is part of the Global Exhibitions Division of Informa PLC
This site is operated by a business or businesses owned by Informa PLC and all copyright resides with them. Informa PLC's registered office is 5 Howick Place, London SW1P 1WG. Registered in England and Wales. Number 3099067.
By Frank Ryan, Ph.D., D (ABCC); Technical Director,
National Reference Laboratory, Abu Dhabi, UAE
13 September 2016
Healthcare providers are so busy these days that they may quickly look at test results and make assumptions that are incorrect. Although it is a responsibility of the laboratory to provide clear reports in order to prevent wrong interpretation of the results, the laboratory doesn’t always have all the patient information needed to clearly interpret a report. Thus, it is incumbent on both the laboratory and the provider to establish simple lines of communication to exchange information and ensure that the results are correctly interpreted.
In my more than 40 years of experience in the laboratory there have been many examples where the initial result reported caused confusion, and if the provider didn’t contact the laboratory for clarification, the patient treatment would have been different. Some examples of these cases are given below. One of the major causes of errors in the laboratory is the fact that providers often order tests by test name only, rather than the test number, which the laboratory can easily misinterpret. Therefore, we strongly recommend that when possible, the provider should include the laboratory test number when ordering a test. Moreover, for the tests discussed in this article, the test number as in the test menu of NRL is noted.
Example 1: False Positive Laboratory Results due to Heterophilic Antibodies
A provider called because a serum pregnancy test was positive in our laboratory while the urine pregnancy test was negative. Additional tests found that the urine test was correct, and the serum test was actually a false positive.
There are very few reasons that such discrepancies can happen. One reason is that the patient is producing antibodies that interact with the antibodies in the immunoassay used by the laboratory. For instance, if a test uses mouse monoclonal antibodies to measure an analyte, and the patient makes antibodies to mouse antibodies, the patient antibodies can cause false positive results (or false negative results). Such antibodies, produced by a patient against the antibodies of another species, are known as heterophile antibodies.
There have been many publications on heterophile antibodies both trying to determine why patients develop antibodies to antibodies from other species and to determine the frequency of such antibodies. In general, people that develop antibodies to mouse or goat antibodies don’t have a history of contact with either goats or mice.
One study found that patients with heterophilic antibodies tended to also have selective IgA deficiency1. In the study, 38% of people with selective IgA deficiency had anti-goat antibodies, and 18% had anti-mouse antibodies. False positive test results occurred in 30% of subjects with selective IgA deficiency.
Selective IgA Deficiency is one of the most common primary immunodeficiency diseases. Studies have indicated that as many as one in every 500 Caucasian people has Selective IgA Deficiency. The rate of occurrence may be different in other ethnic groups.
The good news is that test methods developed after 2005 have intricate testing schemes to eliminate the effect of heterophile antibodies on the tests. Even so, the possibility of heterophilic antibodies should be considered if unusual immunoassay results are obtained. Since antibodies are not present in the urine, heterophilic antibodies do not affect urine assays when the serum assays are affected.
Example 2. False Negative Celiac Disease Test
A provider called our laboratory to ask the significance of an abnormal low Immunoglobulin A result in a celiac disease panel (e.g., NRL test #165126). Studies have demonstrated that IgA endomysial antibody tests have greater than 99% specificity for celiac disease and 70-80% sensitivity. The potential problem with using just the IgA endomysial test to detect or rule out celiac disease is that 0.5% of some ethnic groups have no IgA antibodies. Such individuals will always test negative for the IgA endomysial antibody even if celiac disease is present.
To address the problem with selective IgA deficiency causing false negative endomysial antibody IgA tests, the total IgA test is run whenever the endomysial antibody test is ordered. If the IgA result is negative or significantly low, an alternative to the endomysial IgA test is required. In such cases, the next best tests to detect celiac disease tissue transglutaminase IgG (tTG IgG) and deaminated gliadin-derived peptides IgG (DGP IgG) should be run (NRL test #s 164988 and 161687). Since these tests do not require the patient to have IgA antibodies, these tests can be used in conjunction with clinical findings in the diagnosis of IgA-deficient celiac disease.
To answer the provider’s original question, the low result for the total IgA test is not a significant finding relating to the presence of celiac disease. The low result simply tells the provider that the IgA endomysial antibody test is not an appropriate assay to use to assist in the diagnosis of celiac disease in this patient.
Example 3: Carcinoembryonic Antigen (CEA) Result That Didn’t Make Sense
A provider called our laboratory to say the CEA test (NRL test #002139) at the laboratory was not correct and wanted to know if there was a problem with the assay or a heterophilic antibody present (see example 1 above). The patient had a breast tumour removed two years previous and an elevated CEA prior to the surgery. After surgery, the CEA went to 2.4 ng/mL, a normal value. In May, the value started to increase and in June, the result was significantly elevated. The provider could not find any signs of tumour recurrence and called the laboratory to question the result.
The June sample was repeated with a good repeat and the provider was told the result was good. In July, the result had increased three-fold and the provider called to say there must be some problem with the Laboratory. The July sample was sent to a competitor laboratory and also tested the sample for heterophilic antibodies (NRL test #140657) to determine if such antibodies were producing false positive results. The competitor laboratory matched our laboratory results and no heterophilic antibodies were detected.
