Education Blog for Healthcare Professionals

Pradaxa: Interruption for Colonoscopies, Dental Procedures, Surgery, etc.

When to stop the drug – General comments

The patient on Pradaxa® (Dabigatran) may need interruption of therapy for dental work, a medical procedure such as colonoscopy, or minor or major surgical procedure. As to when exactly to take the last dose of Pradaxa® before the procedure depends on (a) what type of procedure is planned and how much bleeding to expect with it, and (b) whether the patient is at high or low risk for thrombosis if he/she is off anticoagulants for some period of time.

Renal function is important

About 80 % of Pradaxa® is excreted unchanged via the kidney, the remainder via the bile. The half-life of Pradaxa® is majorly determined by renal function (see table below).


How quickly is Pradaxa® excrected?

In most patients with normal renal function, Pradaxa®’s anticoagulant effect is mostly gone within 1-2 days after the last dose has been taken, as the drug is relatively quickly excreted in the kidney. Typically, after 3 half-lives most of the drug (87.5 %) is gone. Most surgeries with a normal (standard risk of bleeding) can be safely done when the majority of anticoagulant effect is gone. For people with normal renal function this is ca. 24 hours after the last dose has been taken (see table). However, in patients with impaired renal function, Pradaxa® needs to be discontinued earlier because of its prolonged half-life, such as 2 or more days before surgery (see table). Also, if surgeries with a high risk of bleeding are planned, one will want all the drug out of the system and, depending on the renal function, the last dose of Pradaxa® would have to be 2 or more days before the surgery (see table below).

Table.  Guide to the discontinuation of Pradaxa® before procedures or surgeries 
(adapted from reference 2)

Renal Function

(CrCL mL/min)

Half-life (hours),

mean (range)

Timing of Discontinuation Prior to Procedure (Minimum)

Standard Risk of Bleeding

High Risk of Bleeding*

> 80

13 (11 – 22)

24 hours

2 – 4 days

50 – 80

15 (12 – 34)

24 hours

2 – 4 days

30 – 50

18 (13 – 23)

> 48 hours

> 4 days

< 30

27 (22 – 35)

48 – 120 hours

> 5 days

Examples: heart surgery, neurosurgery (brain, spinal cord, nerves), abdominal or surgery involving major organs; or procedures requiring guaranteed no bleeding  (i.e., spinal anesthesia), or when additional risk factors for bleeding are present (low blood platelets, bleeding disorder, previous major bleeding, etc).

Dental Procedures

Many dental procedures can be done on full dose anticoagulation [ref 1]. Detailed recommendations as to which dental procedures can be done on full dose anticoagulation (teeth cleaning, root canal, one or two teeth extractions) and for which the anticoagulant level needs to be reduced have been published [ref 1 and discussion here].  A similar approach can likely be taken in patients on Pradaxa®. However, it is also easy to tell the patient not to take his/her evening dose of Pradaxa® on the day before the procedure and not to take the morning dose on the day of the dental procedure; and then to restart the evening of the day of the procedure. However, individualized recommendations need to be given.

Colonoscopies and EGDs (=upper endoscopy)

Colonoscopies and EGDs can typically safely be done on full-dose anticoagulants. However, as biopsies and polypectomies may have to be performed, many gastroenterologists prefer to see the patient off anticoagulation. Once again, the renal function needs to be taken into consideration as to when to have the patient take the last dose of Pradaxa® –  the table above gives guidance.  The patient with normal renal function may not want to take Pradaxa® one day prior to surgery, nor in the morning of the day of the procedure (i.e. will have taken the last dose of Pradaxa® more than 36 hours before the colonoscopy or EGD).

Surgery, minor or major

When to take the last dose before minor or major surgery, again, depends on the degree of expected peri-surgical bleeding. The table above gives guidance.

When to restart Pradaxa® after surgery

Pradaxa® is rapidly effective after oral intake, reaching its maximum anticoagulant effect within 2-3 hours of intake. Therefore, in the patient with significant risk for bleeding after a procedure or surgery, the re-initiation of Pradaxa® may have to be delayed by a few days. Again, individualized decisions need to be made, depending on a patient’s risk for bleeding and clotting.


