FERTILITY AND STERILITY威

REPRODUCTIVE ENDOCRINOLOGY VOL. 72, NO. 5, NOVEMBER 1999

Copyright ©1999 American Society for Reproductive Medicine
Published by Elsevier Science Inc.
Printed on acid-free paper in U.S.A.

Efficacy of methods for determining

ovulation in a natural family planning
program
Maurizio Guida, M.D., Giovanni A. Tommaselli, M.D., Stefano Palomba, M.D.,
Massimiliano Pellicano, M.D., Gianfranco Moccia, M.D., Costantino Di Carlo, M.D., and
Carmine Nappi, M.D.
Department of Obstetrics, Gynecology and Physiopathology of Reproduction, University of Naples
“Federico II”, Naples, Italy

Objective: To evaluate the efficacy in ovulation detection of methods used in natural family planning in
comparison with pelvic ultrasonography.
Design: Prospective analysis of ovulation detection by natural family planning methods and ultrasonography.
Setting: Natural family planning clinic, Department of Obstetrics and Gynecology, University of Naples
“Federico II”.
Patient(s): Forty healthy women who were highly motivated to use natural family planning.
Intervention(s): None.
Main Outcome Measure(s): Transvaginal ultrasonographic findings, urinary LH levels, salivary ␤-glucu-
ronidase activity, salivary ferning levels and characteristics of cervical mucus, and BBT.
Result(s): Urinary LH level determination yielded a 100% correlation with the simultaneous ultrasono-
graphic diagnosis of ovulation. Mucus sensations and characteristics yielded a 48.3% correlation when
simultaneously evaluated with ovulation. ␤-Glucuronidase levels yielded a 27.7% correlation. The salivary
ferning test had a 36.8% ovulation-detection rate the day of ovulation, but 58.7% of results were uninter-
pretable. Body temperature measurements yielded a 30.4% correlation with the simultaneous ultrasonographic
diagnosis of ovulation.
Conclusion(s): Measuring urinary LH levels is an excellent method for determining ovulation. Although
variations in mucus characteristics and basal body temperature correlate somewhat with ovulation, the length
of the fertile period is overestimated with these methods. The salivary ferning test and measurement of
␤-glucuronidase levels are not good methods for home ovulation testing. (Fertil Steril威 1999;72:900 – 4.
©1999 by American Society for Reproductive Medicine.)
Key Words: NFP methods, ovulation self-detection, fertility control

Received November 23,
1998; revised and
accepted June 18, 1999.
Reprint requests: Maurizio
Guida, M.D., Via U. Ricci 3,
80100 Naples, Italy (FAX:
39-81-645893; E-mail:
pellican@unina.it).
0015-0282/99/$20.00
PII S0015-0282(99)00365-9
Controlling reproduction is one of the major
issues in social economics and preventive med-
icine today. Limitation of the number of births
is essential in underdeveloped countries. Un-
fortunately, economic problems there will
hinder the use of new, expensive hormonal or
surgical contraceptive methods in the next de-
cade (1). In developed countries, pharmaco-
logic and instrumental interventions are be-
coming less accepted, there is a need
for nonmedical contraception, and researchers
seek to gain a better understanding of physiol-
ogy; in these countries, the need for nonhor-
monal, noninvasive methods is well docu-
mented (1).
New developments in the area of natural
family planning have emerged in recent years
(1). The World Health Organization (2– 6) has
played a leading role in the organization of
studies for the evaluation of methods for con-
trolling human fertility. At the same time, basic
and clinical research are being carried out to
discover new technologies that can support,
from a scientific point of view, natural methods
in family planning.
We organized a natural family planning pro-
gram to support a group of selected couples
with information about and training in two
sympothermal methods and to evaluate the ef-
fectiveness of these two methods. Evaluation

900
of the results of this program is still in progress. In addition,
we also evaluated single methods of self-detection of ovu-
lation compared with an objective ovulation-determination
technique (pelvic ultrasonography). We report the efficacy of
these methods in detecting ovulation in the study population.
