Restorative reproductive medicine for infertility in two family medicine clinics in New England, an observational study

Our study adds to the growing literature supporting an integrated RRM approach to infertility. It suggests that the RRM approach in a primary care setting can identify underlying causes or contributing factors for infertility and that treatment results in healthy births for a significant proportion of couples.

This is the first study that has examined the impact of woman’s BMI within RRM, which we investigated as a secondary analysis. Women who had BMI < 25 had 40% probability of live birth, compared to 16% probability for women with BMI ≥25. This is concordant with studies showing lower live birth rates among infertile women with a high BMI who undergo other types of fertility treatment, including intrauterine insemination, IVF, or simply ovulation induction [24‐26]. Future studies of RRM and all fertility treatments should continue to examine this important risk factor.

Women > 35 years of age who had been trying from 6 months to 1 year based on standard clinical definitions and recommendations [17], had substantially higher rates of pregnancy leading to live birth than all women who were trying for more than 1 year. Because we focused on infertility, we did not include women under 35 years of age trying for less than one year. Regardless of age, we believe it is reasonable that all women with greater than 6 months of trying begin tracking their fertility cycles and consider RRM evaluation to facilitate evaluation and achieving healthy pregnancy more quickly.

Without a control group we cannot identify the untreated spontaneous birth rate, which limits the ability to infer the impact of treatment. Because the patients were received predominantly by referral rather than being population-based, we believe a spontaneous birth rate for a referral population is a more relevant comparison than one for a population-based primary care practice; the former is about 50% lower than the latter [27]. The proportion with live birth (10%) among those who did no fertility charting might be one surrogate comparison for a minimal intervention group, but the number of couples in this group was small (22 couples). Future studies should seek more robust comparison groups, possibly of different treatments, because couples seeking medical attention for infertility are usually not willing to go without any treatment.

The diagnoses prior to RRM evaluation (Table 2) were reported by patients (who may not remember all diagnoses they were given) or sometimes available from prior medical records. Diagnostic criteria and the intensity of diagnostic evaluation will vary between different practices, and the patients came from many prior practices. Therefore, the comparisons between diagnoses before and after evaluation should be considered as descriptive and perhaps suggestive, and certainly not definitive.

Although all patients received RRM evaluation, not all couples availed themselves of fertility tracking. Creighton Model Fertility Care System tracking is the foundation of NPT [13, 28, 29]. Most study patients (80%) used the Creighton Model. However, 14% of couples used other types of fertility tracking, primarily the Sympto-Thermal method, which tracks cervical fluid, bleeding, and basal body temperature [22, 23]. We did not find a difference in proportions of live births between these 3 groups, (30% vs. 30%, respectively). Further research is needed to define the potential impact of different types of fertility tracking [30, 31].

Encouraging infertile couples to continue for a full trial of treatment represents a challenge. Over half of our study couples discontinued treatment before 2 years. This rate of discontinuation is similar for that of other infertility treatment cohorts, both RRM and ART [9, 15, 16, 32, 33].

These are the first published data on RRM for infertility from family physicians in the United States, and complements previously published data from family physicians in Ireland and Canada. The cumulative adjusted proportion of live births in those studies was 53% in Ireland (32% for a separate study of only couples who had previous IVF), and 66% in Canada [9, 15, 16]. The reported risk factors measured in the Canadian study were similar to those reported here, and the reasons for the lower proportion with live birth in the present study are unclear. One possible difference between populations could be the women’s BMI, which was not reported in the Irish or Canadian studies. There is also variation in interventions. For example, in the Irish group, follicular ultrasound tracking was used routinely, whereas it was rarely used in these New England practices. This may correlate with a less aggressive approach to ovulation stimulation in the patients in this study.

Another difference relates to inclusion criteria. The criteria for this study were intentionally broad (only 1 visit and 1 lab test). In both the Irish and Canadian studies, at least 2 clinic visits were required for inclusion. The survival analysis should adjust for early drop out, but if those who dropped out after one visit from the present study had a lower potential for pregnancy than those who continued for two visits or more, this could contribute to a lower cumulative pregnancy probability in the present study, compared to studies that required two visits to be entered in the study, and thus excluded the couples who had only one visit. Finally, differences in the duration of follow-up may contribute to differences in survival analysis probabilities.

A broad range of interventions was used in these patients (see Table 3). A review of the available evidence for each specific intervention to improve normal reproductive function is beyond the scope of this paper. We also lack adequate timing data to construct the statistical models to identify treatments that may be most successful for couples with various underlying diagnoses. These issues should be addressed in future and larger prospective studies.

The proportions of births with prematurity (8%) and low birth weight (2%) were very low, and compared very favorably to the recent Massachusetts state average prematurity rate of 8.6% in 2014 [34]. This supports the concept that RRM treatment, by identifying and rectifying underlying chronic disease processes, can result in better maternal and newborn outcomes compared to ART. Even when comparison is restricted to singleton births, IVF in the U.S. was associated with rates of prematurity and low birth weight of 30.9 and 26.7%; and artificial insemination with rates of 15.9 and 12.2%, respectively [35]. One of the fundamental principles of RRM is that value is added to the treatment process by increasing the probability of a healthy pregnancy and neonate [9, 10, 15, 16, 36]. This study gives insight into common underlying conditions that are diagnosed and treated with an RRM approach. We suggest that for best outcomes for the woman, the couple, and the newborn, infertility should be approached as a symptom resulting from multiple, identifiable, chronic underlying causes. This perspective is well suited for primary care settings. Another advantage of RRM is lower cost, particularly relative to IVF. Future cost-effectiveness analyses should include the costs of prenatal, perinatal, neonatal and pediatric care. It is important to study the long-term outcomes of these techniques to ascertain whether RRM treatment leads to better future health for women, men, and their children, and perhaps also lower healthcare costs.