Introduction

Infertility is defined as a failure to conceive after 12 months of regular and unprotected sexual intercourse and is estimated to affect 8–12% of couples of reproductive age worldwide1. With declining birth rates reported in advanced countries, infertility issues will add to economic hardship and are of significant public health importance2. Internationally, it has been estimated that 56% of couples have sought medical care including fertility-related issues and treatment3which is believed to be an underestimate of the overall population that experiences infertility challenges. In such context, Japan is recognized as one of the most active countries in reproductive medicine globally4where nationwide surveys have shown that 1 in 4.4 couples of reproductive age had undergone fertility treatment5. For reproductive endocrinologists, predictive analytics for future treatment are based on the results of numerous examinations and responses to previous treatments, so effective data accumulation and summarization are essential to support decision-making for further fertility treatment.

In Japan, the Japanese Society of Obstetrics and Gynecology (JSOG) authorizes the facilities for assisted reproductive technology (ART) provision, and imposes the responsibility to report treatment procedures and outcomes. This registry, the JSOG registry, cooperates with the public health insurance coverage system for ART and includes nearly all ART facilities (621 of 622 facilities) with more than 99% of ART treatment cycles having been entered in this registry since 20074,6. When patients want to receive fertility treatment that is covered by the public health insurance, they have to receive their care at medical facilities that report their treatment information to the JSOG registry7. Through this mechanism, JSOG provides data accumulation and monitoring for fertility treatment. JSOG implemented an online cycle-based registry system and reports on ART data and analysis annually4,8,9,10,11. In 2020, the total number of registered ART cycles was 449,90012. A strength of this registry is that it comprises accurate clinical data from medical facilities without selection bias because of a high compliance rate. However, there are shortcomings in terms of providing a comprehensive understanding of fertility treatments; information from facilities that do not provide ART is missing, and clinical data for non-ART procedures, such as timed intercourse and AIH, are not collected. Furthermore, additional details experienced by patients are not captured in this registry system. Understanding how patients were treated prior to starting ART affects the interpretation of ART outcomes, so the establishment of a fertility registry to record non-ART related treatment is warranted. In order to obtain a more complete picture of fertility treatment in Japan, data from patients is quite meaningful as it may capture treatments performed at various medical facilities, including those that do not provide ART, in the sequence experienced by the patients. Innovative tools that can capture information directly from patients can complement the shortcomings of the existing registry.

In Japan, there is a female health care smartphone application platform called Luna Luna, which has been downloaded by about 19 million people. Using this application, we recently reported large-scale surveillance results and provided evidence that such a platform can be an effective modality for fertility treatment education in Japan13. In this current study, we have extended the study to address whether a smartphone application platform can assemble clinical data from patients directly, with the expectation that it can contribute to comprehensively understanding an individual’s fertility treatment combined with the existing registry data. We also propose that the findings from this study may provide insight for other countries faced with similar challenges regarding fertility treatment.

Results

Recruitment of participants

More than 7,000 participants were recruited within the first week after notifying the users of this study (Fig. 1a). Extending the recruitment period to cover a full month, the length of a typical menstrual cycle, resulted in a total of more than 13,000 participants. Next, we examined the applications as a mechanism for collecting basic participant information, such as content typically gathered in a medical checkup. The characteristics of the participants are described in Table 1. Most of the participants were female, and fertility treatment was experienced by 92.5% of participants. The participants were observed to be at a variety of stages of fertility treatment. The basic information obtained from participants were age, body mass index, duration of infertility and hospital visits, and lifestyle information such as alcohol intake and smoking. We also asked the participants to report medical conditions that they and their partners had experienced and could obtain information about from their medical history (Supplementary Table 1).

Fig. 1
figure 1

Recruitment of participants. (a): Timing of participating from recruitment. Note that more than 7,000 participants were recruited within the first week after notifying the users of this study and resulted in a final total of more than 13,000 participants. b-c: Location of (b) the participants and (c) the medical facilities where the participants received fertility treatment. Participants were scattered throughout the country, but were especially concentrated in urban areas such as Tokyo.

Full size image
Table 1 Characteristics of participants.
Full size table

Location and medical facilities

To examine the geographical distribution of the participants in Japan, we plotted the locations of the participants and the medical facilities where they underwent fertility treatment. Participants were scattered across the country, but were especially concentrated in urban areas such as Tokyo (Fig. 1b-c). Furthermore, we asked the participants about the features of the medical facilities where they received fertility treatment, and found that the most important factor for selecting the facility was that it was “near home” (Table 2). Transitioning between medical facilities was common; 42.6% of participants changed at least one time, with 4.4% of the participants having transitioned more than 10 times.