MATERIALS AND METHODS
Subjects
Forty women were enrolled in the natural family planning
program (mean [⫾SD] age, 28.4 ⫾ 3.8 years; range, 21– 42
years). Exclusion criteria were infertility or subfertility (each
women had had a previous pregnancy and had no history of
recurrent miscarriage), any medical illness (e.g., diabetes,
cardiovascular disease, or gastrointestinal disease), oral con-
traceptive use in the last 3 months, and any behavior that
indicated that the couple was unreliable in terms of follow-
ing natural family planning rules. To increase the possibility
of compliance, we recruited our subjects from groups with
high motivation in this area, such as religious groups and
university students. Couples were allowed 6 months to be-
come familiar with all of the techniques used in the study.
We did not request institutional review board approval,
because all patients referred to our clinic had requested
natural family planning methods for contraceptive purposes
and did not receive any drugs during the study. All women
gave informed consent.
Ovulation-Detection Methods
During each cycle, the following ovulation-detection
methods were used:
1. Transvaginal ultrasonography. This technique was per-
formed to determine ovulation objectively. An ultrasonic
scanner (SSA250-A; Toshiba, Medical Systems, Rome, It-
aly) with a 7.5-MHz vaginal probe was used, and the scans
were performed daily, starting 7 days after the onset of
menses, by the same operator (G.A.T.), who was blinded to
the results of the other tests. The day of ovulation was
retrospectively identified as the day before a corpus luteum
was observed.
2. Daily morning urinary LH level determinations. Levels were
determined by the subjects themselves, beginning on day 6 of
the cycle. Subjects used a stick system (Clearplan; Farmades,
Rome, Italy) based on monoclonal antibodies directed versus
LH bound to a colorimetric substance.
3. Determination of salivary ␤-glucuronidase activity. This
evaluation was done in our laboratory on specimens that
were collected by the subjects starting on day 6 of the cycle
and then frozen. The phenolphtalein test and titolation with a
colorimetric assay (7) were used. The occurrence of ovula-
tion was hypothesized when the salivary concentration of the
enzyme reached, for the first time from the beginning of the
cycle, 500 pmol of free phenolphtalein per microgram.
4. The salivary ferning test. Salivary ferning was evaluated
daily throughout the cycle with an optical microscope. Dried,
unstained specimens were used. A score was assigned; 0 was
FERTILITY & STERILITY威
assigned when salivary ramifications were absent, 1 when
there was first-order ferning, 2 when there was second-order
ferning, and 3 when there was third-order ferning (8). A
score of 1 or 2 was considered indicative of the periovulatory
period, and a score of 3 suggested ovulation.
5. Determination of cervical mucus levels and characterization
of cervical mucus. The vulva was classified as dry, moist, or
wet, and mucus was classified as absent, sticky, or stretchy
(9). Identification of the last day in which fluid mucus was
observed (the so-called mucus peak (9) was performed ret-
rospectively, considering the last day in which there were a
wet, slippery sensation and/or transparent, stretchy mucus.
6. Measurement of rectal or oral BBT. Measurements were
done with a digital thermometer (Terumo, Tokyo, Japan)
after at least 2 hours of bed rest. Each subject measured BBT
at the same site throughout the study.
Sexual Activity
Couples were instructed to abstain from sexual inter-
course as soon as one of the methods indicated the beginning
of the fertile period. They were allowed to resume sexual
activity only after all methods had indicated the end of the
fertile period. To avoid the necessity for prolonged sexual
abstinence, the use of a barrier method (i.e., condoms) was
permitted.
