Relationship of Ethinylestradiol/Drospirenone Prescription on Work Productivity and Activity Impairment Among Women With Menstruation-Related Symptoms A Multicenter Prospective Observational Study

Objective: Theaimofthestudyistoexamine changesinworkproductivityand daily activity impairment among women by starting ethinylestradiol (EE)/ drospirenone(DRSP)forperimenstrualsymptoms. Methods: Participantswere women who were newly prescribed EE/DRSP at 25 gynecological clinics in Japan. Eligible participants recorded daily intake of EE/DRSP and the Work Productivity Activity Impairment Questionnaire General Health every 2 weeks for 3 months by smartphone app. A linear mixed-effects model was used to see changes in work productivity impairment and activity impairment relative to baseline. Results: A total of 222 participants were eligible. Work productivity impairment recovered by 20.0% (95% confidence interval, 14.1% – 26.0%) at 1 m and maintained for 2 months. Activity impairment recovered by 20.1% (95% confidence interval, 15.5% – 24.7%) at 1 m and thereafter. Conclusions: Improvements in work productivity and daily activitieswere observed at1 maf- ter EE/DRSP initiation, with a sustained effect thereafter.

R ecently, more women are expected to show leadership in politics and the economy, and gender equality is included in goal 5 of the Sustainable Development Goals. 1 In this context, there is also a rapidly growing interest in the relationship between women's health and work productivity. Modern women with few children have a menstrual cycle that lasts nine-fold more than that in past women with many children. 2 Women often experience menstrual-related discomfort, with 80% experiencing dysmenorrhea and 5.4% experiencing moderate or higher premenstrual syndrome (PMS) in Japan. [3][4][5] The low-dose ethinylestradiol (EE) and progestine controls and relieves menstruation-related symptoms by stopping the natural ovarian hormone cycle, and EE/drospirenone (DRSP) is particularly known to improve symptoms of dysmenorrhea and PMS. [5][6][7] However, only 3% of 15 to 49 years old women in Japan use oral contraceptives (OCs), while 17.8% in Europe and North America. 8,9 This low percentage of OCs use in Japan may be due to its some unique circumstances as follows. Japanese society views OCs somewhat negatively, for example, as women who use OCs may be considered "easy" or casual about having sex, or as using OCs is harmful to women's health although thromboembolic events per 10,000 person-years in OC users was 3.17 (2.81-3.57). [10][11][12] Furthermore, accessibility for women to use OCs is limited because most primary care physicians in Japan do not prescribe the OCs.
Recently, the work productivity loss due to health issue has gained great attention particularly in developed countries. The loss due to menstruation-related presenteeism is estimated to be 491.1 billion yen (approximately 42.2 billion US dollars) per year in Japan. 13 As working women increase, interest in the relationship between women's' health and their work productivity is rising. There is a concern that many women may be forced to work while experiencing menstrual discomfort without using OCs. Neglecting menstruation-related symptoms may reduce women's social performance, and may not only be detrimental to individual women's health, but may also be a social burden. Using OCs is potentially able to improve women's activities with menstruation-related symptoms; however, there has been no report investigating to what extent it affects women's work productivity and activity impairment in reality. This study aimed to analyze the work productivity and activity changes before and after being prescribed OCs for menstruation-related symptoms in reproductive age women.

Study Design
This was a multicenter prospective observational study using a smartphone app designed to assess the changes in work productivity and activity due to menstruation related symptoms after starting EE/DRSP.
It has been established as common practice that EE/DRSPs are immediately prescribed to patients with PMS to alleviate menstrual-related symptoms as a chief complain for the doctor visit. Therefore, in general clinical practice, it is ethically and practically unacceptable to establish a control group without prescribing or the use of placebo. Recognizing the limitations of the validity, a single-arm before/after comparison was used from the perspective of feasibility of the study.

Participants
Women 18 years or older who were newly prescribed EE/DRSP at 25 private clinics between December 2019 and July 2021 were included in this study. The clinics were selected from among those located in medium to large cities in Japan that prescribe OCs for therapeutic purposes. The criteria for initiating a new EE/DRSP were based on the clinical decisions of each physician. Exclusion criteria were women who used ovulation suppressants within 2 months, those with irregular menstruation, those who had comorbid psychiatric disorders, and those without a smartphone or tablet. Participants who were newly started on EE/DRSP received a flyer from their physicians with instructions on how to download the smartphone app, which was developed for this study.

