SUMMARY: OVULATION TEST POSITIVE THEN NEGATIVE NEXT DAY
Women who test positive for ovulation, using an ovulation predictor test (or OPK), and then negative the next day, usually means an LH surge occurred in the past 24 hours and that ovulation may have already happened.
Many women often use commercially-available urinary ovulation predictor tests while trying to conceive. These tests aim to predict ovulation by measuring the rise in luteinizing hormone (LH) in the urine.
In the past, it was assumed that LH peaks 1-2 days before ovulation, however recent studies have shown this peak is more like a wave which can stay high even after ovulation. Therefore, despite being inexpensive, these tests can be easily misinterpreted by their users.
Although several studies have confirmed the validity of these tests among normal women, evidence correlating the optimal urinary LH concentration level with ovulation is still lacking.
The end result is a wide variation in LH test thresholds among the manufacturers of these ovulation predictor tests (or OPK) which potentially influences a couple’s success rate.
To determine the ideal luteinizing hormone (LH) thresholds which predict ovulation within 24, 48 and 72 hours.
Initial recruitment of patients was carried out across 8 natural family planning clinics in France, Italy, Germany, Belgium and Spain.
Women aged between 19-45 years with previous menstrual cycle of 24-34 days were assessed for study inclusion.
The exclusion criteria included:
- Abnormal cycles
- History of anovulatory cycles
- Hormonal treatment for infertility (last 3 months)
- Use of hormonal contraception
- Hormonal replacement therapy (last 3 months)
- Tubal sterilisation
- Pelvic inflammatory disease
- Breastfeeding or postpartum mothers (less than 3 months)
Following assessment, 107 women were selected for this study, with demographic and cycle characteristics recorded.
Hormone assessment was carried out using the first urine discharge of the morning. The urine samples were then frozen and later analysed for estrone-3-glucuronide (E1G), pregnanediol-3a-glucuronide (PDG), follicle-stimulating hormone (FSH), and LH using time-resolved fluorometric immunosorbent assays.
Ovarian scanning was carried out on the first day of cervical mucus, or when an LH surge was detected by the LH home tests, and continued every second day, until the follicle reached 16µm in diameter, and then every day until there was evidence of ovulation.
The outcomes measured included 9 scenarios, specifically:
- If ovulation occurred within 24, 48 or 72 hours after a single positive LH test (followed by a negative test)
- If ovulation occurred within 24, 48 or 72 hours after 2 days of consecutive positive LH tests
- If ovulation occurred within 24, 48 or 72 hours after 3 days of consecutive positive LH tests
A test result was labelled as a ‘positive test’ only if the LH concentration obtained was above a defined threshold else it was considered a ‘negative test.’
If ovulation occurred 24, 48 or 72 hours after the positive test, it was categorised as a true positive. Similarly, if no ovulation occurred 24, 48 or 72 hours after a negative test, it was categorised as a true negative.
The accuracy of these tests was then calculated, specifically the Positive Predictive Value (PPV), which is the proportion of ovulations within 24, 48 or 72 hours after a positive test, and the Negative Predictive value (NPV), which is the proportion of non-ovulations within 24, 48 or 72 hours following a negative test.
The authors also tested 2 hypotheses:
- Whether a given threshold can confirm the end of the fertile window (ovulation within 24 hours), that is by obtaining three daily consecutive negative results below the defined threshold after one positive result.
- Whether the addition of peak-type cervical mucus, to a positive LH test, would increase its predictive value.
Initial analysis of participant demographic characteristics did not find anything noteworthy or potential cause of study bias.
|Age at Menarche (years)||13.2||9||17|
|Body Mass Index (kg/m2)||21.2||17.1||28.3|
|Physical activity (hours/week)||1.1||0||9|
|Regular smokers (%)||11|
|Past use of oral contraception (%)||35|
A total of 326 cycles was monitored from the 102 women in this study. However 43 cycles were excluded from the final analysis, due to the first ultrasound occurring after follicle rupture (n=28) or no ovulation (n=15).
