Pregnancy or isolation of an oocyte from the reproductive tract are definitive evidence that ovulation has occurred,1,2 but they cannot be used clinically as reference methods for predicting or confirming ovulation in infertile women. In common clinical practice, other methods to predict and confirm ovulation are used. Direct methods involve seeing follicle growth and rupture by laparoscopy or high-resolution transvaginal ultrasound examination.3–7 Laparoscopy is invasive and ultrasound monitoring requires inconvenient daily monitoring of follicular growth. Both are very expensive. The indirect methods of predicting ovulation rely on measurement of LH or estrogen in blood,8 urine,9 or saliva,10 ascertainment of basal body temperature (BBT) nadir,11 or analysis of cervical mucus.12 Methods used to detect the LH surge in urine have a higher incidence of false-negative results than radioimmunoassay applied to serum samples, but repeated measurement of circulating LH is invasive and expensive. Indirect methods used to confirm ovulation are the measurement of serum progesterone,13 related changes in BBT,11 or endometrial histology.14 Serum progesterone has at least two drawbacks, the cutoff serum levels of ovulatory progesterone are not well defined and the variability in cycle length dictates the need of two or three samples to avoid underestimation of ovulation frequency. Basal body temperature is an easy, inexpensive, and self-administered method, but can be affected by many factors other than hormonal changes. Endometrial biopsy and dating, besides being invasive, have no correlation with impaired fertility,15 and is used less.
Most studies16–19 estimated ovulation during physiologic and clomiphene citrate–stimulated menstrual cycles using repeated blood or urinary samples for hormonal evaluation or ultrasound assessment of ovarian morphology, and evaluation of BBT. The limitations of those studies are small samples and inclusion of stimulated cycles and of women with menstrual disorders into the final analysis. The major study conducted on women with normal menstrual cycles was the World Health Organization probit analysis8 that analyzed hormonal assessment of 107 women using laparotomy for diagnosing ovulation based on ovarian histology findings.
The aim of our study was to assess the reliability of the most widely used clinical methods for predicting or confirming ovulation in a large group of infertile but ovulatory women. Transvaginal ultrasound monitoring was chosen as the true indicator of ovulation, and we investigated sensitivity, specificity, and accuracy in predicting or confirming ovulation of urinary home stick system for LH surge detection and serum progesterone and of BBT charts in a population of eumenorrheic infertile women.
Materials and Methods
Between June 1994 and June 1998, female members of 101 infertile couples were recruited at the First Department of Obstetrics and Gynecology of the University of Milan. Inclusion criteria were 18–38 years of age at screening, recorded evidence of regular spontaneous cycles between 26 and 34 days since menarche or for at least 12 regular cycles immediately before the study, ovulatory serum progesterone value in the midluteal phase of a previous cycle (at least 8 ng/mL), good physical health, and normal body weight (body mass index = 19–24 kg/m2). During preliminary screening, medical histories were recorded and FSH, LH, progesterone, and prolactin serum concentrations were measured. Women were excluded if their serum FSH and LH concentrations in early follicular phase were higher than 10 mUI/mL and 12 mUI/mL, respectively, or if their prolactin exceeded 20 ng/mL in the midluteal phase. Exclusion criteria also included clinical signs of polycystic ovary syndrome (acne, hirsutism, oligomenorrhea, obesity) or ultrasound evidence of polycystic ovaries according to the criteria of Adams et al,20 any ovarian or abdominal abnormalities that would interfere with adequate ultrasound investigation, and evidence or history of endocrine or other diseases that might influence the menstrual cycle.
When inclusion criteria were met, women were considered eligible and a single menstrual cycle was investigated in each with the use of repeated vaginal ultrasound scans, home LH urinary stick system tests, BBT, and serum progesterone measurements. One hundred one women entered the study and were considered in the final analysis. The mean age (± standard deviation) of subjects was 31.8 ± 3.4 years (range = 24–38 years). During the cycles studied, two spontaneous pregnancies were observed. Each subject was asked to record daily BBT with appropriate thermometers (with an accuracy of 0.10 of a degree and temperature range of 36–38C). Three to fourminute BBT readings were taken on first awakening, at the same time every morning, during the entire study cycle. Subjects were asked to perform the LH urinary stick system test (Clearplan, Farmades, Rome, Italy) at home starting from the day of the cycle on which the ultrasound mean follicular diameter was 14 mm. The LH test was done every 12 hours on the first morning and early evening urine samples (according to the instructions outlined in the package insert in the LH test) until two consecutive positive results were achieved.
