Outcomes From Treatment of Infertility With Natural Procreative Technology in an Irish General Practice

Methods

During the study period, the clinic was a single physician office, with the exception of January 2000 to August 2001, when a second physician was practicing in the same office. The 2 physicians were licensed family physicians in Ireland who had received additional training in NPT through Creighton University (Omaha, Nebraska) and had successfully passed all evaluations and examinations of the NPT continuing medical education course. The clinic primarily focuses on providing women's health services.

Data for the NPT treatment cohort were collected from patients during their initial assessment visit and at subsequent follow-up visits and recorded in the routine medical record for the practice. Data from routine follow-up telephone contacts were also included in the medical records. The relevant data for this study were abstracted from medical records and entered into a computerized database, with manual verification of all entered data.

Patients were primarily self-referred couples experiencing difficulty conceiving and achieving a live birth. All patients from Ireland who presented between February 1998 and January 2002 (inclusive) and who proceeded beyond the initial explanatory consultation were eligible for inclusion in this study. Patients who had previously attempted ART (including in vitro fertilization and intracytoplasmic sperm injection) were also included. A few patients were advised after the initial consultation that they were not eligible for NPT, mostly because of azoospermia or menopause; these patients were not included in this analysis. We also excluded patients that had been trying for less than a year or who did not continue with the evaluation that was recommended at the initial consultation.

The NPT infertility treatment implemented in this clinic is a systematic multilevel investigation and treatment program.¹³ It begins with an initial consultation in which reproductive physiology and the various stages of NPT investigation and treatment are explained in detail to patients. They next learn specialized fertility tracking of daily observations of vaginal discharge biomarkers (bleeding and cervical fluid), according to the Creighton Model FertilityCare System.^15,16,30 After patients gained competence in fertility tracking (no earlier than the second month), timed blood samples were taken with respect to the estimated time of ovulation as prospectively identified by the woman's own fertility chart.^13,31 Reproductive hormones, including estradiol and progesterone, were measured and interpreted in the context of the woman's fertility tracking chart. Medications to correct identified hormonal abnormalities were prescribed after the completion of these investigations, usually in the fourth month. The goal was to optimize physiologic conditions for natural conception in each menstrual cycle, including mucus discharge quality, luteal hormone levels, ovulation, and the timing of intercourse. It often takes several cycles of treatment to reach optimal physiologic conditions. Once reached, there may still be some suboptimal cycles mixed with the optimal ones; for example, a cycle where intercourse does not occur during the fertile time. In addition, some patients conceive and subsequently experience a spontaneous abortion, after which they continue in the treatment program. Patients were advised that a total of up to 24 months may be required to complete an adequate trial of NPT to achieve pregnancy leading to a live birth. Evaluation and treatment of the male partner also took place during this time, as indicated. Hormonal assessment and support continued through pregnancy, as indicated.²⁵ Although specialized NPT surgical treatment may also form part of the NPT program, it was not available in this clinic in this time frame. A more detailed discussion of the NPT approach to infertility is available elsewhere.^13,19,32

Per-cycle pregnancy rates are not an appropriate outcome to assess for NPT because of the longitudinal and incremental nature of the NPT treatment program.^33,34 Therefore, we chose a cohort approach with the couple, rather than the cycle, as the unit of analysis. The main analyses for this study were the cumulative proportion of couples experiencing a first live birth during discrete time periods after the beginning of treatment. For comparability with other reports of infertility outcomes, we assigned the time of this outcome as the time of conception leading to first live birth.³⁵ Secondarily, we also analyzed the cumulative proportion of first conceptions and cumulative proportion of withdrawals. Life-table analysis was used to adjust for couples withdrawing from treatment, or that completed 24 months of NPT treatment without pregnancy. Among the live births, we analyzed the proportion of multiple births. Secondarily, we analyzed the proportion with low birth weight and prematurity.

The study protocol was reviewed and approved by the Linacre Centre for Health care Ethics in London, and the Institutional Review Board for Human Subjects at the University of Utah. Because data were abstracted from usual clinical data sources and patient anonymity was maintained, there was no requirement for written informed consent of participants.

