How real is real world data? - lessons from CVD Real study

Dr Chee L Khoo

Have you notice that our patients never really behave like patients in clinical trials? That is to be expected as subjects in clinical trials are carefully selected, extensively supported, followed up closely during the entire trial period. There is hardly any chance of non-compliance as every single tablet is accounted for and subjects are prompted to ensure they don’t miss their appointments. Trials also test one single drug brand and you rarely see head to head trials between different brands.world_population_africa_europe_magnify_400_clr_7653

In real life (i.e. patients in our practice), we have a mix match of patients with different motivation and different busy schedules. They have a real life to live and tablets are omitted for a number of reasons. They also have more than one GP and anyone in the chain of command can swab and change their medication regimen.

Nonetheless, randomised controlled trials (RCT) are insightful as they inform us what certain drugs can theoretically do and what the maximum effect of those drugs can achieve. This provides us with guidance that patients in real life will more than likely respond. This is the basis of REAL WORLD DATA series that we will be seeing more of in the near future. They are cheaper to run and involve much larger number of patients. However, in order for us to believe the results, we need to carefully look at the source of data, the completeness of the data and what are we actually interrogating.

EMPA-REG OUTCOME was a prospective RCT in patients with T2D and established atherosclerotic CVD, demonstrated a substantial reduction in CV death and hospitalisation for heart failure (HHF) with the SGLT2 inhibitor, empagliflozin within a short follow-up period. Particularly because the mechanism of this benefit is unclear and is still being debated, it has open up a number of questions:

Is this a class effect? In other words, will the other SGLT2 inhibitors do the same? You can imagine which drug company will say what.

Will this apply to patients in the real world? These are patients in our practice receiving “standard” care with their normal non-perfect compliance.

What about patients with normal CV risks? Will they benefit from being on SGLT2 inhibitors?

Using data from multiple countries, The Comparative Effectiveness of Cardiovascular Outcomes (CVD-REAL) study compared the risk for HHF, death, and the combined endpoint of HHF or death in patients with T2D who were new users of SGLT-2i versus oGLDs in real-world practice. Any SGLT2 inhibitors were included. 166,033 SGLT-2 inhibitor users were matched with similar number of patients on other glucose lowering drugs (oGLD). What did they find?

  • Compared with oGLD, treatment with SGLT-2i was associated with a 39% relative risk reduction in HHF, a 51% reduction in all-cause death, and a 46% reduction in the HHF or death composite, consistent with what was found in the EMPA-REG OUTCOME trial
  • Approximately 87% of patients did not have known cardiovascular disease, suggesting possible cardiovascular benefits for a broad population of T2D patients
  • The lower rates of and death associated with SGLT-treatment appear likely class related as all brands of SGLT2 inhibitors were used

Before you say wow, let’s look at the: Data sources

De-identified data from 6 countries were analysed:

  1. US – primarily encounter, claims and benefit based data from >300 large self-insured US employers and >25 US health plans
  2. Germany – from the Diabetes Prospective Follow-Up (DPV) initiative with 452 centres participating
  3. UK – primary care data from >670 general practices linked with hospitalisation and mortality registries
  4. Sweden, Norway and Denmark - mandatory national full-population registries of each respective country were used, with linked Prescribed Drug Registers covering all drugs . dispensed, National Patient Registers covering all hospitalisations and specialised outpatient care, and Cause of Death Registers

As you can see, datasets were different according to which country they were sourced from. The Scandinavian countries were pretty complete while the US data were least complete. The US contributed almost 234,000 patients, the Scandinavians countries 60,000, UK 10,000 while Germany had 3000 patients. The event rates were all different in different countries. However, the CV benefits were consistently seen across countries regardless of the variability in the health systems.

Dapagliflozin were the dominant SGLT2i used while Canagliflozin was more commonly used in Europe. This suggest that the CV benefits are likely to be class effect rather than brand specific. This is encouraging for our patients on  SGLT2i other than empagliflozin.

As this Real World Data include patients with all levels of CV risk, SGLT2i may provide benefit in patients with lower risk than those in the EMPA-REG OUTCOME trials, i.e. average risk.

We have evidence that empagliflozin have CV benefits (EMPA-REG). CANVAS (canagliflozin), DECLARE (dapagliflozin) and VERTIS (ertugliflozin) are yet to report as we go to press. CVD REAL plugs the gap in the meantime.


Kosiborod et al Lower Risk of Heart Failure and Death in Patients Initiated on SGLT-2 Inhibitors Versus Other Glucose-Lowering Drugs: The CVD-REAL Study. Circulation. 2017;CIRCULATIONAHA.117.029190

The views expressed are views of the authors and do not necessarily represent the views of the board of SSW GP Link

[Home] [TOBOGM] [DiabetesCaseConference] [Old&Weak?] [Diarrhoea & Diabetes] [SpecialisedWomenScreening] [RealWorldData] [Pesticides] [LowDoseCT] [FreeRetinaScreen] [LowCostPodiatry] [DELIGHT]