Transfusion Reaction Benchmarks

[bkrobinson]

Our Cancer Center is asking us to provide benchmarks to compare our transfusion reaction data to. I have looked just to be able to say that there is none and can not find any. Do any of you know of some that we can get access to?

[RR1]

Could your (U.S) Biovigilance folk help you with this?

I'v never come across this before either and don't quite understand how you can benchmark transfusion reaction data. I'm assuming majority of your T.R's are not caused by lab errors/ missed RBC antibodies.

Also, for what exact purpose does the Cancer Centre want the data for?

Interesting post- thanks.

[Bill]

In New York State, every transfusion service must report all transfusion reactions to the state Health Dept each year. There may be some data at www.wadsworth.org, which is the lab/blood bank website. I will try to search in next 2-3 days and post more if I find the info.

[bkrobinson]

Thanks to you both. The Cancer Center is using transfusion reactions as a quality metrix regarding our service to them. I explained that the reactions we see are not laboratory related and that we really should use some other measure but they are stuck on this one. Thanks about the website. I will try and look on it too.

[bhunterlabguy]

Here is a link you can use...

http://www.cdc.gov/nhsn/bio.html

[bkrobinson]

Thank you so very much.

[SusanM]

Do any of these (or is there somewhere else) give actual benchmarks? Like we expect to a rate of 1%, etc... Expected values vs. over/under reporting?

Thanks!

[bkrobinson]

I emailed them about this and they responded back that their data is not available on this yet and it will be sometime in 2012 until it is. I also am checking with CAP and the California Blood Bank Association. I have never really ever heard about a rate to expect since they are all patient dependent, except when the Transfusion Service makes a mistake and here we expect that to be zero.

[cthherbal ☆]

I emailed them about this and they responded back that their data is not available on this yet and it will be sometime in 2012 until it is. I also am checking with CAP and the California Blood Bank Association. I have never really ever heard about a rate to expect since they are all patient dependent, except when the Transfusion Service makes a mistake and here we expect that to be zero.

We also have to report our data to the state annually, who reports the findings on their website. If I recall the rate range was somewhere around 3-4% of total transfusions for 2009. 2010 data is not updated yet.

[Eoin]

Reports on the SHOT website in the UK will be handy (www.shotuk.org ). Have a look at their annual reports. Here in Ireland, you can access annual reports of the National Haemovigilance Office through the Irish Blood Transfusion Service website direct link to the publications is http://www.giveblood.ie/Clinical_Services/Haemovigilance/Publications/.

Obviously if you are looking at a specialised population (i.e. Cancer Patients) in particular, that may skew expected rates. Keep in mind that there is more reporting over this side of the big pond because of legislation making such reporting mandatory. You would also have to take in to account rules for reporting, as a lot of minor reactions are not reportable (mild allergic springs to mind). If you are dealing with advanced stage cancer patients, a lot may have pyrexia going on underneath related to their disease, not the transfusion. The variables are endless (types of product etc). I haven't seen any comprehensive study on a large population, but would be very interested if you do come across any.

Good luck with it. I think they are asking a question about which they have little or no knowledge - evidenced by just asking the question anyway.

Cheers

Eoin

[wardj]

Maybe a better quality indicator would be tracking the reporting of suspected reactions by the transfusionists. We this as a % of transfused components. We used our historical practice and some comparitive research and determined our benchmark of % reporting by product, and by total components.

[Mabel Adams]

If they just want a number to say they looked at something, then give them something that is harmless, essentially true and not demanding of many resources to produce. If they want actual meaningful information, you need to find out what they hope to learn and then find a way to assess it. Using the rate of suspected transfusion reactions as a metric for measuring the lab's quality (as opposed to the rate of reactions that could be somehow associated with lab error) could motivate the Lab to discourage reporting of suspected reactions--not good science.

The FDA's annual report of transfusion fatalities and BPD reports might provide some good benchmarks for things that really could be lab error. (Sorry, I am home and don't have them readily at hand.) The problem with counting real reactions as opposed to suspected reactions is that simple statistics don't work very well for rare events.

By tracking the rate of suspected reactions, you are not tracking Lab quality but probably how well nursing is noticing and documenting reaction symptoms. This may be valuable to track but it is not a measure of Blood Bank quality.

[L106]

Excellent post, Mabel!