Table 2 Medical facilities providing fertility treatment.
Full size table

Fertility testing

We asked participants about the examinations that were performed for assessing their fertility (Supplementary Table 2). We were able to determine the actual status of what percentage of the participants had received blood tests including hormonal and immunological assays, less invasive examinations such as hysterosalpingography, and more invasive examinations such as laparoscopy. We were able to grade the results qualitatively as normal or abnormal, although the patients did not include detailed results of the testing (Table 3). The percentage of abnormalities in laparoscopic examinations was higher than in other tests. Laparoscopic examination may be performed when a specific abnormality is suspected because of its invasiveness; on the other hand, this procedure may also be part of a protocol for fertility evaluation in some institutions. These differences should be taken into account when interpreting the patients’ reports. Some participants provided quantitative values for AMH and semen analysis, suggesting that the participants manage their own data, as well as their partner’s. We found that 65.4% (n = 3,954) of the 6,044 participants have measured AMH, and 941 participants (15.6%) also recalled their own values. Semen analysis data was often reported, with 3,710 participants revealing their partners’ values (Table 4). We also found that the participants often reported on their partners’ fertility data, as well as basic characteristics and lifestyle (Table 1).

Table 3 Participants responses, clinical testing and examination.
Full size table
Table 4 Participants response, semen analysis.
Full size table

Experience of fertility treatment

We then assessed whether critical information gathered during clinical visits can be obtained by the smartphone application. We recruited 9467 participants who had tried only timed intercourse (n = 3533, 37.3%), those who had experienced up to AIH (n = 2437, 25.7%), and those using ART (n = 3343, 35.3%). We also obtained details of the treatment and found that 63.4% of the participants had experienced timed intercourse up to six cycles (Table 5). In Japan, patients are generally advised to proceed to the higher-level treatment when pregnancy has not occurred after six ovulatory cycles by AIH14. Accordingly, 86.9% of them have experienced AIH up to six cycles. Of the participants who experienced oocyte retrieval, 83.8% attempted up to three cycles. For embryo transfer, 76.4% of the participants experienced up to three cycles.

Table 5 Types of fertility treatments that the participants received.
Full size table

Fertility treatment outcomes

We asked the participants about their most recent fertility treatment and the details to determine whether the smartphone application could capture the outcomes. We found that the participants were at a variety of treatment stages (Fig. 2a). We also collected information on ovarian stimulation for corresponding treatments including timed intercourse, AIH, and oocyte retrieval (Fig. 2b-d). Notably, approximately 20% of participants answered “Not sure”, suggesting that the name of treatment protocol might not be precisely understood by the patients. Outcomes of each treatment including timed intercourse, AIH, and embryo transfer were also reported by the participants (Supplementary Table 3).

Fig. 2
figure 2

Recent fertility treatment. (a): Type of recent fertility treatment. Note that the participants were at a variety of treatment stages. b-d: Treatment protocol of ovarian stimulation in (b) timed intercourse, (c) AIH, and (d) oocyte retrieval. Approximately 20% of participants answered “Not sure”, suggesting that the name of treatment protocol might not be precisely understood by the patients. AIH artificial insemination with husband’s semen; OR oocyte retrieval; ET embryo transfer; CC clomiphene citrate; hMG human menopausal gonadotropin; FSH follicle-stimulating hormone; LTZ letrozole; PPOS progestin-primed ovarian stimulation

Full size image

Discussion

The primary finding of this study is that this smartphone application is a promising method for data collection from fertility treatment. In this study, we employed the Luna Luna platform, which is widely used in Japan and has been shown to have some scientific utility13,14,16. In Japan, there is an academic society-based registry for ART, the JSOG registry4,8,9,10,11; however, it does not possess patient-derived data and non-ART data; thus, we conducted this study to establish of a more patient-centered data collection modality for fertility treatment information. In this study, we successfully recruited more than 13,000 participants in one month using a smartphone application platform that is readily available to the general population, including those who have already received fertility treatment. These results indicate that such applications have the potential to rapidly recruit large numbers of participants for medical studies, at least in Japan. Furthermore, it is also valuable for medical professionals to have access to the information included in this current study, which may help them recommend appropriate therapeutic strategies and reasonable timelines. Patient-centered care, good communication and a strong patient–healthcare provider relationship are essential for effective management of patients with infertility17. Theoretically, this platform may also improve patient adherence to treatment schedules, reduce the physical and emotional burden associated with treatment, and decrease the rates of treatment discontinuation18,19. Combined with the information obtained at the hospital, patient-derived information, including their partners’, collected using smartphone applications, can contribute to our understanding of healthcare utilization in Japan. Overall, we found that a vast range of important fertility-related information from couples could be obtained by this smartphone application, suggesting that this modality may be a promising tool for information management in fertility treatment.