Data Recording
Women recorded urinary LH levels, salivary ferning
scores, characteristics and sensations at the vulva of cervical
mucus, and daily BBTs along with occurrences of discharge
of vaginal blood and frequencies of sexual intercourse. All
values were entered with ultrasonographic data and ␤-glu-
curonidase levels onto a spreadsheet program (Excel; Mi-
crosoft, Redmond, WA.), to compare the coincidence of the
ovulation day determined by each method with the ovulation
day determined by ultrasonography.
The day of ovulation determined by ultrasonography was
considered day 0. If a method indicated ovulation on the
same day, simultaneous correlation with ovulation was con-
sidered to exist. For each subject, we calculated the percent-
age of simultaneous correlation for each method in each
cycle throughout the study. We then calculated the overall
simultaneous correlation for all methods. We also evaluated
the presence of indication of ovulation for each method over
a length of time, day by day, from 6 days before to 6 days
after ovulation as diagnosed by ultrasonography.
Statistics
To identify significant differences between the methods
used for detection of ovulation, we used the instrumental
estimation test. With this test, the accuracy and the precision
of each method were compared with those of the other
methods. The desired outcome was that the method have
relatively the same accuracy (as measured by the mean) and
the same precision (as measured by the variance).
In this study, to compare transvaginal ultrasonography
with other methods for detection of ovulation, each method
901

FIGURE 1
Timing of ovulation as determined by the different methods
throughout a 12-day periovulatory period. } ⫽ urinary LH
kit; ■ ⫽ ␤-glucuronidase level determination; Œ ⫽ salivary
ferning test; ⫻ ⫽ determination of mucus characteristics;
ⴱ ⫽ BBT measurement.
Guida. Determining ovulation. Fertil Steril 1999.
was used in every cycle. Successively, on a cycle-by-cycle
basis, the values from the two methods were added and
then subtracted. A statistical test was previously devel-
oped by Maloney and Rastogi (10), who showed that sums
and differences such as these could be used to detect
differences in the variance (precision) of two methods.
The F test for comparing variances from two independent
samples was not used, because values were obtained in the
same cycle from each method; thus, the sample was not
independent.
Differences in accuracies (mean) were examined with the
use of the matched pairs test.
RESULTS
A total of 148 menstrual cycles were completed by the
subjects studied. The mean (⫾SD) number of cycles per
woman was 3.82 ⫾ 0.78 and the median number of cycles
per women was 4, with a lower quartile of 3, an upper
quartile of 4, and a quartile range of 1. Ultrasonography was
performed in all 148 cycles. Urinary LH level determinations
were performed in 98 cycles (66%), ␤-glucuronidase level
determinations in 95 (64%), salivary ferning observation in
125 (84%), and mucus evaluation and BBT measurements in
all 148 cycles (100%).
Figure 1 shows the timing of ovulation throughout the
cycle as determined by each method.
Urinary LH and ␤-glucuronidase levels yielded a 100%
and 27.7% correlation, respectively, with the ultrasono-
graphic diagnosis of ovulation. The salivary ferning test had
902 Guida et al. Ovulation and natural family planning
a 36.8% simultaneous correlation with ovulation but had a
high percentage (58.7%) of uninterpretable pattern. Mucus
sensation and characteristics and body temperature measure-
ments yielded a 48.3% and 30.4% correlating, respectively,
when simultaneously evaluated with ovulation.
Ovulation timing determined by urinary LH level mea-
surement always coincided with ultrasonographically de-
tected (actual) ovulation measurement of salivary ␤-glucu-
ronidase levels and indicated ovulation 3 days after actual
ovulation in 27.7% of cases and 2 days after actual ovu-
lation in 16.6% of cases. The salivary ferning test indi-
cated ovulation the day before actual ovulation in 21% of
cases and the day after in another 21%. Determination of
mucus sensation and characteristics indicated ovulation
the day after actual ovulation in 27.4% of cases. Finally,
body temperature measurements indicated ovulation the
day before ultrasonographically detected ovulation in
32% of cases.
There was no difference in the accuracy and precision of
urinary LH level determination and those of transvaginal
ultrasonography (P⬎.05). For the other methods, analysis
of variance showed that precision differed significantly from
that of ultrasonography.