To monitor follicular and endometrial growth, transvaginal ultrasound scans were done on day 8 of the cycle for the first time, then every 2 days, and daily from the day on which follicular diameter of 14 mm was seen until there was evidence of follicular rupture or disappearance. Corpus luteum–appearing structures were also recorded. Physicians doing transvaginal ultrasound scans were masked to LH stick results. Women informed physicians of LH stick results only at the end of the cycle. Serum progesterone assessments were scheduled in the anamnestic midluteal phase, on days +6, +8, +10, day 0 defined as 14 days before expected menses, based on the mean length of the menstrual cycle, irrespective of expected ovulation day of the current cycle. Serum progesterone levels were measured with the commercial immunoassay Enzymun-Test ES 607 (Roche/Boehringer Mannheim, Monza, Milan, Italy); the sensitivity of the assay was 0.2 ng/mL (0.64 nmol/L).
Sensitivity, specificity, and accuracy (combined rate of true-positive and true-negative results) were calculated for LH urinary stick results, BBT readings, and serum progesterone assessments.
Follicular development and ultrasonographic evidence of ovulation (follicular rupture or disappearance or presence corpus luteum–appearing structure) were confirmed in 97 of 101 cycles considered (96%). In three cycles (3%), the dominant follicle, after LH surge, did not show morphologic changes that indicated the ovulatory event. In those women the urinary sticks were positive on cycle days 11, 13, and 14, respectively. Cycle length was 28, 30, and 27 days, respectively, serum progesterone levels were 16.4, 13, and 11.7 ng/mL for the first patient, 10.9, 19.2, and 15.5 ng/mL for the second, and 12.5, 9.9, and 4.5 ng/mL for the third. No ultrasonographic evidence of follicular rupture was recorded in those subjects. In one cycle (1%), no evidence of follicular growth was observed.
LH surge was recorded in 100 of 101 cycles (99%). In 99 cases, ultrasonography and LH readings were in agreement (97 ovulatory and one anovulatory cycles). In three cases, no evidence of follicular rupture at ultrasonography was found, despite positive LH readings. Sensitivity, specificity, and accuracy for LH readings were 1.00, 0.25, and 0.97, respectively. In temporal relationship between urinary LH surge and ultrasonographic diagnosis of ovulation and establishing the day of ovulation (day 0) by ultrasonography, the urinary surge of LH was identified on day 0 in 10% of cycles, on day −1 in 46% of cycles, day −2 in 31% of cycles, and day −3 in 13% of cycles. LH surge preceded rupture of the dominant follicle in all cycles. The mean length of the luteal phase of the cycles was 12.9 ± 2.8 days (range = 8–18) according to ultrasonographic findings of ovulation and 14.5 ± 2.6 days (range = 9–19) according to LH surge.
Basal body temperature was not assessable in 11 cases. In the remaining 90, BBT showed a biphasic pattern in 69 (68.3%) cycles and a monophasic or doubtful pattern in 21 (20.7%) cases. Sixty-five ovulatory cycles and two anovulatory cycles according to BBT were confirmed by ultrasound findings. Basal body temperature readings showed 0.77 sensitivity, 0.33 specificity, and 0.74 accuracy for ovulation detection compared with ultrasonography. The timing of nadir of BBT, the lowest point of the curve, followed by 2 or more days with consecutive rises, of which at least one is above baseline, showed a wide variability ranging from 8 days before to 4 days after ovulation; in 26 cases the nadir was recorded the day of the ultrasound rupture of the follicle or afterward.
Circulating progesterone levels were compared with ultrasonographic evidence of ovulation. Sensitivity, specificity, and accuracy of three different concentrations of serum progesterone as confirmatory tests also were calculated by comparing progesterone levels and ovulation based on ultrasonographic readings in the 97 ovulatory cycles (Table 1). Sensitivity, specificity, positive predictive value, negative predictive value, and accuracy for urinary LH, BBT readings, and serum progesterone are summarized in Table 2.
