DOI: https://doi.org/10.21203/rs.3.rs-2049133/v1
Background: The EULAR recommendation requires that the disease activity of systemic lupus erythematosus before pregnancy be stable because pregnancy complications and disease flares increase if pregnancy occurs while systemic lupus erythematosus activity is high. However, some patients have ongoing serological activity even after treatment. Herein, we aimed to investigate how physicians make decisions on the acceptability of pregnancy when patients with systemic lupus erythematosus only have serological activity.
Methods: A questionnaire was administered online to physicians from December 2020 to January 2021. It included the characteristics of physicians, facilities, and the allowance for pregnancies for patients with systemic lupus erythematosus using vignette scenarios.
Results: The questionnaire was distributed to 4,946 physicians, and 9.4% of physicians responded. The median age of respondents was 46 (range: 38–54) years, and 85% were rheumatologists. Pregnancy allowance was significantly affected by duration of the stable period, and mild or high serological activity, respectively (duration: proportion difference, 11.8 percentage points [p.p.]; p < 0.001; mild: proportion difference, -25.8 p.p; p < 0.001; high: -65.6 p.p.; p < 0.001). Even though patients had a high level of serological activity, 20.5% of physicians allowed pregnancy if there were no clinical symptoms for six months.
Conclusions: Serological activity had a significant effect on physician judgment on the acceptability of pregnancy. Conversely, some physicians allowed patients who only had serological activity to become pregnant. Further observational studies are required to clarify the prognosis of such patients.
We investigated the effect of SLE serological activity on physician acceptance of pregnancy. Serological activity alone was found to negatively affect pregnancy acceptance. Conversely, we also found that some doctors approve pregnancy only with serological activity. It is necessary to collect further information on the prognosis of pregnancy in the future.
Systemic lupus erythematosus (SLE) is a systemic autoimmune disease that presents with chronic inflammation in multiple organs, including the skin, joints, kidneys, lungs, blood, and central nervous system. Because SLE primarily affects women of childbearing age, pregnancy and its outcomes are significant concerns for most SLE patients. Patients with SLE are still at high risk of adverse pregnancy outcomes (APOs) compared with the general population [1–3], and the rate of disease flare during pregnancy and one year postpartum is from 30 to 50% [4–6]. Several studies have identified active diseases at conception, hypocomplementemia, anti-ds-DNA antibodies, history of lupus nephritis, and antiphospholipid antibodies (aPL) as risk factors for APOs [7–13]. Therefore, patients with SLE are advised to consider pregnancy during inactive or stable disease periods, the so-called “planned pregnancy” [14]. The EULAR recommendation requires stabilization of disease activity for 6–12 months before pregnancy [15].
While normal levels of complementary and ds-DNA antibodies are desirable before conception because of the association of hypocomplementation with poor fetus prognosis [16], we encounter patients with only persistent serological activity but no clinical activity in clinical practice. Previous studies have shown that approximately 6–12% of SLE patients exhibit high serologic activity, anti-ds-DNA antibodies, and/or hypocomplementemia without clinical activity [17, 18]; we may encounter difficulties regarding pregnancy allowances in patients who only have serological activity. Herein, we aimed to investigate physician judgment on the acceptability of pregnancy in patients with SLE showing only serological activity by a vignette survey study.
This study involved an online cross-sectional survey conducted from 14 December 2020 to 17 January 2021 according to the Checklist for Reporting Results of Internet E-Surveys (CHERRIES) guidelines [19]. Our study sample consisted of physicians from the Japan College of Rheumatology (JCR). Members of JCR are practicing physicians, research scientists, and academic scholars from all major scientific and hospital institutions in Japan, and all members are subscribed to the mailing list of JCR. Since this study aimed to conduct an exhaustive survey, a questionnaire was administered to all physicians (n = 4,946) affiliated with JCR. The data were obtained from anonymous online cloud-based questionnaire development software (Survey Monkey; Momentive Inc., San Mateo, CA, USA; https://www.surveymonkey.com). We sent the voluntary questionnaire survey URL to the JCR mailing list.