Analysis of accuracy (mean) showed that the accuracy of
determination of mucus sensation and characteristics was not
significantly different from that of transvaginal ultrasonog-
raphy (t ⫽ ⫺1.23; P⬎.05, two-tailed). Measurement of
salivary ␤-glucuronidase levels (t ⫽–3.38) and measurement
of basal temperature (t ⫽ 4.31) are not good methods for
determination of ovulation, because their accuracies differ
significantly from that of transvaginal ultrasonography
(P⬍.05). The salivary ferning test did not significantly
differ from ultrasonography in terms of accuracy (t ⫽ 0.66;
P⫽.506); however, 58.7% of results were uninterpretable
and thus excluded from analysis.
During this study there were two unwanted pregnancies,
yielding a Pearl index ([number of pregnancies ⫻ number of
cycles observed]/1,300) (1) of 0.22.
DISCUSSION
Natural family planning methods are widely used today.
In Peru and Mauritius, nearly one in five women reported
using natural family planning methods (11). In 1998, 4% of
fertile married women in the United States relied on these
methods (12). In Italy, 16% of fertile couples use periodic
abstinence as a means of contraception (13). Nevertheless,
most of these couples use improper methods to determine the
fertile period of the cycle, mostly because of inadequate
information or training. It is therefore important to find out
which of the methods used in natural family planning is the
most effective for determining ovulation.
From the data gathered in this study, it is clear that the
Vol. 72, No. 5, November 1999
most accurate method is self-determination of urinary LH
levels using a home kit. Another method for determining the
fertile period, which seems to be correlated with the peri-
ovulatory period, is observation of mucus characteristics.
Measurement of salivary ␤-glucuronidase levels cannot be
considered reliable for detecting the ovulation day, given
that it indicated ovulation both 2 and 3 days after ultrasono-
graphically detected ovulation in significant percentages of
subjects. Basal body temperature is capable of indicating the
fertile period, even though detected ovulation days may be
scattered throughout the periovulatory period, thus inducing
extended periods of sexual abstinence. The salivary ferning
test proved capable of indicating ovulation, but only after the
exclusion of more than half of the results (58.7%), which
were uninterpretable.
The effectiveness of measurement of urinary LH levels
for the purpose of monitoring ovarian activity and detecting
ovulation is well documented and was first confirmed in the
area of infertility. Martinez et al. (14) evaluated the results of
IUI timed either after the detection of a urinary LH surge at
home by the patients or after a positive LH test interpreted by
a gynecologist. The authors concluded that a home urinary
LH test was a reliable device for the prediction of ovulation,
because the patients’ and gynecologist’s results agreed in
89% of cases. A similar study was conducted by Robinson et
al. (15), who tried to determine whether a home urinary LH
detection method could predict ovulation and thus decrease
the number of controls required for the management of a
donor insemination cycle. They observed that the use of such
a method reduced the number of visits required and did not
worsen monthly fecundity or cumulative conception rate.
The ability of home urinary LH level determination to detect
ovulation in spontaneous and induced cycles has also been
tested (16). In that study, the relationship between urinary
LH levels, serum LH levels, serum E2 levels, and ultrasonog-
raphy was examined. The results indicated that urine LH
level determination was a good method of predicting ovula-
tion in spontaneous cycles.
In a previous study, our group evaluated the results of the
use of two commercial immunoenzymatic kits for the deter-
mination of urinary LH levels comparing results with ultra-
sonographic findings regarding ovulation (17). Kit A was a
qualitative assay and kit B was quantitative and monophasic.
The first had a sensitivity of 30%, a specificity of 99%, and
an accuracy of 92%, whereas the second had a sensitivity of
80%, a specificity of 90%, and an accuracy of 88%. It is
evident that these results are in agreement with the literature.
