The prognostic power of the post-coital test for natural conception depends on duration of infertility

Cathryn M.A. Glazener1,3, W.C.L. Ford2 and Michael G.R. Hull2,*

1 Health Services Research Unit, University of Aberdeen, Aberdeen and 2 University of Bristol, Division of Obstetrics and Gynaecology, Bristol, UK


    Abstract
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
Controversy about the value of the post-coital test (PCT) has prompted us to re-analyse data from 207 couples, originally studied between 1982 and 1983, with at least 12 months' infertility at presentation, complete diagnostic information and exclusion of female factors, to clarify the effect of duration of infertility on the prediction of conception. In couples with less than 3 years infertility and a positive PCT, 68% conceived within 2 years compared with 17% of those with a negative result. After 3 years, corresponding rates were 14% and 11%. The relative risks of conception [95% confidence interval (CI)] calculated using the Cox's proportional hazards model were 0.23 (0.12–0.43) for a negative PCT (reference positive PCT) and 0.25 (0.13–0.51) for more than 36 months infertility (reference 12–23 months). Semen analysis had no extra predictive power given the duration of infertility and the PCT. The PCT is an effective predictor of conception where defined female causes of infertility are absent and duration of infertility is less than 3 years. Once infertility is prolonged (beyond 3 years) the conception rate is low even with a positive test because a large proportion of couples remaining childless so long have true unexplained infertility. Use of the PCT will enable clinicians to allocate scarce, expensive and invasive resources effectively.

Key words: cumulative conception rates/duration of infertility/post-coital test/semen analysis/sperm function


    Introduction
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
To manage infertile couples effectively, the clinician needs a realistic and practical prediction of their chance of conception without treatment. If this is poor, the couple should proceed to definitive treatment immediately, otherwise they can be advised to continue trying to achieve a natural conception. Thus, diagnostic investigations that refine the clinician's predictive power will promote appropriate and efficient use of resources. Sperm dysfunction is the commonest defined cause of infertility (Hull et al., 1985Go; Cates et al., 1985Go), but its diagnosis remains uncertain. Sperm concentration and morphology, assessed according to World Health Organization criteria (WHO, 1992), were related to the monthly chance of conception in a group of Danish trade unionists attempting to become pregnant (Bonde et al., 1998Go) but in infertility practice the conventional semen analysis is poorly predictive of outcome except when results are very poor (Glazener et al., 1987Go; Polansky and Lamb, 1988Go; Tomlinson et al., 1999Go).

Sperm function tests might offer greater predictive power but at present are recommended only for use in specialized centres (Royal College of Obstetricians and Gynaecologists, 1998). The post-coital test (PCT) requires no expensive equipment and a negative PCT reduced the cumulative chance of conception at 24 months after first attending an infertility clinic almost four-fold (Glazener et al., 1987Go). It could provide a generally applicable test of sperm function. However its value is the subject of intense controversy and a recent trial that compared outcomes in patients randomized to undergo PCT or not appeared to demonstrate that inclusion of the test increased the number of investigations the couples were subjected to but had no effect on outcome. The authors concluded that the test was of no value and indeed could be actually harmful (Oei et al., 1998Go). This study was heavily criticized in subsequent correspondence to the British Medical Journal: Hull and Evers (1999) pointed out that a diagnostic test can only affect outcome if treatment is modified in the light of the result. In Oei's study treatments were applied non-specifically and inconsistently. Since the PCT result did not alter the management of the patients it is hardly surprising that it had no effect on outcome. Oei also failed to give a clear definition of a positive PCT result. The biased selection of references to substantiate the poor prognostic power of the PCT and the ineffectiveness of intrauterine insemination (IUI) was also criticized (Cohlen et al., 1999Go). Attention was also drawn to the value of the test in revealing antisperm antibodies (Hendry, 1999Go). Therefore despite its powerful statistical design, the Oei study does not answer the question of the efficacy of the PCT and this deserves further analysis.

In judging the value of any test of sperm function a number of points must be borne in mind. First, infertility has a range of discrete causes and any test of a single component (e.g. sperm function) can only effectively predict a successful outcome once other causes have been eliminated. To take an extreme example, even a couple with perfect sperm function cannot achieve a pregnancy if the woman's tubes are blocked. Second, incidental factors like age (Schwartz and Mayaux, 1982Go) and duration of infertility (Hull et al., 1985Go; Eimers et al., 1994Go) can be critical and must be controlled for. Furthermore, in the case of the PCT, correlated functions such as sperm production, mucus secretion and coital competence must be accounted for.

