This promotional information is intended for healthcare professionals based
in the UK.
If you are not a healthcare professional in the UK, click here.
For Prescribing and Adverse Event reporting information, click here.

ELIQUIS® (apixaban) is indicated for prevention of stroke and systemic embolism in adult patients with non-valvular atrial fibrillation (NVAF), with one or more risk factors, such as prior stroke or transient ischaemic attack (TIA), age ≥75 years, hypertension, diabetes mellitus, symptomatic heart failure (NYHA Class ≥II); treatment of deep vein thrombosis (DVT) and pulmonary embolism (PE), and prevention of recurrent DVT and PE in adults; and, prevention of venous thromboembolic events (VTE) in patients who have undergone elective hip or knee replacement surgery.

NVAF – Real World Data

ELIQUIS® (apixaban) RCT results, complemented by real-world evidence1,2

Real-world evidence complements RCTs by providing key insights into diverse clinical settings and patient populations3–5

Summary of differences between clinical trials vs. real-world studies5

RCTs are the ‘gold standard’ for determining the efficacy and safety of a medicine, while real-world evidence can provide insight into the effectiveness in routine clinical practice – together, these can support treatment decisions3–5

Rationale and study overview

ARISTOPHANES was a large-scale retrospective, US real-world analysis examining both DOAC vs. warfarin, and DOAC vs. DOAC in 434,046 propensity score-matched patients with NVAF.2

Study population:2

  • Adult patients with a diagnosis of NVAF and a first OAC prescription in the study period (1 January 2013 to 30 September 2015)* and no OAC exposure for 12 months prior
  • Pooled population from US Medicare and four US commercial claims databases


  • To compare the rates of stroke / systemic embolism and major bleeding among patients with NVAF who were newly prescribed ELIQUIS, rivaroxaban, dabigatran, or warfarin


  • Effectiveness outcome: time to first stroke / systemic embolism (including ischaemic stroke, haemorrhagic stroke, and systemic embolism)
  • Safety outcome: time to first major bleed (including gastrointestinal bleeding, intracranial haemorrhage, and major bleeding at other sites)


  • Retrospective cohort analysis, using Cox proportional hazard regression to evaluate outcomes across propensity score-matched cohorts§
  • Follow-up period: 30 days after discontinuation date, switch date, death, end of continuous medical or pharmacy plan enrolment, or end of study period (whichever occurred first)
  • Additional analyses included patient subgroup and dose-specific outcomes

What are the limitations of the ARISTOPHANES real-world analysis?2

  • Only statistical association rather than causal relationships could be inferred
  • Although cohorts were propensity-score matched, potential residual confounders exist, which are not available in the dataset – this limitation is especially important for interpreting DOAC vs. DOAC, which are for hypothesis generation given the lack of head-to-head trials
  • Outcomes were identified using ICD-9-CM codes, without further adjudication using precise clinical criteria or further validation against healthcare providers’ medical records
  • There was no evaluation of the dose-reduction criteria for DOACs due to the absence of comprehensive data on body weight or serum creatinine / creatinine clearance
  • Laboratory values, such as INR measurements, are not available in the database, therefore it is not possible to determine time in therapeutic range for patients prescribed warfarin
  • Many clinically important outcomes, such as mild-to-moderate bleeding associated with oral anticoagulant use, were not evaluated because they cannot be reliably measured in claims databases
  • The study relied on prescription dispense records to characterise anticoagulant drug exposure, but patients’ actual drug-taking behaviours could not be measured
  • Observed and unobserved heterogeneity may exist across the 5 data sources
  • There is no guarantee that patients were dosed according to the US prescribing information for each DOAC and warfarin
  • Results may not be generalisable to the overall NVAF population in the US, because the study did not include uninsured patients and patients solely covered by other public health insurance plans

ELIQUIS was associated with a reduced risk of stroke / systemic embolism and a favourable bleeding profile vs. warfarin in real-world settings2

There are no head-to-head RCTs comparing the DOACs.

As ARISTOTLE is a clinical trial and ARISTOPHANES is a real-world analysis, the two studies cannot be directly compared.