The only difference found was a slightly higher accuracy,
which can be explained both by higher motivation in the
group and by the fact that not all of the subjects used kits in
all the cycles, which may have affected the overall simulta-
neous correlation.
Among the other methods evaluated in the current study,
only determination of mucus characteristics showed good
FERTILITY & STERILITY威
correlation with ultrasonography in terms of diagnosis of
ovulation. Billing’s method (or the ovulation method), based
on the observation of the modification of cervical mucus
throughout the cycle, is one of the main natural family
planning methods used and its efficacy has been proved in
several studies [see Guida et al. (1) for a review]. There is a
high degree of skill involved both in teaching and in using
the method. Because patients must be skilled and dedicated
to use this method, it is highly effective.
Determination of BBT in the present study did show a
slight correlation with ultrasonography with regard to diag-
nosis of ovulation. This finding is in contrast with findings
by Martinez et al. (14). These investigators found a true-
positive rate of 90% and a false-negative rate of 2%; 8% of
readings were uninterpretable, reflecting measurement prob-
lems. They also found that the thermal nadir occurred within
1 day of urinary LH surge in 75% of cases and in 90% when
2 days were considered. This discrepancy can be explained
by the fact that BBT measurement has been proved effective
for detecting fertile periods if performed in association with
observation of other signs and symptoms of ovulation and
has been shown to indicate only the periovulatory period.
Indeed, in 94% of cases, BBT-determined ovulation days
were scattered from day–1 to day ⫹3 of actual ovulation,
giving a fairly good indication of the fertile period but
inducing prolonged sexual abstinence.
The salivary ferning test had a good percentage of simul-
taneous correlation, but the high percentage of uninterpret-
able results must be considered as well. Most of the records
did not show the characteristic triphasic pattern reported in
the literature (8), but they frequently showed a monophasic
pattern. This may be due to the inconstant estrogen dilution
in the saliva, yielding different patterns between women and
within the same subject between cycles. This discrepancy
may also be due to the fact that the slides were interpreted by
the subjects themselves only, not also by a gynecologist,
who would be more skilled in interpreting slides; as a result,
there was a higher percentage of uninterpretable results. It is
evident from these data that salivary ferning observation by
natural family planning users cannot yet be relied on for
accurate identification of the fertile period.
There are very few published data regarding the method
of salivary ␤-glucuronidase level determination. The pattern
of this enzyme throughout the cycle has been found to be
triphasic, with the second peak coincident with ovulation (7).
In the present study, however, we observed a biphasic pat-
tern, with a lower peak of coincidence on ovulation day and
a higher peak on day ⫹3. Women may not recognize the first
peak and may thus have unprotected sexual intercourse in
the fertile period. These data seem to indicate that determi-
nation of levels of this salivary enzyme cannot be regarded
as a method for self-identification of ovulation and these
findings confirm the unreliability of salivary enzyme levels
as indicators of ovarian activity.
903
 We believe that the low Pearl index in this study is not
indicative, because our study involved a limited number of
motivated women and a restricted number of cycles. We are
currently conducting a study involving a larger number of
cycles and are evaluating these methods combined into a
modified sympothermal method.
The cost of urinary LH kits is an important issue, given
that one of the aims of natural family planning is to decrease
medical expenses. The prices of LH kits are still high (ap-
proximately $40 –$45 per cycle in Italy), but use of natural
methods as well to determine ovulation should help reduce
the number of determinations per cycle, thus limiting costs.
Moreover, women with regular cycles can use alternative
methods (determination of mucus characteristics, BBT mea-
surement) to determine the period of increased fertility,
limiting the need for LH level determinations to a few days
per cycle. Furthermore, the use and number of LH level
determinations are also linked to the acceptable period of
sexual abstinence. Couples who wish to have the shortest
period of sexual abstinence will need to monitor urinary LH
levels accurately. On the other hand, subjects who accept a
longer period of sexual abstinence can rely also on other
methods that indicate the beginning and end of the fertile
period.