A key element in the recent debate has been the effect of duration of infertility on the predictive power of the PCT (Hull and Evers, 1999Go; Oei et al., 1999Go). Our basic hypothesis is that although these factors may be inter-related, duration of infertility is likely to affect prognosis only when a cause for infertility cannot be defined. Couples with a defined cause will always be unlikely to conceive. Those without a defined cause will include fertile couples who have failed to conceive through chance. As time progresses these will achieve pregnancy, leaving those with an unknown but real cause for their infertility who will not conceive even if the defined tests are normal. Consequently, any test of sperm function will be most effective in changing the predicted chance of a couple conceiving when the duration of infertility is short.

In this study, we present the effect of duration of infertility by re-analysing data from a previous study that investigated the power of the PCT and semen analysis to predict natural conception in otherwise normal couples, after complete investigations and control for age (Glazener et al., 1987Go). This paper addresses the question whether the PCT can refine the ability of the clinician to predict the chance of conception in practice.


    Materials and methods
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
We re-analysed previously published data on couples fully investigated as described to confirm ovulatory cycles, normal pelvis (by laparoscopy), normal uterus, normal cervical mucus secretion and normal coital performance and frequency (Glazener et al., 1987Go). In the present study, couples were included only if they had at least 1 year of infertility at first presentation and a complete set of data for analysis. Duration of infertility before presentation and time to conception or loss to follow-up subsequently were based on the date of first presentation at the clinic.

Based on standard semen microscopy, semen quality was classed by an overall estimate of motile, normally developed spermatozoa. This was derived as a composite score by multiplication of the sperm concentration, the percentage motile and the percentage morphologically normal, assessed according to WHO criteria (WHO, 1980), giving the motile normal sperm concentration (MNSC). Previous work showed that it had a better ability to predict natural conception than any other semen analysis parameter or combination thereof (Glazener et al., 1987Go). The cut-off value which best discriminated for fertility was 4x106/ml.

A pre-ovulatory phase PCT was done 6–18 h after coitus as previously described (Hull et al., 1982Go). Mucus was extracted with a standard narrow 1 ml, 10 cm long disposable plastic syringe, without needle, after introducing the neck of the syringe up to its shoulder into the cervical canal. Suction was maintained as the syringe was withdrawn to avoid loss of content and to assess ductility (spinnbarkeit). The volume of mucus was noted before its transfer to a glass microscope slide, where its clarity or turbidity was observed. It was covered with a glass slip and examined at low (x100) and high (x400) magnification. Fully developed mucus (at least 0.3 ml with stretch to 10 cm, and mostly clear to the naked eye) was required to define a negative result, which also required confirmation in a second cycle.

In accordance with evidence presented in a recent review (Oei et al., 1995Go), a negative PCT result was more strictly defined in the current re-analysis as the absence of at least one progressively motile spermatozoon in most high power microscope fields (hpf) of the mucus. The PCT was classified as follows: positive: at least one forward-progressing spermatozoon per field in most (at least five) hpf; negative: less than one forward-progressing spermatozoon in most (at least five) hpf.

Cox proportional hazards regression analysis was used to estimate the relative predictive power for conception of duration of infertility, age, the PCT result and semen analysis parameters. Cox's regression allows the effects of multiple factors to be compared in the same model. It was designed to handle censored data. The coefficients represent the change in the logarithm of the baseline survival function produced by a unit change in the parameter, in this case the extent to which a change in the parameter alters the probability of pregnancy in a given time. The relative risk (RR) was calculated as the natural anti-logarithm of the coefficient. The method assumes that the effect of the parameters remains constant with time. Here this was confirmed by parallel log-minus-log plots (Norusis, 1993Go). Univariate relationships were investigated first, after which all parameters were offered together with their interactions.

Outcomes were graphically presented by time-specific cumulative conception rates to first pregnancy using life-table methods and compared between groups using the Wilcoxon (Gehan) statistic.


    Results
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
When the analysis was restricted to those couples for whom complete data were available (duration of infertility, age, PCT result and MNSC), 207 couples were available for study. Their median duration of infertility prior to first attendance was 24 months (range 12–144) and median age of the women was 28 years (18–42). Although 33% of the women were parous, this did not necessarily relate to their current partner, and therefore was not taken into account.