Important insights were gained by observing the pattern of responses reported through the application and the characteristics associated with them. Even though most of the participants were women, we were able to collect detailed information regarding their partners. Considering the low involvement of men in fertility treatment in Japan13the gender distribution observed in this study reflects the current situation. In fertility treatment, encouraging more male involvement is acknowledged as a problem that needs to be addressed20. In this study, we showed that information regarding partners can be obtained from the female users. Indeed, it may be ideal to collect the information from the male partners themselves through this platform would be beneficial. In fact, several applications are available for the management of data regarding semen21,22but further improvements are necessary in order to manage a much wider variety of tests and more robust data. Although we need to carefully address the quality of the data, it is beneficial that we were able to obtain the patient-derived male data through our platform. This application may be a feasible option for managing the couple’s information.

We were also able to collect detailed information regarding the participants’ current fertility treatments. To assess the effectiveness of the treatment, information including protocols and outcomes, such as pregnancy and delivery, should be obtained. In this study, we demonstrate the feasibility of using a smartphone application as an information management tool for bringing together important details related to fertility treatment. In addition, our platform may offer further advantages in capturing subjective data or data not traditionally noted, such as the use of add-ons during treatment, patient experiences, and views and predictions.

With the increase in fertility issues experienced by many countries, the findings from this Japanese study may suggest strategies for addressing similar circumstances in other regions of the world as well2. In Japan, fertility care including AIH and ART that are offered by reproductive endocrinologists have been covered by health insurance since April 202223. Until then, medically-assisted reproduction was not covered in principle, but was subsidized. With the transition to universal health insurance coverage, further regulatory mechanisms are needed from the viewpoint of the Japanese government. For example, there are set limits for the number of ART procedures eligible for financial support24; thus, efficient management for treatment of individual patients should be critical. Furthermore, data management is important when assessing the effectiveness of this health insurance coverage. Although this study targeted only people of reproductive age in Japan, it provides important insight into associated with fertility treatment in general. In order to determine the efficacy of public healthcare policies, a smartphone-based data collection system that can reflect real-world circumstances of patients who receive fertility treatment may be worthwhile. Findings from patient-derived data may help in the development of the ideal financial support system, may improve in fertility treatment strategies, and contribute to evidence-based policy-making. Furthermore, long-term public health strategies are essential, and support for infertile patients is provided via various modalities by the government, based on the Act on Assisted Reproductive Technology Offering and the Special Provisions of the Civil Code Related to the Parent-Child Relationship of a Child Born As a Result of the Treatment in Japan23,25. This smartphone application may have the potential to contribute significantly to the well-being of couples and effect health for the long-term.

Establishment of an information management modality using a smartphone application may be acceptable to society since it is in line with modern lifestyles, particularly in developed countries26. In addition, compared to paper-based data collection, digital registration using a smartphone application has potential to offer various benefits, such as time savings and improvements in data quality. The effectiveness of a digital health registry was shown in a recent study27. In another study, a chatbot was found to be a more effective tool rather than standard provision of information28 and this would help data collection using smartphone application when the patients are confused as to how to use the application. In addition, a smartphone application may offer opportunities to enhance the data collection and registration experience for users such as incorporating a chatbot or video recordings.

While a smartphone application approach offered advantages in targeting a broad population, this study had certain limitations. First, smartphone application-based approaches to participant recruitment and data collection is very efficient, but faces challenges with maintaining high follow-up rates13. Without face-to-face contact or instilling a heightened sense of commitment to the study, we experienced a lot of incomplete answers. The burden of manual input may have also affected the low response rate. Development of further sophisticated data gathering strategies such as automated input using optical character recognition system may be warranted. This study may provide important insight for future studies using similar approaches, and issues related to withdrawal may be improved by exploring automated input systems and enhanced digital features. In addition, incentives (e.g., links to financial support systems) for data entry and use can encourage greater participation. Second, the participants may have forgotten and/or lost the results of examination by the time of the participation in this study. Despite the high numbers of qualitative answers collected from the participants, a low proportion of participants provided actual clinical values, even though we requested participants to gather the data in advance. It is not possible to control the quality of data reported by participants, so we recognize that the quality of responses is the primary limitation of the current study. As data management technology develops and more widespread data collection becomes standard, strategies to ensure the accuracy of these data are critical. To ensure accuracy by comparing the agreement rate between data entered by medical institutions and data entered by users of our smartphone application for the same test values, the development of a classifier that can determine whether data are accurate and reject inconsistent values through verification testing may be an option. Creating a platform where data collected at medical institutions is shared directly with patients may also be an alternative, but we need to ensure that test results are not interpreted incorrectly, and security issues should be addressed. Furthermore, it should be acknowledged that approximately 20% of the participants could not answer questions about detailed treatment protocols (Fig. 2b-d). For routine use, there needs to be a user-friendly registry system that enables users to simultaneously input clinically obtained data and treatment information. Third, selection bias should be acknowledged. The participants who responded are essentially self-selected and motivated and may not reflect the general population. For example, the participants in this study may tend to be more proactive, and this is a characteristic that affects the data entry. When combining the data from self-selected patients with the officially collected data, we need to be aware that the data from the individuals who participate may be biased. Less tech-savvy individuals will tend to be excluded from this type of study, and those who are overwhelmed by their infertility diagnosis and may choose not to track their results consistently. Fourth, in the current study, we must address the fact that participants’ reported treatment outcome is only one point, a snapshot in time, rather than continuous data, which is crucial aspect of fertility treatments. Further investigation is warranted to see whether the chronological sequence of treatment outcomes can be collected and can lead to the establishment of novel therapeutic strategies that produce improved outcomes (e.g. a successful pregnancy). Moreover, when more advanced data management technology is developed, it will be possible to link the patient-derived data with the results of examination and treatment in medical facilities, leading to the establishment of users’ health “lifelogs” and precision medicine. Fifth, issues of generalizability to worldwide circumstances should be acknowledged. As a study conducted in Japan, the current results may not be directly generalizable to populations in other countries where social support systems for fertility treatment may be different. However, infertility issues are prevalent worldwide, especially in developed countries, and this study conducted in Japan may provide insight into efforts in other countries as well.

