JC The REVERT trial: Dip or doom for SVT in the Emergency Department?

Dip or Doom?SVT (supraventricular tachycardia) is something we see a lot of in Emergency Medicine.  Any emergency physician will tell you exactly how satisfying it is to treat a patient with SVT.  There must be close to a 100% successful cardioversion rate, one way or another, and after cardioversion patients can usually go straight home soon afterwards.  It’s one of those things that gives you a warm glow and reminds you why you did Emergency Medicine in the first place.

But not everything is great about the way we manage SVT in practice.  More often than not we revert to using adenosine.  This is a great drug – rapid onset, short-acting, safe – and we all know how effective it can be.  You just have to remember to warn the patient that they might feel that they’re about to die when you give it.

But hang on.  How much would you like to feel like you’re about to die, even it’s just for a few seconds?   I can’t say that I have a desperate urge to feel that way.  Wouldn’t we prefer a better solution?  Well, many people do – and that’s why things got pretty heated at the RAGE podcast last year when they debated the use of verapamil versus adenosine…



Since these podcasts Simon has been using verapamil as first line for his low risk patients with SVT for a while and swears by it.

But wouldn’t it be nice if we didn’t have to use any drugs?  Wouldn’t that be better?  We all know that we could.  We could try carotid sinus massage, accepting the small but finite risk of causing a stroke.  Hmmm.  We could dunk the patient’s face in a bucket of ice cold water.  Hmmmm.  Or we could try a Valsalva manoeuvre.  It’s very safe and it works in almost 1 in 5 patients.

We all like to try that – but wouldn’t it be great if it worked for more than 1 in 5?  Well, 5 years ago two emergency physicians described a modified Valsalva manoeuvre that might improve success rates in the EMJ.  Moving on from that, an inspirational group from the south west of England led by Andy Appelboam put down their PASTIES and set to work on a definitive trial.  And their most awesome work has just been published in the Lancet.   Let’s take a look at what they did and what they found…

First, what is a modified Valsalva and how do you do it?

Andy and his team proposed to modify the Valsalva manoeuvre to improve venous return and vagal stimulation in the ‘relaxation phase’ (after the straining bit).  Essentially, the patient would puff into a sphygmomanometer at a pressure of 40mmHg for 15 seconds, then lie back in the Trendelenberg position with legs elevated.  You can see a video of Andy in action doing the modified Valsalva manoeuvre below…


What was the aim of the trial?

Essentially, they aimed to test the null hypothesis that, when compared to a standard Valsalva manoeuvre, the modified Valsalva manoeuvre will not improve cardioversion rates measured 1 minute after the procedure.  That means this was a superiority trial – they were trying to disprove that hypothesis and show that the modified Valsalva manoeuvre is better than the standard one for achieving cardioversion.

Who did they include?

This was a multicentre trial run at two teaching hospitals and eight district general (community) hospitals in the UK.  The multicentre nature of the trial will help to improve the external validity of the findings, making it less likely that they’re specific to patients from a particular geographical area presenting in a certain way.  Being able to run trials like this is one of the key benefits of the truly amazing National Institute for Health Research (NIHR) portfolio in the UK.  The NIHR provides the resources needed to deliver studies like this.  Without the NIHR, we just wouldn’t have studies of this nature.

Adults (>18yr) with regular, narrow complex SVT were included provided that they weren’t shocked and in need of immediate cardioversion, and providing that the diagnosis wasn’t atrial fibrillation or flutter.  This seems pretty reasonable.

How did they randomise patients?

They randomised patients to receive either the modified or standard Valsalva manoeuvre in equal (1:1) proportions.  The patients were randomised using sealed envelopes.  I’ve never been a great fan of this technique because of the potential that it might be open to abuse.  Given that it’s just not possible to blind clinicians to the allocation in a trial of this nature, what if you open the envelope and find that your patient is allocated to a group that you didn’t want them to be allocated to?  Could you exclude them from the trial and use that envelope for the next patient?  Could you just open the next envelope?  Does it therefore mean that we’ve lost allocation concealment?

