult to predict as clinical knowledge evolves and new technologies are introduced.

Additional considerations

Point Item
6.1 The Committee agreed the NOACs should not be used in people with valvular AF (especially rheumatic mitral stenosis), or people with prosthetic heart valves, given these people have not been well-studied in clinical trials. Given the potential for inappropriate use of NOACs, the Committee recommended listing on the MAF rather than on the Standard Drug List (SDL).


Point Item
7.1 Based on evidence presented in 2016, the Committee recommended rivaroxaban 15 mg and 20 mg tablets be listed on the MAF for preventing stroke and systemic embolism in patients with non-valvular atrial fibrillation (NVAF) who meet certain clinical conditions, given its superior reduction in ICH and acceptable cost-effectiveness at the price proposed by the manufacturer, compared with warfarin.
7.2 In April 2018, the Committee also recommended apixaban 2.5 mg and 5 mg tablets be listed on the MAF in line with the same clinical criteria as rivaroxaban, following an acceptable price discount offered by the manufacturer.
Back to Top