Allen J. Wilcox, Donna Day Baird, David Dunson, Ruth McChesney, Clarice R. Weinberg. Natural Limits of Pregnancy Testing in Relation to the Expected Menstrual Period. JAMA. 2001;286(14):1759–1761. doi:10.1001/jama.286.14.1759
Author Affiliations: Epidemiology Branch (Drs Wilcox and Baird) and Biostatistics Branch (Drs Dunson and Weinberg), National Institute of Environmental Health Sciences, Durham, NC; and Department of Biological Sciences, Barnard College, Columbia University, New York, NY (Dr McChesney).
Context Pregnancy test kits routinely recommend testing "as early as the first
day of the missed period." However, a pregnancy cannot be detected before
the blastocyst implants. Due to natural variability in the timing of ovulation,
implantation does not necessarily occur before the expected onset of next
Objective To estimate the maximum screening sensitivity of pregnancy tests when
used on the first day of the expected period, taking into account the natural
variability of ovulation and implantation.
Design and Setting Community-based prospective cohort study conducted in North Carolina
between 1982 and 1986.
Participants Two hundred twenty-one healthy women 21 to 42 years of age who were
planning to conceive.
Main Outcome Measures Day of implantation, defined by the serial assay of first morning urine
samples using an extremely sensitive immunoradiometric assay for human chorionic
gonadotropin (hCG), relative to the first day of the missed period, defined
as the day on which women expected their next menses to begin, based on self-reported
usual cycle length.
Results Data were available for 136 clinical pregnancies conceived during the
study, 14 (10%) of which had not yet implanted by the first day of the missed
period. The highest possible screening sensitivity for an hCG-based pregnancy
test therefore is estimated to be 90% (95% confidence interval [CI], 84%-94%)
on the first day of the missed period. By 1 week after the first day of the
missed period, the highest possible screening sensitivity is estimated to
be 97% (95% CI, 94%-99%).
Conclusions In this study, using an extremely sensitive assay for hCG, 10% of clinical
pregnancies were undetectable on the first day of missed menses. In practice,
an even larger percentage of clinical pregnancies may be undetected by current
test kits on this day, given their reported assay properties and other practical
The practical reference point for the detection of pregnancy is the
first day of a woman's missed period, ie, the day on which she expects her
next period to begin. Unless a woman carries out special tests for ovulation,
her only basis for predicting the onset of her next menses is her usual cycle
length. Most women are able to report their usual cycle length, despite the
natural variability from cycle to cycle. This variability in cycle length
is due largely to variation in the number of days from the beginning of menses
There is no practical way to identify conception before implantation
of the blastocyst. The time of implantation has its own natural variability,
ranging from 6 to 12 days after ovulation.2
Thus, the exact interval from onset of previous menstrual period to ovulation
and then to the detectability of a pregnancy is inherently difficult to predict.
The trophoblast cells of the conceptus produce human chorionic gonadotropin
(hCG) in amounts that increase exponentially following implantation. This
production of hCG provides the basis for all pregnancy tests, including home
test kits. In 1999, about 19 million over-the-counter pregnancy test kits
were sold in the United States, with sales of about $230 million.3 Kits typically instruct women to test for pregnancy
"as early as the first day of the missed period." To evaluate the number of
pregnancies detectable at the missed period, we estimated the day of implantation
relative to the expected first day of the missed period among a group of women
with naturally conceived pregnancies.
We studied 221 healthy North Carolina women who were planning to become
pregnant.4 Women with no known fertility problems
were recruited in 1982-1986 from the local community and enrolled at the time
they discontinued their method of birth control. Women ranged from 21 to 42
years of age; mean age was 30 years, with 5% older than 35 years. Most were
college educated, and 96% were white; further details are provided elsewhere.4 All participants provided informed consent, and the
study was approved by the National Institute of Environmental Health Sciences
institutional review board. Women kept menstrual diaries and froze daily urine
specimens for later analysis. During the study, 151 women conceived a clinical
pregnancy (defined as a pregnancy lasting 6 weeks beyond the last menstrual
Urine samples were assayed for hCG with an extremely sensitive immunoradiometric
assay.5 Detection limit was 0.01 ng of hCG
per milliliter, which is approximately 0.13 mIU/mL.4
This assay was directed primarily at intact hCG, but also cross-reacted with
the free-beta subunit of hCG. The hCG produced by clinical pregnancy was easily
distinguished from background levels by its pattern of exponential rise. We
defined the onset of pregnancy-related hCG as the earliest day of sustained
hCG elevation in which hCG concentration exceeded 0.015 ng/mL. We regard this
as the best available marker of implantation.2
Among the 151 clinical pregnancies, we were unable to determine the
onset of hCG for 9 (in most cases because crucial urine samples were missing).
