Real world data on symptomology and diagnostic approaches of 27,840 women living with endometriosis

Endometriosis is a chronic disease that requires a suitable, lifelong treatment. To our knowledge, the Visanne Post-approval Observational Study (VIPOS) is to date the largest real-world, non-interventional study investigating hormonal management of endometriosis. We describe women’s experiences of endometriosis in the real world by considering their symptoms and the diagnostic process in their healthcare setting. Overall, 27,840 women were enrolled from six European countries via networks of gynecologists or specialized centers. Of these, 87.8% of women were diagnosed based on clinical symptoms; the greatest and lowest proportions of women were in Russia (94.1%) and Germany (61.9%), respectively. Most women (82.8%) experienced at least one of the triad of endometriosis-associated pain symptoms: pelvic pain, pain after/during sexual intercourse, and painful menstrual periods. The most frequently reported endometriosis-associated symptoms were painful periods (61.8%), heavy/irregular bleeding (50.8%), and pelvic pain (37.2%). Women reported that endometriosis impacted their mood; 55.6% reported feeling “down”, depressed, or hopeless, and 53.2% reported feeling like a failure or having let down family/friends. VIPOS broadens our understanding of endometriosis based on real-world data by exploring the heterogeneity of symptoms women with endometriosis experience and the differences in diagnostic approaches between European countries. Trial registration: ClinicalTrials.gov, NCT01266421; registered 24 December 2010. Registered in the European Union electronic Register of Post-Authorisation Studies as number 1613.


Results
In total, 27,840 women were enrolled through 1012 gynecologists or specialized endometriosis centers across six European countries (Germany, Switzerland, Russia, Poland, Ukraine, and Hungary). Women could be receiving one of the approved or non-approved (off-label) treatments for endometriosis. Study population. The mean overall age at study enrollment was 32.9 years (standard deviation [SD] 9.0), with the lowest age reported by women in Hungary (27.6 [7.6]) and the highest reported by women in Russia (36.1 [8.1]) and Ukraine (37.0 [7.9]). The mean overall body mass index was 23.5 kg/m 2 (SD 4.3 kg/m 2 ), and individual country-specific values were similar. Baseline characteristics and demographics of the study population are shown in Table 1.

Diagnosis of endometriosis.
Overall, endometriosis was diagnosed in 87.8% of women based on clinical symptoms (Fig. 1). Diagnosis based on clinical symptoms was most frequently reported in Russia (94.1%), followed by Hungary (88.0%). In contrast, surgical-based diagnosis was most common in Germany (38.1%) and Poland (34.5%). Overall, 69.6% of women were reported to have received diagnosis of endometriosis within the first year after the occurrence of their first endometriosis symptoms.
Medical treatments. Of the approved treatments prescribed, 11.7% of women received dienogest, 9.3% of women received GnRH agonists, and 3.2% of women received danazol. Non-approved medications for the treatment of endometriosis were prescribed to 62.9% of women receiving combined hormonal contraceptives, 12.4% of women receiving other progestins, and 0.5% of women receiving other non-approved medications for the treatment of endometriosis.
Disease symptoms and impact on mood before treatment. Overall,82.9% of women reported experiencing at least one of the triad of pain symptoms typical of endometriosis (pelvic pain, pain after/during sexual intercourse, and painful menstrual period) via questionnaire at baseline. This was reported by more than 68% of women in each country investigated; the highest percentage was reported by women in Poland (91.0%) and the lowest by women in Hungary (68.8%) (Fig. 2). Only a small proportion of women (8.5%) reported expe- www.nature.com/scientificreports/ riencing all three pain symptoms typical of endometriosis; the highest percentage was reported by women in Ukraine (25.1%) and the lowest by women in Hungary (2.2%). Before treatment, 39.4%, 48.3%, and 9.9% of women reported their endometriosis-related pain as mild (0-3), moderate (4-7), and severe (8-10) over the preceding 4 weeks, respectively, using a numeric pain scale of 0 (no pain) to 10 (unbearable pain) (Fig. 3). Pain rated as mild was most commonly reported by women in Hungary (60.9%), as moderate in Russia (58.5%), and as severe in Poland (29.0%).
Women reported on their mood at baseline via questionnaire, and how they felt endometriosis (and treatment, if received before study enrollment) impacted them by using a scale to indicate how often they experienced a given mood or feeling. Ratings were: 0 (never), 1 (rarely), 2 (sometimes), 3 (often), and 4 (always) ( Table 3). Overall, 55.6% of women reported feeling down, depressed, or hopeless (scale values ≥ 1), with the highest proportion in Poland (69.9%) and the lowest in Hungary (49.9%). In total, 53.2% reported feeling like a failure and having let down family (scale values ≥ 1), with the highest proportion in Russia (59.3%) and the lowest in Germany (38.1%).
A large proportion of women (91.4%) reported feeling happy or optimistic about the future (scale values ≥ 1) via questionnaire at baseline, with the highest proportion from Hungary (94.6%) and lowest from Germany (83.0%).
The given percentages represent all women who reported an impact on mood (scale values > 0).