We developed a six-page survey questionnaire examining permission of pregnancies with SLE patients. The survey booklet and cover letter highlighted to physicians that, while treating pregnancy complicated by collagen disease is critical, it is unknown how well clinicians are aware of the therapy and how it is conducted. We conducted a scenario study (vignette survey) to determine whether physicians focus on age, stable period, or ds-DNA value, when making a judgment to allow pregnancy. The vignettes varied in age (28 or 35 years), duration of stable disease activity (3 or 6 months), and anti-ds-DNA titer (0 IU/mL or 30 IU/mL or 120 IU/mL). For the latter two variables, the EULAR recommendations [15] were used as the basis for selection. Based on these three characteristics of patients, we created 12 (2 × 2 × 3) different vignettes. We asked physicians whether they would allow or disallow pregnancy in each scenario. An example scenario is as follows:
A female, aged 28, was diagnosed with systemic lupus erythematosus at age 20 due to rash, arthritis, leukopenia (2,300 /µL), and anti-ds-DNA antibody 200 IU/mL (baseline: <10.0 IU/mL). After steroid intervention, the anti-ds-DNA antibody result was negative and serum complement titer (CH50) was normal. For the past six months, there has been no skin rash, arthritis, or hematopenia, but anti-ds-DNA antibody (30 IU/mL) and serum complement titer (CH50) has remained around 24 IU/mL. She is currently only taking prednisolone 5 mg/day and is not taking any immunosuppressants or other medications contraindicated for use during pregnancy. She and her family wish to have a baby.
We asked participants if they had completed all the questions at the end to increase the completion rate. The questionnaire also asked about the characteristics of the physicians: age; sex; specialty; availability of rheumatology certification; work area; duration of medical examination of rheumatology; number of collagen disease patients managed (per year); and number of collagen disease patients who want to become pregnant (controlled, per year). The questionnaire also asked about facilities (the availability of the departments of obstetrics and gynecology, pediatrics, and neonatal intensive care units).
To develop clinically acceptable scenarios, we had multiple steps. We conducted a meeting regarding the survey with four rheumatologists specializing in pregnancy-related issues, refined the questionnaire, and increased its clarity based on the multiple discussions. Then, we conducted an assessment of the validity of the vignette scenario with two external experts and modified the questions to have clinical and scenario validity. Finally, we performed a pilot test for 10 physicians by sending them the surveys to see if the questions were difficult to answer and if the number was adequate; we, again, refined the questionnaire. The completed questionnaire was uploaded to Survey Monkey.
The physicians (n = 4,946) surveyed were randomly assigned six of the 12 vignettes. To increase the response rate, flyers were distributed to encourage input in the middle of the survey period, and reminders were sent out via the mailing list before the deadline. Participants who completed the questionnaire in entirety were provided with materials on pregnancy in patients with collagen diseases as an incentive.
Our analysis included participants who provided answers to all questions. We recorded the time participants needed to fill in a questionnaire and excluded questionnaires that were submitted too quickly. Three minutes were chosen as the cut-off period based on the time required to complete the questionnaire during the pre-test.
Data are expressed as frequencies and percentages for categorical data and medians with interquartile ranges [IQR] for continuous variables. The prevalence of acceptable pregnancies for each scenario is described. As primary analysis, generalized estimating equations (GEE) with an identity link function and the robust variance for binary outcome variables were used to investigate the relationship between determining permission for pregnancy and the scenario patient characteristics (age, period of stable disease, titer of anti-ds-DNA antibody). Next, we investigated the background of physicians who allowed pregnancy in cases with high anti-ds-DNA antibody titers by chi-square test. To prevent missing data, all responses were required in the online survey. Therefore, it was not necessary to address missing values. Statistical significance was defined as a two-sided p-value of < 0.05. The post-hoc analysis of GEE was performed with the Bonferroni correction when group differences were found (p < 0.01). All statistical analyses were conducted using STATA 17.0 (Stata Corp LP, College Station, TX, USA). Data were collected via Survey Monkey and were downloaded and imported to STATA for analysis.