The best method available for ovulation prediction, detection, and confirmation is still being debated. Ovulation detection is fundamental in diagnostic work-up of infertile couples, family planning programs, and optimal artificial insemination timing. Other authors studied infertile women, but frequently included in their analyses women with irregular cycles, those who had induced cycles, or those with anovulation.17–19 Samples in all of those studies were limited; the maximum was 66 women.
Serum LH assessment is a good indirect index of impending ovulation because it rises to a median of 32 h (23.6–38.2) and peaks at 16.5 hours (9.5–23.0) before ovulation.8 We evaluated urinary LH surge as an ovulation predictor using a widely available commercial kit. Ovulation was confirmed by ultrasonographic findings and serum progesterone assessment in 97 of 101 cycles analyzed. In three cases, ultrasound examination was critical to differential diagnosis between ovulation and luteinized unruptured follicle syndrome. Ovulatory follicle diameter varied greatly19; therefore, when it reached 15 mm, daily examinations were needed to establish precise ovulation timing. To confirm ovulation, fewer examinations were needed, but corpus luteum identification was sometimes difficult.
The LH urinary surge showed ovulatory events in 99 cycles (four false-positive findings). In all cycles with follicular ruptures at ultrasonography, events were preceded by urinary LH surges. The LH surge was detected in urine from 72 hours before ovulation to the same day of ultrasonographic disappearance of the follicle. That variability might be explained partly by different frequencies of LH measurements (every 12 hours) and ultrasonographic examinations (every 24 hours). Detection of LH surges in urine samples also might have been delayed with serum LH surge (eg, morning LH detected in evening urine sample). It has been hypothesized that a false early positivity of urinary sticks is caused by possible urinary clearance of unsustained premature LH surges that occasionally occur during late follicular phases.19 Those data confirm the value of LH urinary surges for predicting ovulation (100% sensitivity and 96% accuracy), at least in this study, in which LH kits were initiated when mean follicle diameters on ultrasonography were 14 mm.
Several authors suggested that the earliest serum progesterone in most is within 2 days of follicular rupture. The major problem is the selection of blood sample timing and cutoff values for each sample. Values most reported as indicative of ovulation are those that exceed 16 nmol/L (equivalent to 5 ng/mL) for a minimum of 5 days, or a single value exceeding 32 nmol/L (corresponding to 10 ng/mL) in the midluteal phase.21,22 Hull et al13 suggested that midluteal progesterone concentrations of less than 10 ng/mL are associated with lower pregnancy rates per cycle than progesterone levels above 10 ng/mL, but many others suggested that a single midluteal phase progesterone assessment is insufficient, especially to judge luteal phase adequacy.23,24 In this study we evaluated efficacy of three different cutoff levels of serum progesterone concentrations in three differently anamnestictimed blood samples in midluteal phase to confirm ovulation and test the best day of the cycle for serum progesterone measurement. The results consistently showed good reliability of that method for confirming ovulation. One progesterone assessment in the midluteal phase seems highly effective for confirming ovulation.
Our results confirm that BBT nadir is an inaccurate method for predicting ovulation time25,26 because ovulation was between 6 days before and 4 days after nadir. Biphasic BBT is still a desirable method, although with low specificity, for the confirmation of ovulation because it is a simple, self-administered test for at-home use by patients.
1. Noyes RW, Clewe TH, Bonney WA, Burrus SB, De Feo VJ, Morgenstern LL. Searches for ova in the human uterus and tubes. I. Review, clinical methodology, and summary of findings. Am J Obstet Gynecol 1966;96:157–67.
2. Clewe TH, Morgenstern LL, Noyes RW, Bonney WA Jr, Burrus SB, DeFoe VJ. Searches for ova in the human uterus and tubes. II. Clinical and laboratory data on nine successful searches for human ova. Am J Obstet Gynecol 1971;109:313–34.
3. Queenan JT, O'Brien GD, Bains LM, Simpson J, Collins WP, Campbell S. Ultrasound scanning of ovaries to detect ovulation in women. Fertil Steril 1980;34:99–105.
4. Collins WP. The evolution of reference methods to monitor ovulation. Am J Obstet Gynecol 1991;165:1994–6.
5. Irons DW, Singh M. Evaluation of transvaginal sonography combined with a urinary luteinizing hormone monitor in timing donor insemination. Hum Reprod 1994;9:1859–62.