Analysis of included cycles revealed a normal distribution of cycle characteristics with expected mean (average) values.
|Cycle length (days)||28.07||22||44|
|Follicular phase (days)||14.76||9||33|
|Luteal phase (days)||13.35||7||17|
Similarly, the greatest proportion of ovulations occured on cycle day 14, however this is not the norm for most women. In this group, the majority of ovulations (>70%) occured over 5 days, from cycle day 12 to 16.
Statistical (ROC) analysis of true positive rates to false positive rates, between days 9 to 17, showed that out of the 9 original scenarios, the best one to correctly predict ovulation at random was ‘ovulation occurring within 24 hours after a single positive test.’
This indicates women are most likely to ovulate within 24 hours of a single positive ovulation predictor test (OPK) during their menstrual cycle.
Similarly, according to this (ROC) analysis, the ability of ovulation predictor tests to correctly predict ovulation beyond 24 hours of a positive test, or following 2 or more days of consecutive positive tests, was very poor.
In other words, 2 consecutive days of positive ovulation predictor tests (OPK) followed by a negative still means you most likely ovulated within 24 hours of the first positive test.
Further analysis of the optimal scenario showed a distinct change in the accuracy of the ovulation test according to the day of the test. The likelihood of ovulation within 24 hours was most likely to occur following a single positive ovulation test on cycle day 10, which then decreased moderately, every day after that up to day 17.
The authors also noted that participants achieved slightly better accuracy if they used the ovulation tests daily from days 7 to 20 (13 days), compared to only using them from days 13 to 15 (3 days).
The explanation for this is that the LH surge itself (and not LH peak) is a better predictor of ovulation.
Supplementary analysis of all 9 scenarios across varying LH thresholds (15-40 mIU/ml) further confirmed ‘ovulation occurring within 24 hours after a single positive test’ was the most likely scenario to predict ovulation irrespective of LH threshold.
Although, an LH threshold, around 25–30 mIU/ml had the overall best PPV (50–60%), NPV (98%), LR+ (20–30), and LR− (0.5) for this scenario. This also means a first positive ovulation predictor test (OPK) will not correctly predict ovulation within 24 hours in approximately one-third to one-half of cycles.
Next, analysis of the 2 other hypotheses using 25 mIU/ml as the LH concentration threshold revealed:
- That 3 consecutive negative ovulation tests did not correctly predict the end of the fertile window in 31% of cycles (88/283). This means in this group of women, ovulation predictors tests (or OPKs) turned negative before ovulation in 31% of cycles, and turned negative after ovulation in 69% of cycles.
- Peak-type cervical mucus and positive LH test together had better predictive value (Specificity: 97-99%; PPV 34-37%) than either cervical mucus (Specificity: 77-95%; PPV: 20-29%) or LH (Specificity: 91%; PPV: 20%) alone.
Finally, the results demonstrated that while a higher BMI was linked to lower LH levels, it did not affect the predictive value of the ovulation tests.
- Narrow study population limiting the relevance to similar women.
This study was partially funded by a manufacturer of LH home tests, Quidel Corporation, USA.
A laboratory test.
A ratio of the proportion of cycles with ovulation and a negative test, to the proportion of cycles with no ovulation and a negative test.
A ratio of the proportion of cycles with ovulations and a positive test, to the proportion of cycles with no ovulations and a positive test.
The 6 weeks after childbirth.
Receiver Operator Characteristic (ROC)
A graphical plot of sensitivity vs. false positive rate (1-specificity).
The proportion of people without the disease (or condition) who will have a negative result.
Consum Rep. (2003). When the test really counts: part two: the fertility window. https://pubmed.ncbi.nlm.nih.gov/12516589/
Nielsen M S, et al. (2001). Comparison of several one-step home urinary luteinizing hormone detection test kits to OvuQuick. https://doi.org/10.1016/s0015-0282(01)01881-7
Ghazeeri G S, et al. (2000). The predictive value of five different urinary LH kits in detecting the LH surge in regularly menstruating women.. https://pubmed.ncbi.nlm.nih.gov/11092703/
Ovulation disorders are very common and present in 18-25% of couples who see a specialist for infertility treatment. In the general population…. Read more
Polycystic ovary syndrome is a common endocrine condition that affects how a woman’s ovaries work. Described as a low-grade chronic…. Read more