The woman's age had no effect on the chance of conception (Table IGo) possibly because few older women were included in the study group, and this parameter was omitted from subsequent analyses. Compared to couples with 1–2 years infertility, the chance of conception was significantly lower (P = 0.0002) if the duration was over 3 years but not if duration was between 2 and 3 years (Table IGo). A negative PCT was also significantly associated with a reduced chance of conception (Table IGo). However, conventional semen parameters summarized in the MNSC had no significant effect until the MNSC fell below 2x106/ml and then were only just statistically significant (Table IGo).


View this table:
[in this window]
[in a new window]
 
Table I. Univariate relationship between chance of conception and independent risk factors calculated using the Cox proportional hazards model
 
When all the parameters were offered to regression analysis using the Cox proportional hazards method, duration of infertility and the PCT result were the only variables selected as predictive of fertility. No significant interactions were identified. The relative chance of conception in couples with a negative PCT was about a quarter of that when the PCT was positive [RR 0.23, 95% confidence interval (CI) 0.12–0.43] (Table IIGo). Taking the PCT result into account, couples were four times less likely to conceive when duration of infertility exceeded 3 years (RR 0.25, 95% CI 0.13–0.51).


View this table:
[in this window]
[in a new window]
 
Table II. Multivariate relationships between chance of conception and risk factors calculated using Cox proportional hazards model. Motile normal sperm concentration was also offered to the model but was not selected as a significant predictor
 
Graphic depiction of these results by life table analysis shows that only couples with a positive PCT and a duration of infertility <3 years at presentation had a good chance of conception within the next 2 years (68%). Couples with a negative PCT, more than 3 years infertility or both had a poor chance of success (17, 14 or 11% respectively, Figure 1Go). The effect of the PCT was highly statistically significant (Wilcoxon) in couples with less than 3 years infertility (P < 0.0001) but not in couples with more than 3 years (P = 0.682). By contrast, the semen analysis results as encapsulated in the MNSC had relatively little effect on the chance of conception given the duration of infertility (Figure 2Go), and no statistically significant effect given the PCT result (Figure 3Go). Although the effect of MNSC was statistically significant in couples with <3 years infertility (56 compared with 38%, P = 0.023, Figure 2Go), the conception rates in couples with a negative PCT were low, irrespective of the MNSC below or above 4x106/ml (10 versus 23% respectively, P = 0.064, Figure 3Go).



View larger version (15K):
[in this window]
[in a new window]
 
Figure 1. Cumulative conception rates by duration of infertility at presentation and post-coital test (PCT) result; a positive result required the presence of at least one motile spermatozoon in most high power fields. Error bars shown at staggered 8 month intervals represent 95% confidence intervals. <3y = less than 3 years infertility; >=3y = more than or equal to 3 years infertility; pos = positive PCT; neg = negative PCT.

 


View larger version (18K):
[in this window]
[in a new window]
 
Figure 2. Cumulative conception rates by duration of infertility at presentation and motile normal sperm concentration (MNSC) (sperm concentration 106/mlx% progressively motilex% morphologically normal/10 000). Error bars shown at staggered 8 month intervals represent 95% confidence intervals. <3y = less than 3 years infertility; >=3y = more than or equal to 3 years infertility; MNSC = motile normal sperm concentration (x106/ml).

 


View larger version (19K):
[in this window]
[in a new window]
 
Figure 3. Cumulative conception rates by PCT (a positive result required the presence of at least one motile spermatozoon in most high power fields) and MNSC (sperm concentration 106/mlx% progressively motilex% morphologically normal/10 000). Error bars shown at staggered 8 month intervals represent 95% confidence intervals. pos = positive PCT; neg = negative PCT; MNSC = motile normal sperm concentration (x106/ml).

 

    Discussion
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
In a population of infertile couples, with otherwise normal results after complete investigations, the chance of conception was predicted by their duration of infertility at first presentation and the result of the PCT, but not by semen parameters or the woman's age. Regression analysis using Cox's proportional hazards method suggested that duration of infertility and PCT had independent effects: each decreased the probability of conception by about 80%. Thus, the likelihood of conception given a negative PCT would be even less in couples infertile for more than 3 years, but because of small numbers in this category (n = 19), the statistical power to examine this interaction was low. We suspect that in some couples in this group poor sperm function is the sole cause of their infertility and their chance of conception will not be affected by the duration of infertility. However others may have covert causes in addition, the two factors both acting to decrease the chance of conception as predicted by the proportional hazards model.