6. Guida M, Tommaselli GA, Palomba S, Pelicano M, Moccia G, Di Carlo C, et al. Efficacy of methods for determining ovulation in a natural family planning program. Fertil Steril 1999;72:900–4.
7. Pearlstone AC, Surrey ES. The temporal relation between the urine LH surge and sonographic evidence of ovulation: Determinants and clinical significance. Obstet Gynecol 1994;83:184–8.
8. World Health Organization, Task Force on Methods for the Determination of the Fertile Period, Special Programme of Research, Development and Research Training in Human Reproduction. Temporal relationships between ovulation and defined changes in the concentration of serum estradiol-17 beta, luteinizing hormone, follicle-stimulating hormone, and progesterone. I. Probit analysis. Am J Obstet Gynecol 1980;138:383–90.
9. Collins WP, Collins PO, Kilpatrick MJ, Manning PA, Pike JM, Tyler JP. The concentrations of urinary oestrone-3-glucuronide, LH and pregnanediol-3alpha-glucuronide as indices of ovarian function. Acta Endocrinol Copenh 1979;90:336–48.
10. Li TC, Lenton EA, Dockery P, Rogers AW, Cooke ID. The relation between daily salivary progesterone profile and endometrial development in the luteal phase of fertile and infertile women. Br J Obstet Gynaecol 1989;96:445–53.
11. Ferin J. Determination de la pe′riode sterile premenstruelle par la courbe thermique. Bruxelles Med 1947;27:2786–93.
12. Billings EL, Brown JB, Billings JJ, Burger HG. Symptoms and hormonal changes accompanying ovulation. Lancet 1972;1:282–4.
13. Hull MG, Savage PE, Bromham DR, Ismail AA, Morris AF. The value of a single serum progesterone measurement in the midluteal phase as a criterion of a potentially fertile cycle (“ovulation”) derived from treated and untreated conception cycles. Fertil Steril 1982;37:355–60.
14. Jones GS. Some newer aspects of the management of infertility. JAMA 1949;141:1123–9.
15. Wentz AC, Kossoy LR, Parker RA. The impact of luteal phase inadequacy in an infertile population. Am J Obstet Gynecol 1990;162:937–43.
16. Hackelör BJ, Fleming R, Robinson HP, Adam AH, Coutts JR. Correlation of ultrasonic and endocrinologic assessment of human follicular development. Am J Obstet Gynecol 1979;135:122–8.
17. Vermesh M, Kletzky OA, Davajan V, Israel R. Monitoring techniques to predict and detect ovulation. Fertil Steril 1987;47:259–64.
18. Fossum GT, Vermesh M, Kletzky OA. Biochemical and biophysical indices of follicular development in spontaneous and stimulated ovulatory cycles. Obstet Gynecol 1990;75:407–11.
19. Luciano AA, Peluso J, Koch EI, Maier D, Kuslis S, Davison E. Temporal relationship and reliability of the clinical, hormonal, and ultrasonographic indices of ovulation in infertile women. Obstet Gynecol 1990;75:412–6.
20. Adams J, Franks S, Polson DW, Mason HD, Abdulwahid N, Tucker M, et al. Multifollicular ovaries: Clinical and endocrine features and response to pulsatile gonadotropin releasing hormone. Lancet 1985;2:1375–9.
21. Landgren BM, Unden AL, Diczfalusy E. Hormonal profile of the cycle in 68 normally menstruating women. Acta Endocrinol Copenh 1980;94:89–98.
22. Crosignani PG, Collins J, Cooke ID, Diczfalusy E, Rubin B. Recommendations of the ESHRE workshop on ‘Unexplained Infertility.’ Anacapri, August 28–9, 1992. Hum Reprod 1993;8:977–80.
23. Rosenfeld DL, Chudow S, Bronson RA. Diagnosis of luteal phase inadequacy. Obstet Gynecol 1980;56:193–6.
24. Cumming DC, Honore LH, Scott JZ, Williams KP. The late luteal phase in infertile women: Comparison of simultaneous endometrial biopsy and progesterone levels. Fertil Steril 1985;43:715–9.
25. Hilgers TW, Bailey AJ. Natural family planning. II. Basal body temperature and estimated time of ovulation. Obstet Gynecol 1980;55:333–9.
26. Quagliarello J, Arny M. Inaccuracy of basal body temperature charts in predicting urinary luteinizing hormone surges. Fertil Steril 1986;45:334–7.