In this trial the answer is actually no.  This group asked the team to sign and date the seal of the envelope before they broke it.  By doing that, the team couldn’t get up to mischief and abuse the system.  They must sign and date before they break the seal and they must include the patient because the seal has been dated.  It’s a very clever way of making sure that randomisation is simple and pragmatic (essential for a busy Emergency Department) while maintaining allocation concealment.

Randomisation was stratified by centre so that there would be roughly equal numbers of patients in each treatment group at each centre.  This removes the potential for confounding by having unequal numbers in the groups at different centres.  If the centres had different success rates and there were uneven numbers, this could have been a problem.

What did they do?

Consenting patients received either the standard or modified Valsalva after randomisation.  If the first attempt didn’t work, a second was permitted.  An ECG was then recorded 1 minute afterwards.  This was later interpreted by an independent cardiologist to determine whether cardioversion had occurred.  If the ECG was missing, an endpoint committee used all other available information to determine whether cardioversion had occurred.

What OUTCOMES were they interested in?

The primary outcome, which is the most important measure of success, was return to sinus rhythm after 1 minute.  There were also a number of secondary outcomes, such as the rate of use of adenosine.


Ultimately 433 patients were included in the trial.  They seem to have been pretty well matched in terms of baseline characteristics.  None of the patients crossed over between the groups and a similar proportion in each group (84% and 86%) achieved the required pressure for the appropriate time when actually doing the Valsalva manoeuvre.

What did they FIND?

The results are pretty impressive.  The primary outcome of return to sinus rhythm after 1 minute was achieved for 43% of the patients who used the modified Valsalva versus 17% for the standard technique.  This is an absolute risk reduction of 26.2% (p<0.001) and gives us a number .needed to treat of 3.8.  

This finding is pretty impressive and my only slight gripe is that they possibly didn’t need to round that number down to 3 – it’s already extremely impressive and if they wanted to round that number to an integer, it should really be the more conservative estimate of ‘4’.

The findings are equally impressive for some of the secondary outcomes.  There was a 19% difference in the proportion of patients given adenosine (69% vs. 50%, p=0.0002, NNT 5.3).  In terms of safety, there was no difference in the incidence of serious adverse events.  There were slightly more adverse events reported by patients who had a modified Valsalva, but the difference wasn’t statistically significant and they all seem pretty mild in nature (e.g. transient headache) – and nobody experienced a sensation of impending doom – which adenosine notoriously causes.

So what does this MEAN?

Clearly, the findings are pretty impressive.  The modified Valsalva is safe and can cardiovert 43% of patients with SVT.  By using the modified Valsalve instead of a standard technique, we could avoid using adenosine in one out of every 5.3 patients treated, which isn’t bad.  What’s more, this is a pretty robust trial with no major weaknesses.  My only remaining question is whether you need to adopt the Trendellenberg position after the strain or whether being in the supine position all along would be just as effective.  I’ve always been taught that the Valsalva manoeuvre is more effective in the supine position.  However, based on these data, I’ll certainly be changing my practice and using the modified Valsalva technique as shown in the video.

It also has some more implications.  If we teach patients how to do this themselves, maybe they can self-cardiovert without coming to the ED or using a pill in the pocket.  From a patient’s perspective, this could be great.

Want to hear MORE?

If you want to hear more from the first author himself, come to #RCEM15 in Manchester or #EuSEM15 in Torino.  Andy Appelboam, the first author, will be presenting his findings at both conferences – be there if you can!

* Special thanks to Natalie May and Simon Carley for their input into this post. The graphic and title are Natalie’s awesome work. *







  1. Derek Louey

    This makes sense physiologically because you are trying to induce a baroreceptor reflex. I normally attempt Valsalva starting in Trendelenburg but never seen a 1 in 4 success rate. Abruptly moving into Trend might be a better way to give a venous return surge to trigger the reflex.

    The mistake I see in trainees is to believe the straining actually is the part that does the trick (versus the release). The poor patient is asked to strain forever whilst the registrar vainly watches the ECG monitor for something to happen.

    1. Hugo Dowd

      Ive been getting kids to do handstands for SVT for years.Again Baroreceptor response I think coupled with stretching the thin walled right atrium allows for relative ischaemia of the accessory pacemakers versus the more robust SA node. I could be making it up but it works for me.