Data Collection
The smartphone app was used for recording data. Participants downloaded the app using a QR code linked to the App store or Google play URL for download; the app could be opened with the individual login ID and password provided in the flyer. The app first displays instructions for study participation, followed by a questionnaire asking whether the participant is willing to participate in the study, their email address, and whether they fulfill the eligibility criteria. Only eligible participants who were authenticated via email could access the symptom log screen.
Participants completed the premenstrual screening tool (PSST) and the verbal rating scale to measure the severity of dysmenorrhea at the start of this study. Every day thereafter for 3 months, participants were asked to record whether they had taken EE/DRSP, the amount of genital bleeding, and daily symptoms, which were then sent to the server in Kyoto University Hospital. The Work Productivity Activity Impairment Questionnaire General Health (WPAI-GH) was completed on the day of EE/DRSP initiation and once every 2 weeks thereafter. All responses were tap selected, and missing values were not transmitted. A reminder email was sent if data were not entered for three consecutive days. Participants who were confirmed to have started filling out the written consent form as well as the application form were each sent a 3000 yen (approximately 25 US dollars) voucher as reimbursement for their participation. We hoped that sending the voucher before starting the study would encourage participants to fill out the application each time they used it.

Questionnaires and Measurements
The PSST consists of 14 questions regarding menstrual-related symptoms, as well as five items on daily life activities with four levels of severity. Outcomes are described according to an algorithm with the following three levels of PMS: less than mild, moderate or higher, and premenstrual dysphoric disorders (PMDDs). 14 The Japanese version of the PSST was developed and back-translated at McMaster University.
The verbal rating scale of dysmenorrhea asks participants to describe usual menstruation as (1) no pain at all, (2) mild pain, (3) moderate pain, and (4) severe pain.
The WPAI-GH is used for participants with a paid job to assess the actual hours worked in the past 7 days, time missed due to health reasons and that due to other reasons, and how much work performance was affected by the health reasons. For questions relating to daily activities, the WPAI-GH is used for all participants including those without a paid job. 15 Participants were informed that general health reasons include menstrual-related symptoms. The WPAI-GH calculates (1) absenteeism (work time missed due to general health problems/time that should have been worked), (2) presenteeism (how much work performance was affected by the health problems), (3) total work productivity impairment due to general health issues (absenteeism + ((1 − absenteeism) Â presenteeism)), and (4) activity impairment (how much daily activity performance was affected by the health problems). For participants with a paid job, absenteeism, presenteeism, and total work productivity impairment can be calculated regardless of whether they work on a full-time or part-time basis. Total activity impairment can be calculated for all participants including those who do not have a paid job. All results are presented as percentages, with higher numbers indicating greater losses.

Statistical Analysis
A two-tailed paired t test was performed for activity impairment, absenteeism, presenteeism, and total work productivity impairment at baseline and 1 month of EE/DRSP use, respectively. Main outcomes were changes in activity impairment, absenteeism, presenteeism, and total work productivity impairment relative to baseline (ie, before prescription) at 1, 2, and 3 months after EE/DRSP initiation, with 95% confidence intervals (CIs). A linear mixed-effects model with person-specific random intercepts was used to evaluate changes in activity impairment, absenteeism, presenteeism, and total work productivity impairment, adjusting for age, degree of dysmenorrhea, and PMS awareness. Three models were constructed to compare between baseline and 1, 2, and 3 mo, respectively. Stata/SE 15.1 for Mac (Texas, USA) was used for statistical analysis.

RESULTS
A total of 222 participants were eligible (median age, 26 years; range, 23-31 years). Of these, 95 participants had moderate or severe PMS, 74 had PMDD as assessed by the PSST, and 115 had "severe" menstrual pain (Table 1). Those with more than moderate PMS, and those with severe dysmenorrhea, had more than 50% activity impairment before initiating EE/DRSP. Among 170 women with a paid job, 49 who had PMDD and 86 who had severe dysmenorrhea showed 49.3% and 46.4% total work productivity impairment at baseline, respectively.
Roughly 20% of participants dropped out within 1 month after EE/DRSP initiation ( Table 2); compared with those who continued on EE/DRSP, there were no significant differences in age, severity of PMS, and dysmenorrhea at baseline. Roughly half of participants continued the study for 3 months. Comparing baseline with 1 month after the EE/DRSP initiation, there were statistically significant changes in activity impairments, presenteeism, and total work productivity, but not in absenteeism.
Longitudinal analysis using a covariate-adjusted liner mixed model revealed a 20.0% (95% CI, 14.1%-26.0%) recovery of total work productivity impairment at 1 mo, which maintained for 2 months (Fig. 1). In addition, activity impairment recovered by 20.1% (95% CI, 15.5%-24.7%) at 1 mo and thereafter. Absenteeism was not high at baseline, and the amount of recovery after EE/DRSP initiation was small. The recovery of total work productivity impairment was largely related to the recovery of presenteeism.