Table 1 shows the baseline characteristics of participants. The median age of the physicians was 46 (interquartile range [IQR]: 2–10). The reported specialties were rheumatology (84.9%), other internal medicine (8%), and pediatrics (5.6%). More than half of the respondents were doctors at hospitals that offer obstetrics and pediatrics. The median number of patients wishing to become pregnant managed by physicians per year was 5 (IQR: 2–10).
| Age, median (IQR) | 46 (38–54) |
|---|---|
| Men, n (%) | 319 (68.9) |
| Years of healthcare experience years, median (IQR) | 16 (10–25) |
| Majority, n (%) | |
| Rheumatology and collagen diseases | 393 (84.9) |
| Other internal medicine | 37 (8.0) |
| Pediatrics | 26 (5.6) |
| Other | 7 (1.5) |
| Certified rheumatologist, n (%) | 396 (85.5) |
| Adjunct obstetrics and gynecology, n (%) | 327 (70.6) |
| Adjunct NICU, n (%) | 278 (60) |
| Adjunct pediatrics, n (%) | 352 (76) |
| Number of patients treated per year, median (%) | 250 (100–500) |
| Number of patients treated wishing to become pregnant, per year, median (IQR) | 5 (2–10) |
| IQR, interquartile range; NICU, neonatal intensive care unit. | |
The specialties were rheumatology (84.9%), other internal medicine (8%), and pediatrics (5.6%). A total of 70.2% of the doctors worked in hospitals that have obstetrics and gynecology.
Table 2 shows the prevalence of pregnancy allowance for each scenario. If serological activity was stable for 6 months, 99.6% and 98.6% of physicians would allow pregnancy in the case of a 28-year-old patient (scenario no. 1) and a 35-year-old patient (scenario no. 7), respectively. If serological activity was high for six months without clinical activity, 29.0% and 20.5% of physicians would allow pregnancy in the case of a 28-year-old patient (scenario no. 5) and a 35-year-old patient (scenario no. 11), respectively. The percentage of acceptable pregnancies decreased as serologic activity increased.
| Scenario No. | Age (y) | Duration (m) | Serological Activity | Allowance Prevalence (%) |
|---|---|---|---|---|
| 1 | 28 | 6 | Normal | 235 (99.6%) |
| 2 | 28 | 3 | Normal | 191 (84.1%) |
| 3 | 28 | 6 | Mild | 169 (71.3%) |
| 4 | 28 | 3 | Mild | 126 (55.8%) |
| 5 | 28 | 6 | High | 70 (29.0%) |
| 6 | 28 | 3 | High | 52 (23.6%) |
| 7 | 35 | 6 | Normal | 219 (98.6%) |
| 8 | 35 | 3 | Normal | 198 (82.2%) |
| 9 | 35 | 6 | Mild | 171 (77.3%) |
| 10 | 35 | 3 | Mild | 153 (63.2%) |
| 11 | 35 | 6 | High | 50 (20.5%) |
| 12 | 35 | 3 | High | 73 (33.3%) |
Most physicians allow pregnancy if the 6-month serologic activity is normal, 99.6% at age 28 and 98.6% at age 35. The percentage of acceptable pregnancies decreased as serologic activity increased.
Table 3 shows vignette patient factors related to pregnancy allowance using GEE analysis. There were no significant differences in case age (proportion difference, 0.28 percentage points [p.p.]; 95% confidence interval [CI], -1.68 to 1.22; p = 0.42). Physicians of patients who had been stable for six months were more tolerant of pregnancy than those of patients who had been stable for three months (proportion difference, + 11.8 p.p.; 95% CI, 8.78 to 14.8; p < 0.001). Pregnancy was not allowed in cases with mild or high serological activity (mild: proportion difference, -25.8 p.p.; 95% CI, -29.5 to -22.0; p < 0.001; high: proportion difference, -65.6 p.p.; 95% CI, -84.0 to -61.6; p < -0.001).
| Proportion difference, percentage points | 95% CI | P-value* | |
|---|---|---|---|
| Age (28 y vs 35 y) | -0.23 | -1.68 to + 1.22 | 0.42 |
| Stability period (3 m vs 6 m) | + 11.8 | + 8.78 to + 14.8 | < 0.001 |
| Serological activity | |||
| Normal vs. Mild | -25.8 | -29.5 to -22.0 | < 0.001 |
| Normal vs. High | -65.6 | -84.0 to -61.6 | < 0.001 |
| CI, confidence interval | |||
| *Post-hoc analysis of the background of physicians who allowed pregnancy in cases of high anti-ds-DNA antibody titer was performed with the Bonferroni correction when group differences were found (p < 0.01). There were no significant differences in patient age. The length of time that clinical symptoms were stable and serological activity significantly affected the decision of pregnancy allowance. | |||
Table 4 shows physician factors related to pregnancy allowance in patients with high serological activity. Female physicians were significantly less tolerant of pregnancy in high serological activity (12% vs. 37.5%, p < 0.001). There were no significant differences in specialty status or clinical experience.