      “This makes sense physiologically because you are trying to induce a baroreceptor reflex. I normally attempt Valsalva starting in Trendelenburg but never seen a 1 in 4 success rate. Abruptly moving into Trend might be a better way to give a venous return surge to trigger the reflex”


      And could it additionally be improved crossing the lower limbs and even upper limbs with muscle activation in a manner that venous return would be additionally enhanced?
      Or there are another inconveniences (active /non passive Trendelemburg position) that make this additive suggestion non contempled?


  2. Andywebster (@Andywebster)

    Was control trying same position sitting up? Would have been interested to see this technique compared to starting in trendelenburg.

  3. Pingback: Keeping upto date in EM | Emergency Medicine Podcasts

  4. Pingback: LITFL Review 195 | LITFL: Life in the Fast Lane Medical Blog

  5. Sal

    Is it a powered study?

  6. Pingback: Episode 34 – Tachyarrhythmias | FOAMcast

  7. Pingback: The REVERT Trial: A Modified Valsalva Maneuver to Convert SVT - R.E.B.E.L. EM - Emergency Medicine Blog

  8. Pingback: TPSV: usa la manovra di Valsalva modificata - EM Pills

  9. PA Dave

    0 for 1 here. Patient was a great sport in giving it a shot though!

  10. Pingback: A better Valsalva for SVT | FOAM links

  11. Pingback: LITFL Review - Best of 2015 | LITFL: Life in the Fast Lane Medical Blog

  12. Pingback: SGEM#147: This is a SVT and I’m Gonna REVERT It – Using a Modified Valsalva Manoeuvre | The Skeptics Guide to Emergency Medicine

  13. Pingback: Isso é Taqui Supra e vou REVERTê-la – Usando a Manobra de Valsalva Modificada | The SGEM Global

  14. Pingback: #FOAMed on prescription. St.Emlyn's - St.Emlyn's

  15. Pingback: SVT – EMbytes

  16. Pingback: SVT – EMbytes

  17. James Seddon

    A bit late to the party but some interesting physics / physiology resources here. Everything you ever wanted to know about the Valsalva manoeuvre came out this year by Pstras et al in Acta Physiologica http://onlinelibrary.wiley.com/wol1/doi/10.1111/apha.12639/full ; same chap also pointed out that the widely used syringe is unlikely to generate the consistent 40mmHg resistance for 15s that is accepted as being integral to the VM (and used in REVERT). This is due to an increased initial resistance to movement (static force) which is proportional to how long the barrel of the syringe has been at rest for. This was demonstrated by Thornton et al in Leicester which showed large pressures (150/300mmHg for 10 or 20ml size respectively) required to get a new syringe going which reduced once the barrel had been moved (http://emj.bmj.com/content/33/10/748.full) No one, it seems, has measured the glide resistance i.e. the resistance generated once the barrel is moving which is likely to be the force at play over a 15s period.
    REVERT used a BP manometer on wheels and connected it to the patient with bubble tubing. If you’re out in the field and this isn’t practical, I suggest a 20ml syringe that you have given a bit of a wiggle to first and then put the legs up.

  18. Cheryl

    I have svt had for about 17 years was found only by attending hospital when my pulse was 239 I had the drug twice luckily 2nd time worked had no episodes since that I couldn’t fix my self this morning it happened I knew it wouldn’t go back after 30mins of trying blowing in a syringe and cold water on my face couldn’t fix it so husband called an ambulance they came out took me into ambulance and got me to do the syringe for them heart rate was sitting at about 220 probably going on for about an hour they tried this after him doing a drug trial and had read up on this the other guy had never seen it done did it in the ambulance and it worked monitored me for a while as they doing their paperwork so no need to go to hospital so I say as a patient great less scary than the drugs and worked felt funny at the time pressure in chest breathless and sick but as soon as I was sat up again I knew I was fine and I was

  19. Josep Serra

    As a primary care ph I had heard from this mod VM
    Nevertheless one more question I make myself:
    And if we could even improve this Maneuvre .
    Based on physiology:
    Could we add to the Trendellemburg position the leg crossing aimed to increase the active return venous flow in order to enhance barorreflex stimuli?

    I leave this doubt to the forum
    Thank you very much


Thanks so much for following. Viva la #FOAMed

Translate »