The provider continued to test monthly with higher results each month (see figure A) and also called me monthly. In November, the provider called to apologise for all his phone calls, as he was able to find evidence ω
of the tumour recurrence. In this case the laboratory was able to identify the recurrence five months before the provider could find physical evidence of the recurrence.
Example 4. Falsely Elevated Serum Pregnancy Test Caused by Post-Menopausal Production of HCG-Like Peptide
A 55-year old lady needed surgery after a car accident, and the hospital required serum pregnancy tests prior to surgery. Her Beta-HCG (NRL test #004416) result was 12 mIU/mL (reference range is 0-5 mIU/mL). The hospital continued with the surgery, but sent the patient for follow-up with her ob/gyn provider who confirmed the elevated HCG result and lack of pregnancy and then sent the patient to an oncologist to rule out gestational trophoblastic disease.
In this case, the patient was not pregnant and did not have cancer. The problem is that the hospital and ob/gyn provider didn’t understand that HCG reference ranges are not absolute measures of elevated HCG for women of all ages.
Publications2 have shown that post-menopausal women can have serum HCG levels up to 14 mIU/mL as a normal response to lower estradiol levels following menopause. When estradiol decreases, the pituitary responds by increasing FSH and LH due to the loss of the estradiol feedback inhibition. In a few patients, HCG or an HCG immunoreactive material is produced by the pituitary as part of the over-production of LH and HCG.
In this case, the patient could have been spared extra medical expenses had the providers understood that low elevations of HCG are normal in post-menopausal women. Dr. Snyder3developed an algorithm to identify women with elevated HCG due to post-menopausal production of HCG. In the algorithm, the sample is also tested for serum FSH (NRL test #004309) and if FSH is greater than 20 mIU/mL and the HCG is between 5 and 14 mIU/mL, pregnancy is unlikely. Dr. Snyder recommends women over 55 years should not be tested for pregnancy.
Example 5. The Wrong Test is Ordered, and a Patient has Incorrect Gestation Diabetes Glucose Tolerance Result
This is a general example. Many providers want to order one of the gestational glucose tolerance tests and they just order glucose tolerance. This can be confusing to the laboratory staff that have to interpret test orders, and this usually results in a call to the client the next day to assure the correct test is ordered. Different tests have different reference ranges, and therefore a false interpretation can occur.
There are two main reasons to perform a glucose tolerance test. One reason is to determine if a person has type 1 diabetes mellitus. The other reason is to determine if a pregnant lady has gestational diabetes.
The diagnosis of non-gestational type 1 diabetes is now usually made without a glucose tolerance test. The American Diabetes Association diagnosis type 1 diabetes based on any one of the four criteria below:
A1C > 6.5%. The test should be performed in a laboratory using a method that is NGSP certified and standardised to the DCCT assay.* (NRL test #001453)
Fasting Plasma Glucose > 126 mg/dL (7.0 mmol/L). Fasting is defined as no caloric intake for at least 8 h.* (NRL test #001818)
2-h Plasma Glucose > 200 mg/dL (11.1 mmol/L) during an oral Glucose Tolerance Test (OGTT). The test should be performed as described by the World Health Organization, using a glucose load containing the equivalent of 75 g anhydrous glucose dissolved in water.* (NRL test #101200)
In a patient with classic symptoms of hyperglycemia or hyperglycemic crisis, a random plasma glucose > 200 mg/dL (11.1 mmol/L). (NRL test #001302)
*In the absence of unequivocal hyperglycemia, results should be confirmed by repeat testing.
Testing for gestational diabetes is more complicated as a number of organisations have different guidelines to make the diagnosis4. An example is the ADA One-step strategy (NRL test #101000):
Perform a 75-g oral glucose tolerance test, with plasma glucose measurement when patient is fasting and at one and two hour, at 24-28 weeks of gestation in women not previously diagnosed with overt diabetes
The OGTT should be performed in the morning after an overnight fast of at least eight hours
The diagnosis of GDM is made when any of the following plasma glucose values are met or exceeded:
Fasting: 92 mg/dL (5.1 mmol/L)
1 h: 180 mg/dL (10.0 mmol/L)
2 h: 153 mg/dL (8.5 mmol/L)
In summary, good communication practices between medical professionals are important in order to ensure best patient outcome. Providers are encouraged to contact the laboratory whenever results are significantly different from what was expected based on clinical conditions. In some instances, specimens may be retested, and additional comments can be added to future reports as part of the continuous quality improvement processes. Moreover, the laboratory staff will know and advise the provider if any explanations were not considered during the interpretation of the results. Depending on the policy of the laboratory, the communication can be channeled through the laboratory’s customer care service. Alternatively, all issues can be communicated directly with the technical staff of the laboratory, via phone and/or email.
1 Knight et al, Clin & Exp Immun, 2005, 141 333 – 337
2 McCudden CR, et al. Clin Chem 2008 54:209-14
3 Snyder JA, et al. Clin Chem 2005 51:1830-5
4 Diabetes Care, January 2015 Vol 38 Supplement 1