Results

A total of 1239 couples obtained an initial consult for NPT during the study period. Of these, 167 couples were excluded because they had tried for less than a year to conceive or because they did not continue evaluation beyond the initial consultation. The final study cohort consisted of 1072 couples that began treatment between February 1998 and January 2002 (inclusive), with subsequent outcomes followed through February 1, 2008. The women were, on average, 35.8 years old (range, 25 to 48 years), and nearly all were white. One third (33%) had previously attempted ART treatment. The mean length of time that couples had attempted to conceive before NPT treatment was 5.6 years (range, 1 to 20 years). Approximately one fourth (24%) of couples had a previous live birth. As shown in Table 1, the 364 couples who ultimately conceived with NPT treatment were slightly younger (mean age, 34.8 years); had not been attempting conception as long (mean duration, 4.8 years); were likely to have had a previous birth (30%); and less likely to have attempted ART treatment (21%).

In addition to teaching women to track their fertility biomarkers with the Creighton Model instruction, the most common diagnoses given to couples before and after NPT evaluation are shown in Table 2. Through NPT evaluation, more than half of the couples had suboptimal serum levels of progesterone and estradiol during the luteal phase. Approximately one fourth had limited cervical mucus and 10% had polycystic ovarian syndrome.

The most common treatments given to women included clomiphene (75.3%), support of luteal hormonal production with human chorionic gonadotropin (67%) or progesterone (18%), and medications to enhance cervical mucus production (71%). Fifty-four women (5%) conceived without medical intervention, using only Creighton Model fertility charting and optimally timed intercourse.

There were 354 clinically recognized conceptions within 24 months after starting NPT treatment, with cumulative crude proportions of live births of 19.1 per 100 couples up to 12 months, and 25.5 up to 24 months, as shown in Table 3. Adjusting for withdrawals from treatment and continuing treatment at the end of study follow-up, the cumulative proportion of first live births was 27.1 up to 12 months, and 52.8 at 24 months. The proportions with any conception (regardless of its outcome) were higher: 25.9 crude and 35.5 adjusted at 12 months, and 33.0 crude and 64.8 adjusted at 24 months. Of conceptions within 2 years leading to a live birth, 75% (205 of 273) occurred within 12 months and 93% (255 of 273) occurred within 18 months.

Several couple characteristics were associated with the probability of live birth, as shown in Table 4. The cumulative proportion of live births declined with increasing age of the woman. For women under 30, the crude proportion of live birth after 24 months was 33.7. For women over 40 the crude proportion was 13.9. The cumulative crude proportion of live birth also declined with increasing previous attempts to conceive (36.6 for previous time of 1 to 3 years; 11.9 for previous time more than 9 years), and with the number of previous ART attempts (30.8 for none, and 10.3 for 3 or more). Women with a previous birth had a higher cumulative crude proportion of live birth (35.8) than those without (23.1). The adjusted proportions confirm the same trends but should not be considered reliable as precise estimates because most of the subgroup life table analyses involve small numbers of women continuing treatment until 24 months (ie, fewer than 25 women).

Among all live births observed, there were 13 twin births (4.6%) and no higher order births. At least 88% of all births were to term and did not have low birth weight (Table 5). None of the patients in this cohort experienced ovarian hyperstimulation syndrome. The proportion of couples withdrawing from treatment was 44.6 per 100 couples by 12 months, and 62.7 by 24 months (Table 1). Among couples who had a live birth, 51 couples had additional NPT treatment, resulting in 47 additional live births within the study period. However, the results reported here are either for the first conceptions or the first live births only.

Discussion

In this study setting, NPT, an integrative approach to infertility provided by NPT-trained family physicians, resulted in substantial live birth rates with a minimal risk of twin or multiple births. These results would suggest that NPT provided by trained generalist physicians has the potential to improve access to treatment for infertility with favorable outcomes. The question remains, however, as to how to interpret these results in relation to outcomes from other infertility treatments.

