Medical Device Link.

Originally Published IVD Technology April 2003

Commentary

Ulf-Håkan Stenman and Henrik Alfthan

Ulf-Håkan Stenman, MD, is the head of the hormone and tumor marker laboratories, and  Henrik Alfthan, PhD, is a clinical biochemist at Helsinki University Central Hospital (Helskinki, Finland). The authors can be reached via  ulf-hakan.stenman@hus.fi and henrik.alfthan@hus.fi, respectively.

Pregnancy test kits are among the most widely used immunoassays. These tests not only are used by medical professionals but also are available over the counter, and their commercial value is considerable. Since patients can do these tests at home, reliable performance and easy interpretation are important factors that should be taken into account. Consumer Reports examined these factors in a recently published article.¹

Consumer Reports evaluated 18 pregnancy test kits for their detection limit, the time required to perform a test, and ease of interpretation. While most of the study’s conclusions are correct, some are questionable. One questionable conclusion concerns sensitivity. The report emphasizes the importance of a test’s sensitivity. Those tests found to have the highest sensitivity can detect human chorionic gonadotropin (hCG) in urine at concentrations as low as 6.5 mIU/ml. A highly sensitive test is also able to detect pregnancy earlier than less-sensitive ones. Most manufacturers claim their tests can detect pregnancy on the first day of a missed period, although the measured detection limits vary between 7 and more than 100 mIU/ml. 

Other studies looking at the increase of hCG levels after in vitro fertilization found that serum hCG can reach a mean level of 80 mIU/ml 12 days after embryo transfer. Even pregnancies ending in an early abortion have hCG levels above 5 mIU/ml 12 days after embryo transfer, which corresponds to about two days before the first missed period in a normal cycle.² 

However, because of variations in cycle length, some pregnancies start later than expected based on menstrual history. In up to 10% of those cases, hCG is not detected at the time of the first missed period, irrespective of the sensitivity of the test.³

The number of misdiagnosed pregnancies is likely to increase when using tests with higher detection limits. So while the Consumer Reports article concluded that sensitive tests are better than less-sensitive ones, this is not necessarily the case.

Determining the Ideal Sensitivity

The optimal sensitivity of a pregnancy test also depends on hCG concentrations in nonpregnant women. Contrary to what is believed, hCG is not limited to pregnant women. Measurable levels of hCG are present as well in the serum and urine of most menstruating women. This hCG is derived from the pituitary gland, and the hCG concentrations vary according to the luteinizing hormone (LH) surges, the levels being on average 5–10% of those of LH.⁴ Thus hCG concentrations increase up to 2–3 mIU/ml during an LH surge. After menopause, levels up to 5–7 mIU/ml can be observed in the serum of healthy women.⁵ At the same time, the average hCG levels in urine are about half of those in serum. However, because urine flow varies, urine is sometimes quite concentrated, and the hCG concentrations can reach levels of up to 10–15 mIU/ml, which means that highly sensitive pregnancy tests can give false-positive results.⁵

Another factor that could cause confusion is the presence of the core fragment of the beta subunit of hCG (hCGbcf) in urine, where its concentrations often are similar to those of hCG. Some pregnancy tests recognize hCGbcf, but this aspect too was not studied in the article. However, only some pregnancy tests based on polyclonal antibodies can react with hCGbcf.

In the final analysis, pregnancy tests that are sensitive enough to give false-positive results in postmenopausal urine should not be marketed. On the basis of the highest hCG levels observed in the urine of postmenopausal women, the detection limits of tests should not be lower than 25 mIU/ml. Typically, hCG concentrations double every 1.5 days during the early stages of pregnancy. Therefore, a pregnancy will be detected 1–2 days later if the test has a sensitivity of 25 mIU/ml, rather than 7 mIU/ml. However, if the test is used on the first day of a missed period, only a small percentage of the viable pregnancies will be missed. In these cases, the test will turn positive within a few days. This is a minor drawback in comparison with the problems and confusion caused by false-positive results for women reaching menopause.

Other Potential Problems

The use of highly sensitive pregnancy tests has also increased the probability of detecting pregnancies ending in an early pregnancy loss, which occurs in about one out of every four pregnancies.^6,7 Unless detected using a sensitive hCG test, this condition will be experienced as a delayed menstruation. This condition does not require treatment, and there is no need to detect it with home pregnancy tests. Misdiagnosis of these early pregnancy losses causes unnecessary worries, until the hCG levels drops again and a repeated test becomes negative. This situation has caused much confusion about and suspicion of false-positive results from highly sensitive tests.

When ultrasensitive pregnancy tests were first introduced in clinical practices two decades ago, a positive result was occasionally misinterpreted as indicating an ectopic pregnancy in women with early pregnancy loss. This was because there were no signs of intrauterine pregnancy when an endometrial curettage was performed to clarify the cause of the irregular bleedings.⁶ This condition is now well recognized by gynecologists. However, a positive pregnancy test followed by delayed menstruation may still cause confusion. Thus, many women with a positive home pregnancy test may be found not to be pregnant when examined later by a gynecologist. This may be interpreted as a false-positive pregnancy test, and this situation becomes more common when using ultrasensitive tests.

Conclusion

Sensitive hCG assays play an important role in managing early pregnancy disorders and monitoring trophoblastic tumors.⁸ High sensitivity is especially useful in cancer diagnostics. However, such high sensitivity should not be used in pregnancy tests. Tests with a detection limit no lower than 25 mIU/ml are preferred.

References

1. “When the Test Really Counts,” Consumer Reports 62 no.2 (2003): 45–50.

2. P Poikkeus, V Hiilesmaa, and A Tiitinen, “Serum hCG 12 Days After Embryo Transfer in Predicting Pregnancy Outcome,” Human Reproduction 17 (2002): 1901–1905.

3. AJ Wilcox et al., “Natural Limits of Pregnancy Testing in Relation to the Expected Menstrual Period,” Journal of the American Medical Association 286 (2001): 1759–1761.

4. UH Stenman et al., “Serum Levels of Human Chorionic Gonadotropin in Nonpregnant Women and Men Are Modulated by Gonadotropin-Releasing Hormone and Sex Steroids” Journal of Clinical Endocrinology and Metabolism 64 (1987): 730–736.

5. H Alfthan et al., “Concentrations of Human Chorionic Gonadotropin, Its b-Subunit and the Core Fragment of the b-Subunit in Serum and Urine of Men and Nonpregnant Women,” Clinical Chemistry 38 (1992): 1981–1987.
6. UH Stenman et al., “Ultrarapid and Highly Sensitive Time-Resolved Fluoroimmunometric Assay for Chorionic Gonadotropin,” The Lancet 8351 no.2 (1983): 647–649.

7. AJ Wilcox, DD Baird, and CR Weinberg, “Time of Implantation of the Conceptus and Loss of Pregnancy,” New England Journal of Medicine 340 (1999): 1796–1799.

8. J Vartiainen et al., “Elevated hCG and a High Proportion of hCGbeta in Serum Preceding the Diagnosis of Trophoblastic Disease by Seven Months,” The British Journal of Obstetrics and Gynecology 109 (2002): 589–590. 

Copyright ©2003 IVD Technology