In conclusion, this study seems to confirm that self-
determination of urinary LH levels is a reliable way to
determine ovulation. This method, in association with other
fertility indices, minimizes the period of sexual abstinence
necessary for couples using natural family planning methods
and indicates the right days to have sexual intercourse for
those desiring pregnancy. Furthermore, although varia-
tions in mucus characteristics and BBT have a certain
degree of coincidence with ovulation, the length of the
fertile period is overestimated with these methods. Fi-
nally, the salivary ferning test and measurement of ␤-glu-
curonidase levels are not accurate methods for self-deter-
mination of ovulation.
904 Guida et al. Ovulation and natural family planning
Acknowledgments. The authors thank Sergio Scippacerola, Ph.D., Depart-
ment of Mathematics and Statistics, University of Naples “Federico II”, for
his substantial collaboration in the statistical analysis.
References
1. Guida M, Tommaselli GA, Pellicano M, Palomba S, Nappi C. An
overview of the effectiveness of natural family planning. Gynecol
Endocrinol 1997;11:203–19.
2. A prospective multicentre trial of the ovulation method of natural
family planning. I. The teaching phase. Fertil Steril 1981;36:152– 8.
3. A prospective multicentre trial of the ovulation method of natural
family planning. II. The effectiveness phase. Fertil Steril 1981;36:
591– 8.
4. A prospective multicentre trial of the ovulation method of natural
family planning. III. Characteristics of the menstrual cycle and of the
fertile phase. Fertil Steril 1983;40:773– 8.
5. A prospective multicentre trial of the ovulation method of natural
family planning. IV. The outcome of pregnancy. Fertil Steril 1984;41:
593– 8.
6. A prospective multicentre trial of the ovulation method of natural
family planning. V. Psychosexual aspects. Fertil Steril 1987;47:765–72.
7. Bardin GW, Brown TR, Mills NC, Gupta C, Bullock LD. The regula-
tion of beta-glucuronidase gene by androgens and progestins. Biol
Reprod 1978;18:74 – 6.
8. Guida M, Barbato M, Bruno P, Lauro G, Lampariello C. Salivary
ferning and the menstrual cycle in women. Clin Exp Obstet Gynecol
1993;20:48 –54.
9. Billings JJ (ed.). The ovulation method. Melbourne: Advocate Press,
1980.
10. Maloney CJ, Rastogi SC. Significance tests for Grubbs estimators.
Biometrics 1970;26:671– 6.
11. Stanford JB, Lemaire JC, Thurman PB. Women’s interest in natural
family planning. J Fam Pract 1998;46:65–71.
12. Forrest JD, Fordyce RR. U.S. women’s contraceptive attitudes and
practice: how have they changed in the 1980s? Fam Plann Perspect
1988;20:112– 8.
13. Italian National Council for Research. Fertility survey in Italy. General
report. Vol. 1. Italian National Council for Research, 1982.
14. Martinez AR, Bernardus RF, Vermeiden JP, Schoemaker J. Reliability
of home urinary LH tests for timing of insemination: a consumer’s
study. Hum Reprod 1992;7:751–3.
15. Robinson JN, Lockwood GM, Dalton JD, Franklin PA, Farr MM,
Barlow DH. A randomized prospective study to assess the effect of the
use of home urinary luteinizing hormone detection on the efficiency of
donor insemination. Hum Reprod 1992;7:63–5.
16. Kawano T, Matsuura K, Honda R, Nishimura H, Tanaka N, Okamura
H. Prediction of ovulation by urinary LH surge [in Japanese]. Nippon
Naibunpi Gakkai Zasshi 1992;68:1188 –96.
17. Cardone A, Guida M, Lampariello C, Bruno P, Montemagno U. Ob-
jective and subjective data for fertile period diagnosis in women:
comparison of methods. Clin Exp Obstet Gynecol 1991;19:15–24.
Vol. 72, No. 5, November 1999