Of the remaining 142 pregnant women, 136 provided information about their
"usual cycle length." These 136 women provided the basis for the current analysis.
The median reported cycle length was 29 days, with 96% of women reporting
lengths between 23 and 35 days (range, 19-60 days). We previously tested the
validity of self-reported cycle length by comparing it with the observed length
of the woman's follicular phase during the study. The correlation was 0.55.6
Ninety-four of the eligible 136 women reported their usual cycle length
as a single number. We used this usual cycle length to estimate the day on
which a woman would expect her next period (eg, a woman with a usual cycle
length of 28 days would expect her period on day 29.) For women who conceive,
the day on which the next period is expected is the "first day of the missed
period." The remaining 42 women reported their usual cycle length as a range
(eg, "28-30 days"). For these women, we used the upper limit of the range
to determine the first day of the missed period.
Table 1 shows the estimated
day of implantation relative to the first day of the missed period. Implantation
occurred by the first day of the missed period in 90% of pregnancies (95%
confidence interval [CI], 84%-94%). By 7 days after the first day of the missed
period, 97% of all clinical pregnancies had implanted (95% CI, 94%-99%). After
excluding the 21 women who reported that their cycles were "irregular," results
were unchanged (90% and 97%).
Implantation can occur surprisingly late in relation to a woman's expected
menses. For 10% of clinical pregnancies in our study, implantation occurred
after the first day of the next expected period. This represents an insurmountable
limitation of hCG-based pregnancy testing on the first day of the missed period.
A perfectly sensitive assay for hCG could not have detected 100% of these
clinical pregnancies even 10 days after the period was expected (Table 1).
These data describe the physiologic limits of early pregnancy testing.
However, we did not directly test the performance of home test kits. The performance
of a single, qualitative assay in the hands of a layperseron depends on many
factors. One factor is the assay detection limit. While detection limits are
not routinely provided with test kits, a Web site lists assay detection limits
for 40 commercial kits as determined by telephoning the manufacturers.7 These reported assay detection limits of hCG range
from 15 to 100 mIU/mL, with most kits between 25 and 50 mIU/mL. Detection
limits for some kits may be slightly more sensitive than reported by manufacturers.8 Even so, an assay detection limit of 15 mIU/mL is
about 100 times less sensitive than the assay we used to define implantation
(0.13 mIU/mL). This implies that most current test kits would not reliably
detect hCG on the day of implantation. Chard9
estimates that an assay with a detection limit of 25 mIU/mL will begin to
detect pregnancy around 3 or 4 days after implantation.
Other factors also affect the performance of test kits. Urinary dilution
may reduce detection limits.10 Not all test
kits measure the same components of hCG; some measure only intact hCG, while
others measure intact plus the free beta.11,12
The ratio of these hCG components may vary from pregnancy to pregnancy,11 which could affect the detection limits of specific
tests. In addition, user errors have been reported to contribute to false-negative
The instruction to test on the "first day of the missed period" may
not mean the same thing to all women. Women do not know their period is late
until the second day of their expected period. If women think the first day
of their missed period is after they know their period is late, the percent
of false-negative test results would decrease slightly.
How all these factors affect actual test kit performance is hard to
predict. If kits do reliably detect pregnancy by the third day after implantation,
then about one fourth of clinical pregnancies would produce a false-negative
test result on the first day of the missed period. Previous studies do not
provide information on this point because they lacked proper attention to
the timing of the test in relation to expected menses.13
It should not necessarily be the goal of home test kits to acheive the
lowest possible limits of hCG detection. About one fourth of all pregnancies
detectable at implantation fail very early.4
The detection of these events by highly sensitive home test kits would be
of uncertain benefit to women.
The interpretation of a negative pregnancy test result on the first
day of the missed period deserves comment. Some package inserts state that
if a test is negative, "you are probably not pregnant." One says "you are
NOT PREGNANT" (emphasis in the original). This unfounded assurance could have
important consequences. For example, women with a negative test result may
fail to protect themselves from exposures to toxicants in the workplace or
to medications that could damage a developing embryo.
In summary, the timing of implantation varies widely in its relation
to the expected period. Many women will test positive a week or more before
their period is expected, while a few women will test positive only a week
or more afterward. Adolescents and young women are frequent users of test
kits14 but may be especially prone to false-positive
test results because they are at high risk for delayed ovulation.15 Better, information on the limits of early testing
can help balance the costs and benefits of early detection against the risks
of a false-negative test result.