Discussion
Endometriosis is a recognized chronic condition for which an approved effective, well-tolerated, long-term treatment is still an outstanding need. Real-world studies have investigated key areas in the field of endometriosis, including trends in incidence of endometriosis over time, patient and disease characteristics in clinical practice, clinical management trends, and associated diseases [24][25][26] . Additionally, real-world studies have explored clinical and patient-reported outcomes in routine clinical practice 18,27 .  Table 2. Proportion of patients who experienced endometriosis-associated symptoms at study enrollment. a Patients from Switzerland are not presented here as a subcategory owing to low patient numbers (n = 74). a Data obtained from patients receiving dienogest 2 mg/day, or other approved or non-approved medications for the treatment of endometriosis.  Table 3. Proportion of women reporting impact of endometriosis (and endometriosis treatment, if received) on mood. a Patients from Switzerland are not presented here as a subcategory owing to low patient numbers (n = 74). a Included women reporting scale values of ≥ 1 to indicate how often they experienced a given mood or feeling, with 0 (never), 1 (rarely), 2 (sometimes), 3 (often), and 4 (always). www.nature.com/scientificreports/ VIPOS was a non-interventional, observational study conducted between 2010 and 2018 that aimed to broaden the understanding of endometriosis and its management across Europe by describing the real-world user populations of hormonal endometriosis treatments. Of note, women self-reported medical and gynecologic information, thereby providing valuable country-specific and long-term patient-reported data. Here, we explored the range of symptoms that women with endometriosis experience and the diagnostic process they undergo in their local healthcare setting.