DISCUSSION
This multicenter prospective observational study is the first to evaluate changes in work productivity impairment and activity impairment among women who started on EE/DRSP. At baseline, their work productivity and daily activities were greatly impaired; 1 month after EE/DRSP initiation, improvements were observed, which persisted over the following 2 months.
Total work productivity impairment at baseline was as high as 42.5% among our participants. A previous study in Japan reported a 46% loss of work productivity according to the WPAI scale among individuals with diagnosed depression, 16 which is of comparable magnitude to that observed in the present study. In addition, the finding that moderate or higher PMS affects work productivity is consistent with previous reports. [17][18][19] The social burden of depression is widely recognized, and several public measures have been taken in Japan. Although not as widely recognized, menstruation has also been suggested to cause a significant burden in both high-and low-middle income countries. 20,21 Appropriate social measures against menstruation-related symptoms are thus warranted.
In the present study, work productivity impairment was largely attributed to presenteeism, that is, one of two factors that contribute to work productivity other than absenteeism. Absenteeism is defined as nonpresence at work due to health condition, whereas presenteeism refers to lower performance at work due to a health condition even though the employee is physically present at work. When we divided total work impairment into absenteeism and presenteeism, absenteeism was low at baseline and did not improve after EE/DRSP initiation, like as a flooring effect. Presenteeism, on the other hand, was high at baseline and improved significantly after EE/DRSP initiation. Menstrual-related symptoms may have been overlooked because of the low absenteeism associated with it, which may explain why little social action has been taken. Presumably, women try to minimize the negative influence of menstrual-related symptoms on their employment and only see a doctor when things get really hard, as they are afraid that they might have to take a sick day to visit a hospital. In this context, improving access to OCs may have a positive impact at both individual and societal levels.
The present study also found that women with menstrual-related symptoms have greater impairment in their daily life activities. Women  Activity impairment = degree health affected regular activities. Absenteeism = hours missed due to health problems / (hours missed due to health problems + hours actually worked). Presenteeism = degree health affected productivity while working. Total work productivity impairment = absenteeism +(1 − absenteeism) x presenteeism. *P value: tested for difference between baseline and 1 m.
are engaged in more unpaid work (eg, housework) than men, and this tendency is more prominent in Japan than in Western countries. According to a 2016 survey, 22  This study has several limitations. First, this study was conducted in a real practice setting without control group. It was neither feasible nor ethical to have a group without prescribing OCs because effectiveness of OCs against menstruation-related symptoms has been established and widely used in real clinical practices. We cannot deny the regression to the mean and natural change in the result due to the lack of a control group. Because all participants were treated by physicians, some may have received analgesics, daily life modifications, and counseling (eg, cognitive-behavioral therapy), leading to an overestimation of the association with OCs. In addition, because the baseline was set as of participants' first visit to the clinics, they may have been more symptomatic than usual, resulting in a possible overestimation of the effects of OCs. Second, 20% of participants dropped out within 1 month during follow-up. However, this was relatively lower than those of previous studies that 30% to 75% dropped out among women recoding daily PMS symptoms over a 4-to 6-week period. 23,24 A possible reason for dropout is the burden of symptom recording rather than discontinuation of medications. Only five cases of medication switch and no case of severe adverse effect of EE/DRSP such as thromboembolism were reported, although we could not determine whether those who dropped out continued on EE/DRSPs or not. Because there were no statistical differences in background characteristics such as age, severity of PMS, and severity of dysmenorrhea in the dropout group compared with those who continued with the study, the impact of missing data due to dropout is likely to be limited. Third is concerning generalizability. Because the participants were women who visited the participating clinics, which were not specialized but conscious and general gynecological clinics, the findings might be applied for the other clinics similar to the present setting. In addition, the changes shown in this study are for the participants as a whole; the generalizability beyond these settings still needs careful considerations, due to individual differences in changes in work productivity and activity after the start of the EE/DRSP.

CONCLUSIONS
When women visit a gynecological clinic due to her menstrualrelated symptoms, their work productivity is considerably impaired.
Work productivity and daily activities may improve rapidly after initiating EE/DRSP. The use of OCs for symptom relief may benefit both women and society.