In conclusion, this study demonstrates that a smartphone application modality has potential of being an effective collection and management tool for patient-derived data related to fertility treatment. Given the limitations discussed, the availability of this smartphone application as an assessment tool needs to be carefully considered. Nevertheless, we can build on the insights gained from this study to overcome the issues we encountered, and develop more sophisticated strategies that will benefit broader populations.

Methods

Recruitment of participants

We conducted this study among current application users of MTI Ltd.’s Luna Luna platform13,29. Luna Luna is a smartphone application that assists users in predicting their menstrual cycle and ovulation under normal conditions. Luna Luna Taion note is an application for recording basal body temperature, and many of its users are trying to improve their fertility. Luna Luna Baby is an application that specifically targets users who are pregnant and/or raising a child. The basic functions of the Luna Luna platform are freely available on the Android and iOS platforms (detailed description for these applications is accessible on the web site). The Luna Luna platform also has paid services, and users can freely choose whether or not to use the paid services. This Luna Luna platform is used for health management, and targets the Japanese female population in general13,15,16. As described on the website, the content of the service is supervised by medical professionals.

Users were requested to participate in this study through in-application notifications outlining of the research plan and how data collected from the application would be used. Furthermore, MTI Ltd. published a press release about the study, together with information on how to voluntarily participate. We recruited participants from October through November of 2020. At that time, the number of users who accessed the pages requesting participation in this study was 55,288. This number includes users who simply glanced at or unintentionally accessed the pages. We recruited users of Luna Luna, Luna Luna Taion-note, and Luna Luna Baby who were over 20 years of age and had experienced fertility treatment, and excluded those who did not aspire to become pregnant. In this study, we asked the users to participate voluntarily and without compensation, regardless of whether they used the free or paid services. All participants in this study agreed to be enrolled by selecting the consent button in at least one of the three applications of the Luna Luna platform (Luna Luna, Luna Luna Taion-note, and Luna Luna Baby) which equates to signing the informed consent form before inclusion. Ethics approval for the implementation of the present study was obtained from the ethics committee at the National Center for Child Health and Development of Japan (approval number: 2020 − 175).

Data collection

Data were collected from users of the Luna Luna platform through the applications (Development of Fertility-supporting Smartphone Application Project). Users were requested to answer the following items: (1) basic characteristics and medical history, similar to what is asked during the medical checkup, (2) results of the basic examination for fertility such as hormone levels (including anti-Müllerian hormone (AMH)), hysterosalpingography, semen analysis, etc., and (3) the protocol and outcome of fertility treatment. Table 6 presents the list of information collected in this study. We asked the participants to provide their and their partners’ data at the time of registration with the goal of a single profile for the couple. For the items related to experience with each type of fertility treatment (Table 5), participants were asked to provide data on the fertility treatments they had experienced in their entire reproductive life. Participants reported that they had experienced more than one fertility treatment without the chronological sequence of treatment and might desire more than one child. Because the participants were allowed to skip questions, the number of responses varied depending on the question. In addition, because the data was provided by patients and not by medical professionals, we did not require that they enter numeric values for clinical tests, but instead asked whether results were normal or abnormal.

Table 6 List of information collected in the development of Fertility-supporting smartphone application Project.
Full size table

Data analysis

We recorded the timing of participation of the users and plotted their geographic location and the location of the medical facilities where the participants underwent fertility treatment. To summarize the answers, we reported the actual numbers of responses, as well as proportions for categorical variables and mean with standard deviation for continuous variables. Microsoft Excel, SPSS (IBM), and Prism 9.3.0 software (GraphPad, Inc.) were used for data analysis and Figure creation.