Mood, % of patients
Diagnostic delays of 6-10 years are commonly reported for endometriosis [28][29][30] . Delayed referral by healthcare practitioners is a key cause, possibly resulting from confounding symptoms, misdiagnosis, trivialization of women's experiences, and normalization of symptoms 28,[30][31][32] . Real-world studies have also explored the roles of specialist access, cultural influences, socioeconomic status, misdiagnosis, and inappropriate screening as contributory factors in this delay 6,25,31,33 , whereas a lack of disease awareness is also considered a barrier to prioritizing timely diagnosis, especially in young women 32 . In contrast to the published literature, most women enrolled in VIPOS received their diagnosis of endometriosis within 1 year of the appearance of first symptoms. Although the reasons for this remain unclear, most women included in VIPOS were already receiving specialty care, representing a selection bias introduced by the study design. Diagnostic delays caused by lapses of time between the occurrence of first symptoms and referral to specialist care may not have been captured here, which could have contributed partly to the earlier diagnoses observed. Given that our findings are in contrast to those of previous reports on the typical length of diagnostic delays in endometriosis, additional confirmatory investigations are necessary to revise this widely acknowledged phenomenon.
The aim within the healthcare community should certainly be to raise awareness and knowledge of endometriosis among general medical practitioners (including general practitioners, GPs), facilitate early recognition of endometriosis, and address barriers leading to diagnostic delay. For example, in countries where GPs provide basic gynecologic care and are responsible for referral to specialist medical care, GPs may benefit from targeted education. In a publication 32 involving semistructured focus groups of 43 GPs in the Netherlands, many GPs reported limitations in their knowledge about endometriosis and the endometriosis training they received. Furthermore, almost all GPs considered endometriosis to be a rare disease, and a few expressed uncertainty as to where to find appropriate literature 32 . Greater collaborations between gynecologists and other healthcare professionals is therefore important and should include development of joint guidelines on indications, empirical treatment, referrals, and secondary diagnosis of endometriosis 32 .
In VIPOS, most diagnoses of endometriosis were based on the presentation of clinical symptoms than by surgical means. This indicates appropriate uptake of national guidelines in countries that recommend the use of medical diagnosis, and a shift toward clinical diagnosis in countries such as Germany and Poland, where national guidelines recommend a surgical approach 4,34,35 . International guidelines outlining empirical or initial treatment of symptoms before surgical approaches have also contributed to the observed trend toward medical diagnosis 3,4 , alongside the increased availability of imaging modalities (ultrasound and magnetic resonance imaging). Despite this, other guidelines and literature suggest that clinical diagnosis is still inconsistently applied in medical practice 36 . Practitioners should feel empowered to make clinical diagnoses of endometriosis early and without an invasive procedure, as there is potential for early diagnosis and treatment to relieve women's pain, avoid central sensitization and pain persistence, prevent infertility, and change the trajectory of women's lives 36 . Indeed, endometriosis can be suspected without surgical exploration, and medical treatment can be safely prescribed without histologic validation of the disease 37 , and can confirm diagnosis when effective. Therefore, a combination of patient interviews and clinical examinations should be sufficient to enable practitioners to identify women suspected of having endometriosis and who may require imaging assessment 37 , with the effect of appropriate medical therapy also assisting in this identification.
Changes in country-specific behaviors, culture, and reimbursement challenges may also influence this shift in diagnostic approaches. For example, it has been suggested that cultural influences likely affect the clinical presentation of endometriosis 38 and can lead to differences in conceptualization and reporting of pain, and health-seeking behavior. Furthermore, differences in healthcare experiences, expectations, and efficiencies can impact women's reporting of symptoms 39 . Race and ethnic background can also affect the provision of healthcare at all levels, which can thereby influence access to care, specialist referral, diagnosis, and treatment 32,38 . Future investigations into the disparity in country-specific rates of clinical diagnosis of endometriosis observed in VIPOS would be interesting, particularly focusing on whether potential barriers and facilitators to the timely diagnosis of endometriosis, such as patient and physician behaviors, can be identified.
Women in VIPOS most commonly reported heavy/irregular bleeding, painful periods, and pelvic pain, followed by tiredness/weakness. In addition, most women experienced at least one of three pain symptoms typically associated with endometriosis (pelvic pain, pain after/during sexual intercourse, and painful menstrual bleeding), and more than 50% of women overall reported an impact on mood due to endometriosis. These observations add to the existing literature on the diverse symptomology of endometriosis and the impact of endometriosis on quality of life [40][41][42] .
The endometriosis-associated pain symptoms reported in VIPOS are likely already well established in routine clinical questioning of women suspected with endometriosis. Unsurprisingly, the most commonly reported symptoms leading to a diagnosis of endometriosis may be dysmenorrhea and pelvic pain 42 , which form part of the classical triad of endometriosis-associated symptoms on which clinical questioning is traditionally based. Additionally, 50.8% of women experienced heavy/irregular bleeding and 27.2% experienced tiredness/weakness, which has recently been identified as an underestimated symptom of endometriosis 43 . Surprisingly, only 8.5% of women in VIPOS experienced all symptoms of this triad, which is lower than reported elsewhere 42 . These striking reports of symptoms challenge the importance of the symptom triad in the diagnostic questioning of endometriosis. We propose the need for greater awareness of the various and less "traditional" symptoms of endometriosis to ensure a more comprehensive patient-doctor dialogue that may encourage women to seek www.nature.com/scientificreports/ healthcare and empower them to vocalize their experiences. Albeit inconsistent, there is a correlation between pain severity and endometriosis disease severity [44][45][46] ; however, the severity of endometriosis in women enrolled in VIPOS could not be investigated, as disease staging data was not collected. VIPOS provides insights into the current treatment choices for endometriosis made by physicians in the real world. Women were most commonly prescribed off-label use of combined hormonal contraceptives (62.9%), such as OCs; although clinical evidence supporting the efficacy of OCs is limited, they are frequently used to treat endometriosis due to their general safety, affordability, and the comfort and familiarity associated with both their use and prescription 15 . Dienogest was prescribed to 11.7% of the women; it is one of few approved medical therapies indicated for endometriosis, and has been studied extensively, demonstrating favorable safety and efficacy in women with endometriosis by improving dysmenorrhea, premenstrual pain, dyspareunia, and pelvic pain, and decreasing the duration of menstrual bleeding and size of endometriomas [47][48][49][50][51][52] . Other prescribed treatments included GnRH agonists and danazol, although associated side effects restrict their use 4,51 .
From our understanding, VIPOS is the largest real-world non-interventional study of the medical treatment of women with endometriosis, reporting data from 27,840 women with endometriosis across European countries. Detailed questionnaires provided comprehensive assessment of each woman's gynecologic, medical, and endometriosis history, in addition to their mood, lifestyle, and education. VIPOS raise awareness of how frequently women experience the different symptoms of endometriosis, thereby alluding to the need for more informed clinical questioning considering "non-traditional" symptoms that women may experience. VIPOS also highlights overall adherence to diagnosis by clinical means rather than by surgical interventions. Limitations of VIPOS include the country-specific variability in the number of women and treatment centers enrolled. Given the observational study design of VIPOS, results must be carefully interpreted and consideration given to the fact that residual confounding or bias cannot be entirely eliminated, which may lead to limitations in inferring causation 18,53-55 .

