Here are the results of our July, 2015 mixing study correction Quick Question survey:
When performing PTT mixing studies using normal control plasma (NP), what criterion do you use to define correction?
A. Mix results within 10% of NP result—60 votes, 32%
B. Mix results within 5 seconds of NP result—20 votes, 11%
C. Mix results within reference interval—76 votes, 40%
D. Mix results within 5 seconds of reference interval mean—15 votes, 8%
E. Other: please post or email to [email protected]—18 votes, 9%
It is clear we have no consensus except to assert that each institution standardize on their own policy. The comments subscribers added in response to answer E included the Rosner and Chang index, a combination of a dedicated reference interval with the Rosner index, and a variety of absolute time interval-based limits, for example, “mix results within 10 seconds of reference interval mean.”
Observe our current Quick Question, when we ask, “What PTT result is used to initiate a mixing study?”
From Donna Lawler: George,
From Donna Lawler: George, We use a cut off of 10 seconds over the reference interval of the PTT to perform mixing studies. We find any result less than that usually corrects
and may mislead the clinician.
Follow-up question from
Follow-up question from George. Thank you, Dr. Favaloro. From curiosity, if instead of using return to reference interval you used a NP-based correction factor such as straight percentage or Rosner or Chang index, would you still see potentially false correction?
Dr. Favaloro’s answer: Hi George,
We used to use a percentage correction, in fact a sliding scale of percentages, but it was too difficult to apply across our Network of 27 labs, with our 100 plus techs. The percentage correction is probably ‘better’ than correction into the normal range, but is also difficult to get right at extremes of results, and interpretation for the staff was problematic. For example, we used to use 10% correction factor at the lower end of abnormal, but even 10% correction at say 40 sec = 4 sec, so with a normal median of 31 sec, a time of 35 sec would be a correction, but a time of 36 sec would not, and there is little real difference between 35 and 36 sec. So, we chose the within normal reference range as being the most unambiguous option for the staff. It won’t always be right, but then nothing is. Most options are only right about 90% of the time. For more information, please see Kershaw G, Orellana D. Mixing tests: diagnostic aides in the Investigation of prolonged prothrombin times and activated partial thromboplastin times. Semi Thromb Hemost 2013;39:283–90.
(From Dr. Emmanuel Favaloro)
(From Dr. Emmanuel Favaloro) Hi George,
Our response:
For outpatient or emergency admissions, we perform APTT mixing studies when the APTT exceeds the normal reference range; however, as we now use a simplified ‘correction’ interpretation (mix result within the normal reference range), mixing studies are only generally interpretable if the original APTT exceeds the normal reference range by at least 5 seconds (since most abnormal APTT results just above the normal reference range will yield mix APTT results within the normal reference range and thus be reported as ‘correcting’ whether or not they actually have an ‘inhibitor’).
For hospital inpatients we do not perform mixing studies unless these are specifically requested clinically (realistically, ~50% of our hospital inpatients are on some sort of anticoagulant therapy or on heparinised lines, and mixing studies are not indicated; we used to ring up to find out which patients required follow up but basically gave this up as too difficult – (i) took too much tech time for the limited benefit, (ii) some wards refused to give us any information or gave us incorrect information or took umbrage that we would waste their time asking).
cheers,
Emmanuel Favaloro