The strength of the current study lies in its thorough exclusion of `noise' from confounding female infertility factors and their treatment, a problem highlighted in other studies (Oei et al., 1995Go). A strict definition of the PCT was used, including a new requirement for a `poor-positive' PCT to be confirmed in a second cycle, and subsequent reclassification of these couples into the `negative PCT category' as suggested by more recent work (Oei et al., 1995Go). By focusing on couples where the woman was normally fertile (tubal patency, ovulation, coitus and normal cervical mucus production), the effect of variations in less well-defined factors could be explored. Although these measures reduced the numbers available for analysis we can be confident that the women were normal so far as current diagnostic procedures allowed us to judge. This enabled us to focus on male and on unexplained infertility. If defined causes of female fertility had not been excluded, the conception rate in the positive PCT group would be lower and the apparent power of the test decreased.

In a recent randomized controlled trial of post-coital testing as part of routine infertility investigations, it was found that the chance of conception amongst those who did and did not have a PCT was similar (Oei et al., 1998Go). However, this study also reported that the pregnancy rate was the same whether the women had a positive (38%) or negative (34%) result. The low rate in the positive PCT group could be because of failure to exclude women with confounding causes of infertility but the rate in the negative group remains surprisingly high, possibly reflecting the success of treatments such as IUI and in-vitro fertilization (IVF).

The current study describes the natural conception rates without treatment in otherwise unexplained infertility, and describes practice before the advent of IVF and intracytoplasmic sperm injection (ICSI). It would be difficult to repeat these studies today because of the ethical problem of delaying effective treatment to achieve sufficient follow-up time. In-vitro testing of interaction between spermatozoa and mucus has since been shown to be correlated with in-vitro fertilizing ability of spermatozoa for assisted conception (Hull et al., 1986Go; Aitken et al., 1992Go; Berberoglugil et al., 1993Go; Biljan et al., 1994Go), further underlining the functional ability of the mucus to differentiate between fertile and infertile spermatozoa.

We believe that the PCT is a good test of sperm function when sufficient care is taken to eliminate negative results due to poor mucus quality. Conventional seminology offers much poorer discrimination, the chance of conception only decreasing significantly where the concentration of motile normal spermatozoa is very low.

Once the woman has been proven to be normal, the PCT provides a rational basis for patient management. In couples with less than 3 years infertility, those with a negative PCT can proceed at once to effective treatment such as IUI, IVF or ICSI whereas those with a positive result can be advised to continue to try for a natural conception with good chance of success, thus avoiding these complex and expensive treatments. Couples with more than 3 years infertility should be offered assisted reproduction without further delay.


    Acknowledgments
 
Thanks are due to Linda Hunt and Craig Ramsay for statistical advice.


    Notes
 
* After a short illness, Michael Hull, Professor of Reproductive Medicine and Surgery in the Division of Obstetrics and Gynaecology, University of Bristol, died on 22 November, 1999, aged 60 years. He went to Bristol from London in 1976 as Consultant Senior Lecturer in Obstetrics and Gynaecology with a developing reputation in clinical reproductive endocrinology from his time in London and elsewhere. His contributions to knowledge and practice of reproductive medicine, particularly in all aspects of male and female infertility, brought international renown to his unit. He was appointed to a Personal Chair in 1989. He brought strong ethical principles to his research and clinical practice and campaigned locally and nationally for public understanding and funding of infertility services. Back

3 To whom correspondence should be addressed at: Health Services Research Unit, University of Aberdeen, Polwarth Building, Foresterhill, Aberdeen AB25 2ZD, UK.E-mail: c.glazener{at}abdn.ac.uk Back


    References
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
Aitken, R.J., Bowie, H., Buckingham, D. et al. (1992) Sperm penetration into a hyaluronic acid polymer as a means of monitoring functional competence. J. Androl., 13, 44–54.[Abstract/Free Full Text]

Berberoglugil, P., Englert, Y., Van den Bergh, M. et al. (1993) Abnormal sperm-mucus penetration test predicts low in vitro fertilization ability of apparently normal semen. Fertil. Steril., 59, 1228–1232.[ISI][Medline]