Conclusions
VIPOS represents a real-world account of the management of endometriosis in specific European healthcare settings through its reporting on the wide heterogeneity of symptoms and the differences in diagnostic approaches that women with endometriosis experience. In particular, the data presented from 27,840 women in this study suggest that clinical diagnosis forms a key element of the medical management of endometriosis in most healthcare settings studied. This supports the need for an open dialogue between physicians and their patients, which should include comprehensive patient interviews designed to capture the spectrum of diverse symptoms that women with endometriosis may be afflicted with. While further study is required to strengthen these findings, the data from VIPOS are supportive of the recent paradigm shift toward clinical diagnosis in the field of endometriosis.

Methods
The methodology of VIPOS has previously been reported 54,[56][57][58][59] . Briefly, VIPOS was a prospective, observational, long-term cohort study conducted between 2010 and 2018, enrolling women using various forms of hormonal treatment for endometriosis. Women were enrolled in Germany, Switzerland, Russia, Poland, Ukraine, and Hungary. The number of women enrolled in Switzerland (n = 74) was insufficient for separate country analysis, and data obtained from these women have been included only as part of the overall data presented here. In addition, a small proportion of women (n = 22) were prescribed a treatment for endometriosis that consisted of more than one hormonal medication concomitantly; they also were included in the overall data set for the description of baseline characteristics but are not described in detail.
The primary objectives of VIPOS were to evaluate the safety of dienogest 2 mg/day compared with other hormonal treatments in the routine clinical practice setting, with a focus on the occurrence of anemia induced by cyclical bleeding disturbances and requiring medical treatment, clinically relevant depression (new or worsening), and drug discontinuation due to treatment failure, as defined by loss/lack of efficacy or an adverse drug reaction. The primary results will be published elsewhere.
Women initiating a new hormonal therapy regimen for endometriosis and willing to participate were eligible for inclusion, while those not willing to participate in a long-term follow-up or with language barriers were not eligible. Hormonal treatments could be approved or not approved (off-label) for the treatment of endometriosis; dienogest 2 mg/day, GnRH analogs, danazol, combined hormonal contraceptives (most commonly combined OCs), other progestins (e.g., dydrogesterone, progestin-only pills, levonorgestrel-containing intrauterine system), and other treatments (hormonal replacement therapy, unknown allocation of concomitant treatments).
To ensure that study participation was not considered necessary for treatment, physicians discussed the study only after the women received their treatment prescription. Enrolled women received a questionnaire at baseline relating to their personal health status and potential risk factors. They self-reported data on their medical and gynecologic history, endometriosis-related symptoms, diagnosis timelines, disease impact on mood, treatment, and lifestyle and educational status. Each woman's physician completed a subsection of the baseline questionnaire, which included details on the type of endometriosis diagnosis (i.e., based on clinical symptoms or surgical confirmation [direct visualization at laparoscopy and/or histologic assessment of excised endometriosis], with any additional tests, including ultrasound and magnetic resonance imaging) and their recommended prescription and surgical history for disease management. Women were followed up at pre-specified time points, which were 6, 12, 24, and 36 months after baseline and, depending on the date of enrollment, 48, 60, 72, and up to 84 months after baseline. In order to minimize the loss-to-follow-up rate, a multi-faceted, four-level process was utilized: this process has been described and implemented in other studies 54,[56][57][58][59] . Level one activities included mailing followup questionnaires and sending reminders (letters or phone calls), when necessary. If no response was received, www.nature.com/scientificreports/ level two activities were implemented, which involved multiple attempts to phone women and/or named contacts provided during the study. In parallel, searches in national and international directories and social networks were carried out (level three activities). Finally, formal address inquiries were conducted, where possible (level four activities), to enable interviewers to visit women's homes or send registered letters with confirmation of receipt. The ethics information for the VIPOS study has been published previously 55 . The study was approved by one independent ethics committee/institutional review board at each country, where required. In Germany, the Ethics Committee of the Berlin Medical Association approved the study, and in Hungary this was done by the Scientific and Research Ethics Committee of the Medical Research Council. In Switzerland, only one large endometriosis center (Inselspital Bern) took part in the study, and a positive vote from Swissmedic was obtained for this hospital. No ethical approval for non-interventional, observational studies was required by law in Poland, Russia, and Ukraine. Each woman provided written informed consent before participating in the study. The study was conducted in accordance with the Guidelines for Good Pharmacoepidemiology Practices issued by the International Society for Pharmacoepidemiology (2008) Informed consent. Informed consent was obtained from all participants in the study.

Data availability
The datasets generated during and/or analyzed during the current study are available from the corresponding author on reasonable request.