The patient is too dizzy to walk to the bathroom to give you a urine specimen to check a urine pregnancy test. Plus, she admits that she just urinated in the waiting room bathroom a few minutes ago – so no urine now.
Trick of the Trade
Apply several drops of whole blood (instead of urine) into the pregnancy test cassette. In the photo below, the patient was pregnant with a serum beta-HCG level of 250 mIU/mL whose urine and whole blood qualitative tests were both positive.
Both the urine (top) and whole blood (bottom) of the patient was tested on a qualitative pregnancy test kit. After placing several drops of urine or blood in the sample (S) well, a visible bar appeared at the test (T) line similar to the control (C) line result for both, which indicates a positive pregnancy test.
Did you know that most urine pregnancy test kits are approved for both urine and serum samples? On a quick Google search, I found that Accutest, Cardinal Health, ICON, OSOM, and Rapid Response all are approved for both. The question is whether this will work for whole blood.
One study (1) in the Journal of Emergency Medicine by Dr. Fromm from Maimonides Medical Center looked at exactly this issue. Whole blood pregnancy test performed extremely well, especially if positive:
- Sensitivity 95.8%
- Specificity 100%
- Negative predictive value 97.9%
- Positive predictive value 100%
In their study, very low beta-HCG values (<159 mIU/mL) occasionally yielded a false negative for whole blood pregnancy tests. The whole blood testing approach missed a total 9 of 425 pregnancies. Interestingly, the urine pregnancy test was also negative in 5 of those 9 and not performed in the other 4.
- Be sure to wait at least 5 minutes when using whole blood in the kit. It sometimes takes a while.
- Do not apply additional drops of water or saline to the whole blood sample. This causes unnecessary dilution. Just wait for the blood to osmose across the entire test strip.
- This is trick is ONLY for medical professionals and not the lay public. We are discussing an actual blood draw and not a simple cut on a finger to obtain blood.
S = Sample well
T = Test specific (will show bar if +HCG)
C = Control (will always have a bar)
- Fromm C, Likourezos A, Haines L, Khan AN, Williams J, Berezow J. Substituting Whole Blood for Urine in a Bedside Pregnancy Test. J Emerg Med. 2011 Aug 27. Pubmed .
- Habbousche JP, Walker G. Novel use of a urine pregnancy test using whole blood. Am J Emerg Med. 2011 Sep;29(7):840.e3-4. Pubmed .
Expert Peer Review
Thank you so much for your interest in our article, Dr. Lin. To date we have not formalized a process through our laboratory department. Similar to your shop, we use whole blood to make time-sensitive clinical decisions and then try to confirm. But in light of all the recent interest – I have received several inquiries about this matter (I suspect due to the popularity of this blog!) – I am considering taking this the next step, and I will keep you abreast of any developments.
I should like to add some detail to your discussion of the cases in which there were discrepancies between the urine or whole blood POC tests and the quantitative serum hCG testing. You correctly point out that the whole blood testing missed 9 out of 425 pregnancies and that the urine test was also negative in 5 of those (those 5 hCGs were 5, 16, 18, 47, and 50). Keep in mind that the test is intended to be 100% sensitive only for hCG >25, so that probably explains the false negatives for 3 of them, but is somewhat troubling that both the urine and whole blood missed the hCGs of 47 and 50.
On that point, we do know that false negative urine pregnancy testing can occur in the presence of high levels of hCG variants [please refer to the very interesting recent article by Dr. Richard T. Griffey: “Hook-Like Effect” Causes False-negative Point-of-Care Urine Pregnancy testing in Emergency Patients. J Emerg Med 2013;44(1)155-160], but I’m not sure how likely that is to be the case here with such low hCG values.
As you point out, the urine test was simply not performed due to oversight in 4 of the other discrepant cases in which the whole blood tests were falsely negative, so we will never know if the urine would have missed those too, but I suspect the urine would not have picked those up either because of the low hCG values, which were 6, 9, 12 and 22 (all below the test’s threshold of 25).
There were no discrepant cases in which the whole blood test was negative but the urine test was positive. However, there were two discrepant cases in which the urine tested negative but the whole blood was positive (those hCGs were 83 and 159). So, according to our data, there is a suggestion that the whole blood may actually be more sensitive than urine, but that might require a larger sampling to establish.
ALiEM’s Bottom Line is spot on — believe a positive test. When we first discussed this technique in our paper, it was in the context of a fascinating case — vomiting, no abdominal pain, breast-feeding a newborn, and having just eaten sushi — except profoundly hypotensive and very ill appearing. This technique allowed us to confirm what seemed like an unlikely diagnosis at the time very quickly, and get to her the OR with OB-Gyn. We’ve received criticism from our paper from several pathologist PhD’s (none of which are physicians nor clinicians) that our test was not FDA-approved, but we see incredible value in this technique, regardless of its official use with serum vs. whole blood; we echo the second part of the Bottom Line: use an approved confirmatory test.