The results of the ARISTOPHANES real-world analysis should only be used for hypothesis generation and must be interpreted with caution, as the effectiveness and safety outcomes and assessed patient populations may be different to those in the ARISTOTLE clinical trial.

Additional real-world data studies are available. Please click GET IN TOUCH to contact your local BMS / Pfizer representative for more information.


* End date was selected as it was the last date before the US moved from ICD-9 codes to ICD-10 codes. Due to this coding switch, it is believed that the researchers wanted to keep the outcome definitions consistent.

Data in this study were pooled from the US Centres for Medicare and Medicaid Services Medicare data, and 4 commercial claims databases in the US: Truven MarketScan®, IMS PharmMetrics Plus, Optum® Clinformatics® Data Mart, and the Humana Research Database.2 Though commercial datasets may contain duplicate patient records when pooled, the number of duplicate is likely to be small (~0.5%) and therefore unlikely to affect the results.8

‡ Outcomes were based on hospitalisations with stroke / systemic embolism as the principal or first-listed diagnosis.2

§ Patients were matched 1:1 in each dataset, using propensity scores generated by logistic regression based on demographics, CCI score, baseline bleeding and stroke / systemic embolism history, comorbidities, and baseline co-medications.2

The ARISTOTLE clinical trial was a randomised, double-blind, double-dummy, non-inferiority trial in 18,201 patients with NVAF and at least one other risk factor for stroke. Stroke / systemic embolism, including haemorrhagic, ischaemic or undetermined stroke, was the primary efficacy endpoint, major bleeding according to ISTH criteria was the primary safety outcome and all-cause mortality was a key secondary endpoint.1 Pre-specified hierarchical sequential testing was performed first on stroke / systemic embolism (primary efficacy endpoint) for non-inferiority, then for superiority, then on major bleeding, and finally, on death from any cause (secondary endpoint).1

# In the ARISTOPHANES retrospective, US real-world analysis, the primary effectiveness outcome was time to first stroke / systemic embolism, including ischaemic stroke, haemorrhagic stroke, and systemic embolism. The primary safety outcome was time to first major bleed, including gastrointestinal bleeding, intracranial bleeding, and bleeding from other sites.2 Outcomes were based on hospitalisations with stroke / systemic embolism or major bleeding listed as the principal or first-listed diagnosis, as identified using ICD-9 codes.2

** After propensity score matching in the ARISTOPHANES retrospective, US real-world analysis, (ELIQUIS vs. warfarin), 75.4% of ELIQUIS patients received the standard dose of ELIQUIS 5 mg BD and 24.6% of patients received a reduced dose of ELIQUIS 2.5 mg BD.2

BD = Twice Daily   CCI = Charlson Comorbidity Index   CI = Confidence Interval   
DOAC = Direct-acting Oral Anticoagulant   HR = Hazard Ratio   GI = Gastrointestinal   ICH = Intracerebral Haemorrhage   ICD-9 = International Classification of Diseases, Ninth Revision   INR = International Normalised Ratio   ISTH = International Society on Thrombosis and Haemostasis   N = Total number of patients in either the ELIQUIS group or the warfarin group   n = Number of patients with event   NVAF = Non-Valvular Atrial Fibrillation   OAC = Oral Anticoagulant   
RCT = Randomised Controlled Trial   RWE = Real-World Evidence


  1. Granger CB et al. N Engl J Med 2011; 365: 981–992.
  2. Lip GYH et al. Stroke 2018; 49: 2933–2944.
  3. Camm AJ et al. Open Heart 2018; 5: e000788.
  4. Blonde L et al. Adv Ther 2018; 35: 1763–1774.
  5. Kim HS et al. J Korean Med Sci 2018; 33: e213.
  6. ABPI. The vision for real-world data: harnessing the opportunities in the UK. September 2011.
  7. ABPI. Demonstrating value with real world data: a practical guide. May 2011.
  8. Lip GYH et al. Stroke 2018; 49: 2933–2944. Online supplement.