We reported the outcome of conception because this is an intermediate outcome of interest to patients and because it is an outcome commonly reported in infertility literature. However, we focused on live birth as the major outcome for this study because live birth is the outcome that is most clinically meaningful for patients seeking treatment for infertility.^36,37 Crude birth rates substantially underestimate the efficacy of treatment because of high drop-out rates that are found in most studies of infertility treatment; however, life-table analysis overestimates treatment efficacy because it assumes that those discontinuing treatment have the same prognosis as those continuing treatment, an assumption that may not be valid for infertility treatment.^38,39 Thus, the “true” estimates for NPT success probably lie somewhere between the adjusted life-table estimates and the crude estimates that we have reported.

We believe our study can be considered to represent a “real world” effectiveness based on undifferentiated infertility categories because all patients who were eligible for treatment were included, as would normally occur in clinical practice. However, the population presenting in this study for NPT treatment may represent a more difficult group than would be normally seen even in some specialist ART clinics; 33% of those presenting for NPT treatment had previously attempted ART. In addition, the mean duration of time spent previously trying to conceive for this population was 5.6 years. Finally, this was a relatively older population of women, with a mean age at entry to treatment of just over 35 years. Each of these characteristics is a risk factor strongly associated with lower likelihood of conception.^40–44 In comparison, a recent national Dutch cohort of 4928 women undergoing ART had an average age of 33.6 years and a mean duration of attempting conception of 3.6 years, with a 1-year cumulative probability of ongoing pregnancy of 45%.³⁵ Importantly, the duration of time attempting conception in our study is greater than of all the studies with which we compare in the remaining discussion, and the mean age of the women in our study is also greater than most of them.

Comparison of these results to studies of other infertility treatments must be made tentatively, because results vary by location, diagnosis, and both identified and unidentified risk factors.^45,46 Unfortunately, most data about infertility treatment outcomes are reported on a per-cycle basis, so pregnancy or birth rates may be biased by high rates of discontinuation of treatment, especially for cohorts that have liberal entrance criteria.^33,38 However, some studies of infertility have used a cohort approach and can provide some benchmarks against which to evaluate our results, as detailed below.

Two cohort studies have demonstrated substantial rates of spontaneous conception in some patients with infertility, including a population-based study with a live birth rate of 40%⁴⁰ and a referral clinic population with a live birth rate of 20%,⁴¹ both over 24 months. Both studies, however, included women with an average age of 29 and an average infertility duration of 2 and 3.5 years, respectively. Thus, the prognosis for spontaneous conception in these studies was substantially higher than for the patients in our study. None of the patients in the spontaneous conception studies had previously attempted ART, whereas 33% of the couples in our study had done so.

Another comparison of our results would be with population-based studies of infertility treatment. In a 1985 quasi population-based study, there was an overall conception rate of 48 per 100 couples for all causes of infertility after 2 years of treatment, adjusted by life-table analysis (treatments included ovulation induction, artificial insemination, hormonal treatment, and/or surgery).⁴⁴ Again, the population in this study was much younger (women's mean age, 28 years) than our study; had a much lower duration of infertility before treatment (mean duration, 2.4 years); and excluded couples that had previously attempted ART. Closer to comparison with our study, the subgroup of women with unexplained infertility of 5 or more years’ duration had a 2-year adjusted pregnancy rate of approximately 30%, compared with 53% in our study.

How do these results compare to existing data for ART? The Human Fertilisation and Embryology Authority's crude live birth rate for the year 2000 was 23.8 per 100 women,⁴⁷ which is comparable to the crude live births of 22.7 in this study; but our results are for a longer time frame of up to 2 years. In the United States in 2000, registry data for ART indicated a 25.4% live birth rate per ART cycle initiated (excluding donor eggs or frozen embryos), but no data are available on the number of women treated or follow-up over time.^37,48 Although more recent ART data are available, we have referenced 2000 data for comparability to the time frame during which the patients in this study were treated. Given the differences in time frames, the fact that women's characteristics and longitudinal outcomes cannot be extracted from registry data, and other dissimilarities, this comparison must only be exploratory.