Biljan, M.M., Taylor, C.T., Manasse, P.R. et al. (1994) Evaluation of different sperm function tests as screening methods for male fertilization potential – the value of the sperm migration test. Fertil. Steril., 62, 591–598.[ISI][Medline]

Bonde, J.P., Ernst, E., Jensen, T.K. et al. (1998) Relation between semen quality and fertility: a population-based study of 430 first-pregnancy planners (see comments). Lancet, 352, 1172–1177.[ISI][Medline]

Cates, W., Farley, T.M. and Rowe, P.J. (1985) Worldwide patterns of infertility: is Africa different? Lancet, ii, 596–598.

Cohlen, B.J., te Velde, E.R. and Habbema, J.D. (1999) Post-coital testing. Post-coital test should be performed as routine infertility test (letter). Br. Med. J., 318, 1008–1009.[Medline]

Eimers, J.M., te Velde, E.R., Gerritse, R. et al. (1994) The prediction of the chance to conceive in subfertile couples. Fertil. Steril., 61, 44–52.[ISI][Medline]

Glazener, C.M.A., Kelly, N.J., Weir, M.J. et al. (1987) The diagnosis of male infertility–prospective time-specific study of conception rates related to seminal analysis and post-coital sperm–mucus penetration and survival in otherwise unexplained infertility. Hum. Reprod., 2, 665–671.[Abstract]

Hendry, W.F. (1999) Post-coital testing. Male partner should be assessed (letter). Br. Med. J., 318, 1007–1008.[Free Full Text]

Hull, M.G. and Evers, J.L. (1999) Post-coital testing. Criterion for positive test was not given (letter). Br. Med. J., 318, 1007–1009.[Free Full Text]

Hull, M.G., Savage, P.E. and Bromham, D.R. (1982) Prognostic value of the post-coital test: prospective study based on time-specific conception rates. Br. J. Obstet. Gynaecol., 89, 299–305.[ISI][Medline]

Hull, M.G.R., Glazener, C.M.A., Kelly, N.J. et al. (1985) Population study of causes, treatment, and outcome of infertility. Br. Med. J., 291, 1693–1697.[ISI][Medline]

Hull, M.G.R., Glazener, C.M.A., Wardle, P.G. et al. (1986) Male infertility: sperm penetration of mucus related to natural and in vitro fertility. In Advances in Fertility and Sterility Series, 4th edn. Parthenon, Carnforth, Lancs, pp. 31–38.

Norusis, M.J. (1993) SPSS for Windows. Advanced Statistics Release 6.0. SPSS Inc., Chicago.

Oei, S.G., Helmerhorst, F.M. and Keirse, M.J. (1995) When is the post-coital test normal? A critical appraisal (review). Hum. Reprod., 10, 1711–1714.[Abstract]

Oei, S.G., Helmerhorst, F.M., Bloemenkamp, K.W. et al. (1998) Effectiveness of the post-coital test: randomised controlled trial. Br. Med. J., 317, 502–505.[Abstract/Free Full Text]

Oei, S.G., Helmerhorst, F.M. and Keirse, M.J.N (1999) Authors reply. Br. Med. J., 318, 1009–1010.[Free Full Text]

Polansky, F.F. and Lamb, E.J. (1988) Do the results of semen analysis predict future fertility? A survival analysis study. Fertil. Steril., 49, 1059–1065.[ISI][Medline]

Royal College of Obstetricians and Gyaecologists (1998) Guideline on the Management of Infertility in Secondary Care. RCOG Press, London.

Schwartz, D. and Mayaux, M.J. (1982) Female fecundity as a function of age: results of artificial insemination in 2193 nulliparous women with azoospermic husbands. Federation CECOS. N. Engl. J. Med., 306, 404–406.[ISI][Medline]

Tomlinson, M.J., Kessopoulou, E. and Barrett, C.L.R. (1999) The diagnostic and prognostic value of traditional semen parameters. J. Androl., 20, 588–593.[Free Full Text]

World Health Organization (1980) Laboratory manual for the examination of human semen and semen–cervical mucus interaction. Press Concern, Singapore.

World Health Organization (1992) WHO Laboratory Manual for the Examination of Human Semen and Sperm–cervical Mucus Interaction. Cambridge University Press, Cambridge.

Submitted on March 14, 2000; accepted on May 30, 2000.