Use of prothrombin complex concentrate in warfarin anticoagulation reversal in the emergency department: a quality improvement study of administration delays

Anticoagulation with warfarin is very common, and there is clear evidence behind the clinical rationale for anticoagulation to prevent thromboembolic events. Despite the positive balance of risk versus benefit, there are still many adverse events associated with anticoagulation. Typically, fresh frozen plasma (FFP) and PCC are used for reversal in the case of life-threatening bleeding, and there is a significant body of evidence that suggests that speed of reversal is linked with outcomes. With that in mind, optimising the time to reversal of anticoagulation is essential. The authors aimed to analyse the use of PCC in their department with respect to creating an action plan that would minimise delays in the administration of the reversal agent.

The setting was a busy trauma unit, with approximately 65,000 visits per year. The procedure in place for administration of PCC (in this case Octaplex [human prothrombin complex]) was:

1. prescription written by a physician
2. prescription taken to the blood bank where the PCC was stored
3. PCC reconstituted by a technician
4. orderly called to take the infusion to a nurse
5. nurse to begin the infusion.

To review the procedure, the Ottowa model of research use (OMRU) tool was used to identify all stakeholders and to involve them in the review.

For the first phase of data collection, a retrospective audit was carried out on all charts for patients who received PCC to correct warfarin anticoagulation in major bleeding between November 2009 and October 2011. Data were extracted from charts involving PCC prescription for the reversal of warfarin-related bleeding, and the administration delay calculated – the time from prescription to the end of the infusion. Pre-reconstitution delays (time of prescription to start of reconstitution) and post-reconstitution delays (start of reconstitution to end of infusion) were also determined.

Phase two involved identification of barriers that caused a delay; the results of phase one were presented to all stakeholders, and an open discussion was held on the barriers identified. A jointly-produced action plan was implemented, with training sessions and flow charts drawn up to familiarise all staff with the new procedures. This was followed by a third phase, comprising a prospective data collection to evaluate the impact the changes had made.

Pre-intervention, there were 77 charts meeting the criteria for analysis, the mean administration delay was 76.3 minutes [SD 34.1 mins], with a median administration delay of 70 minutes [IQR 55–95 mins]. The median pre-reconstitution delay (n=36) was 25.5 minutes [IQR 12.8–34.5 mins], and median post-reconstitution time was 43.5 minutes [IQR 31.0–60.0 mins]. The main barriers identified were:

• inefficient communication processes between the emergency department and the blood bank at time of prescription
• retrieval delay
• delay to immediate administration once received.

The new procedures involved the prescribing physician calling the blood bank directly, and nurses assigning an orderly to retrieve the PCC at the time of prescription. For the prospective analysis, 39 charts over 6 months contained adequate information for a review. The mean administration delay was found to be 33.2 minutes [SD 14.2 mins], with a median time of 30.00 minutes [IQR 24.25–38.75 mins] – this represented a significant decrease in delay compared to the old procedures (p<0.0001). The new procedures were followed in 82% of cases; the mean delay when only those adhering to new protocols were followed was 28 minutes, with a median pre-reconstitution delay of 0 minutes.

The authors note in their discussion that the barriers to optimal care were due to communication and difficult care processes regarding reconstitution and collection. When these barriers were addressed, there was a significant reduction in the delay. It was decided that reconstitution would continue happening in the blood bank to minimise errors and concentrate expertise there. It was also clear that where staff had failed to follow the new procedure, the administration delay was increased. As a result it was decided that regular in-house training would be carried out. The authors acknowledge that the study had limitations – it was not compared to a control and the assessors were not blinded, leaving the possibility for bias to affect figures. Some measures were charted twice however, both by the emergency department and blood bank, which should limit the possibility of overly optimistic timing being recorded. They also did not assess delays occurring before the prescription of PCC; for example, delays getting a CT scan organised and reported.

This study takes a sensible approach in order to audit and refine departmental procedures. In its own right, it does little that is unique or practice-changing – and its applicability is limited. It is however an important example of the size of effect that can be caused by taking a critical look at departmental processes, and a reminder of the importance of a regular audit cycle – particularly in a time-sensitive clinical setting such as bleeding in patients receiving anticoagulants.