Although life-table analysis is not possible with the existing registry data for ART, other studies have reported cumulative pregnancy or live birth rates in cohorts with multiple attempts at ART. These results varied from a low 32%⁴⁹ to a range of 39% to 60%,^38,50 to a high of 64.7%.³⁹ In this last study the mean age of women was 32.8 years, the mean duration of infertility 3.6 years, and no couples had previously failed ART; all of this suggests that the patients studied had a better prognosis than patients in our study. In a recent population-based cohort of couples receiving ART in the Netherlands, the cumulative pregnancy rate at 1 year was 45%.³⁵ Taken together, these results suggest that the overall success rates of treatment with NPT might be comparable to ART despite greater per-cycle pregnancy rates with ART. Although it may seem paradoxical that a treatment with a lower per-cycle pregnancy rate than ART may have a comparable cumulative rates of pregnancy or live birth, very similar results have also been found in randomized trials of ART versus other less invasive therapies for couples with unexplained infertility.^51,52 A systematic Cochrane review concluded that there are insufficient data to establish that ART is any more effective than less intensive therapies for unexplained infertility.⁵³

Multiple births are one of the most notable and important complications of ART as it is currently practiced.^6,54 The proportion of live births of multiples was much lower in this NPT cohort (4.6%) than for ART treatment in the UK in 1998 to 1999 (27.1%).^47,48 Although multiple birth rates from ART have dropped somewhat, they remain high in most settings, eg, 34% in the US in 2003.^37,48 Longitudinal studies of long-term outcomes after NPT treatment are necessary to determine whether NPT treatment might have fewer long-term health risks for resulting children than ART.

A weakness of our study is that over half of patients withdrew from treatment before completing a full 24-month course of treatment. In comparison, in cohort studies of ART of up to 5 cycles in which cost was not an issue, dropout rates from treatment were 69% and 55%,^38,39 and in a randomized trial of ART versus other treatments, the overall dropout rate was 45%.⁵¹ Thus, the dropout rates from treatment in our study are comparable to those of studies of other fertility treatments. Withdrawing from treatment affects pregnancy or live birth probabilities adjusted by life-table analysis but does not change the crude probabilities that we have reported for all outcomes. Crude probabilities include all couples in the analysis, treating those that dropped out of treatment as if they continued treatment, and therefore represent the most conservative estimate of treatment outcomes. Future studies should seek to minimize the withdrawal rates to the extent possible and to define the prognostic factors among those withdrawing to estimate the differential effect of dropout on adjusted probabilities of pregnancy or live birth.

We emphasize again that comparisons of the results of this study with other available data, as discussed above, are limited by many differences in methods and study populations. More recent studies of ART tend to show slight improvement of pregnancy rates in European countries over the past few years, where there is a strong tendency to transfer fewer embryos, and more substantial improvement in pregnancy rates in the United States, where multiple gestations also remain substantially more frequent than in Europe.^55,56 It remains to be seen whether NPT pregnancy rates may improve with time as well.

Conclusion

This study demonstrates that NPT is an effective integrated system of infertility treatment that can be done by interested and appropriately trained generalist physicians. The cumulative crude pregnancy rates, withdrawal rates, and adjusted pregnancy rates are similar to cohort studies of ART based on the limited comparable data that are available. The treatment program is minimally invasive, with fewer multiple pregnancies. However, it is not an option for couples with azoospermia, ovarian failure, or bilateral fallopian tube occlusion. The probability of live birth with NPT treatment is, as expected, lower for women over age 35, couples who have spent more than 6 years trying to conceive, couples without previous births, and couples who have previously attempted ART; these are risk factors that are also associated with lower live birth rates with other types of fertility treatments.⁴⁴ The most significant risk factor for decreased live birth with ART is increased maternal age.³⁵ Couples choosing NPT need to clearly understand that, although 75% couples that have a live birth with NPT conceive within 12 months (93% within 18 months), it may take up to 24 months for a full treatment course. For women who are approaching the end of their reproductive years, the time frame of up to 2 years required for NPT treatment may be a significant disadvantage of NPT.

Large multicenter prospective studies are warranted to confirm these results, to explore further the characteristics associated with successful NPT treatment, and to directly compare NPT to other forms of infertility treatment.