  1. Herman WW et al. Current perspectives on dental patients receiving coumarin anticoagulant therapy. Journal of the American Dental Association. 1997;128:327-335.
  2. Van Ryn J et al. Dabigatran etexilate – a novel, reversible, oral direct thrombin inhibitor: Interpretation of coagulation assays and reversal of anticoagulant activity. Thromb Haemost 2010;103:116-1127.
For Patients: the same information, written for patients, can be found here.

Disclosure:  I have no relevant financial conflict of interest with this blog entry.

Last updated: April 15th, 2011

Tags: , , , , , ,

This entry was posted by Stephan Moll on at and is filed under Anticoagulants, Pradaxa (dabigatran), Therapy. You can follow any responses to this entry through the RSS 2.0 feed. Both comments and pings are currently closed.

8 Responses to “Pradaxa: Interruption for Colonoscopies, Dental Procedures, Surgery, etc.”

  1. Dr Lachie Hayes says:

    While looking through the excellent guideline you have generously offered for wider use I came upon the following question about potentially “bridging” in patients on dabigatran (Pradaxa).

    Although currently recommended / approved for AF, I am aware that off-label use has occurred anecdotally (mechanical valves etc) and will likely occur increasingly in the future. For a patient estimated to be at HIGH thrombotic risk leading up to a major procedure with normal renal function, is there any recommendation on whether bridging strategies are required? The offset of dabigatran is much quicker than warfarin.

    The Van Ryn article recommends holding for 2-4 days for these cases but the same article indicates that the anticoagulant effect should be markedly reduced by 12-24 hours and that LMWH should be started within 12 hours of the last dose of dabigatran.

    Does this mean that for HIGH thrombotic risk patients, they should “bridge” with some form of parenteral anticoagulant, i.e. if they stop 4 days before, should they start LMWH that evening etc?


    Lachie Hayes

    • Stephan Moll says:

      No, bridging would not make any sense. The half-life of Pradaxa is very similar to that of the LMWHs. Thus, if one were to bridge, one would just replace one anticoagulant with another, still having to wait the same amount of time after the last LMWH dose for the drug to be completetely cleared to minimize the risk of bleeding with surgery.

      • Dr Lachie Hayes says:

        Thanks for the reply.
        I suppose I am concerned by the wide range given in the van Ryn article for major procedures. For LOW thrombotic risk patients this is not an issue probably but I’m worried about patients on dabigatran / Pradaxa with AF who have other high risk features or concomitant recent VTE.
        I don’t usually stop LMWH 2-4 days before a procedure in HIGH thrombotic risk patients with normal renal function as it has a similar short half-life (although half full therapeutic dose 24 hours before may be warranted in patients with HIGH bleeding risk to avoid any significant residual anti-Xa effect).
        I’m unclear why the article recommends this approach if the half-life of dabigatran is short as it says – to me this strategy would result in a 2-3 day period of no anticoagulation at all. Granted perioperative anticoagulation is more of an art than a science but this seems to contradict itself.

        • Stephan Moll says:

          For the issues we are discussing – when to discontinue Pradaxa prior to surgeries – I would point our that the half lives of LMWHs and Pradaxa are quite different: 3-5 hours for LMWHs after s.c. dosing [Hirsh J et al. Chest 2008;133:141S-159S], and 11-22 hours for Pradaxa (van Ryn J et al. Thromb Haemost 2010;103:1116-1127]. Given that most of any drug effect is gone within 5 half-lives, this would be 15-25 hours after s.c. injection for LMWHs (in the person with normal renal function), and 55-110 hours (i.e. 2-4 days) after discontinuation of Pradaxa (in the person with normal renal function). So, the vanRyn table makes sense to me. However, a problem in regards to the high risk thrombosis patients – and I agree with you on your concern – is that the variability of the half-life for Pradaxa is so wide (11-22 hours) that one does, indeed, worry about that patient if he/she is off anticoagulants, if that person were to have a short Dabigatran half-life, yet would be kept off anticoagulation for 4 days. In these high risk thrombosis patients a monitoring test would be nice to have to know when most of the Pradaxa is out of the system (or, phrased differently, if significant amounts of Dabigatran are still circulating).

  2. Dr Lachie Hayes says:

    I think that’s cleared up my confusion – you mentioned the half-lives of LMWH and dabigatran were similar and I suppose they are in terms of making bd dosing feasible but when I remembered the actual figures I realise they are significantly different enough in this scenario (particularly accounting for the variability of dabigatran) to be a potential issue.