| Allow (n = 92) | Not allow (n = 309) | P-value | |
|---|---|---|---|
| Female, n (%) | 11 (12) | 116 (37.5) | < 0.0001 |
| Rheumatologist, n (%) | 83 (90.2) | 260 (84.8) | 0.15 |
| Change in practice due to revision of the attached document, n (%) | 64 (69.6) | 206 (66.7) | 0.89 |
| Number of patients treated wishing to become pregnant per year, median | 4 | 5 | 0.18 |
| Number of patients treated per year, median | 250 | 300 | 0.60 |
Female physicians are significantly more cautious about pregnancy than male physicians when patients have a serologically high activity. There were no significant differences in specialty status or clinical experience.
We clarified the current situation of how physicians make decisions on allowing patients to become pregnant in cases where they had serological activity. We found that serological activity negatively impacted physician pregnancy allowance. However, we also found that approximately 19.2% of physicians allowed pregnancy even with high serological activity. To our knowledge, this study was the first to investigate physician pregnancy allowance focusing on serological activity. Serological activity should be an essential factor for physicians in deciding their allowance for pregnancy because several studies reveal that disease activity is a risk factor for APOs and SLE flare related to pregnancy [20, 21]. Our results revealed that serological activity significantly reduces the chances of physician allowance for pregnancy. This result indicates that physicians should carefully assess serologic activity when allowing patients to become pregnant, and follow EULAR recommendations in their clinical practice.
It was also clear that some physicians would allow pregnancy even if serological activity was present. We considered this result to be different from the evidence-practice gap because there was no significant difference in the annual number of patients who wanted to have a baby between physicians who allowed pregnancy and those who did not. In the present vignette scenario, the patient had no history of lupus nephritis or antiphospholipid antibodies, which have been reported as factors associated with a poor prognosis [22, 23]. Therefore, it was possible that the physician would allow the pregnancy based on clinical expertise and preferences of patients, which are components of clinical decision-making [24].
There were several strengths of this study. First, this study was the first to investigate physician pregnancy allowance focusing on serological activity. We determined how physicians allow pregnancy when SLE patients who have serological activity wish to become pregnant. Second, we used a vignette-based study which is useful to assess treatment quality and practice variability. This is frequently used to evaluate clinical care quality and measure variance in practice among nations, health care systems, specializations, and physicians [25, 26]. Vignettes enable comparisons between clinician judgments of pregnancy when the only serological activity remains.
This study had several limitations. First, a selection bias might have resulted from a poor response rate to the survey (9.4%). Considering the number of patients who want to have a baby managed by clinicians per year, and the percentage of doctors who change their practice policy on immunosuppressive drugs during pregnancy due to revisions of package inserts, it is possible that the rheumatologists who agreed to participate were probably those who were most interested in pregnancy with SLE. However, according to various studies, lower response rates do not always imply a higher likelihood of non-response bias [27, 28]. Second, the studies were not patient-based, such as cohort studies. Pham et al. showed that, in evaluating clinical practice using a clinical vignette, a multiple-choice format rather than an open-ended format overestimates physician performance [29]. This theoretical approach is at greater risk of social desirability bias and may thus result in overestimates of guideline adherence and appropriate management. Although the present study was not conducted on patients, we believe that this study is significant because it is a first step toward establishing better guidelines for pregnancy complicated SLE with only serological activity. Third, there was the possibility of duplicate responses, and we did not obtain an IP address. Therefore, we cannot eliminate the possibility of duplicate responses. Still, we believe that the rate of same responses is low because the questionnaire was conducted through the academic society. Finally, our study has limited generalizability to other countries since the survey was conducted only on Japanese physicians. However, the Japanese guidelines for SLE pregnancy [30] also refer to the EULAR recommendations, and we believe that they can be used as a reference in other countries as well.
Since it was found that patients who want to become pregnant with only serological activity are allowed to become pregnant, clarification of the pregnancy prognosis of such patients will lead to better recommendations in the future. We believe that this study provides an opportunity for physicians to become more aware of pregnancy-related issues, which are essential in SLE treatment.
Mild serological activity alone had a significant adverse effect on physician judgment on the acceptability of pregnancy. Conversely, some physicians allow patients who have serological activity with no clinical activity to become pregnant. It is essential to clarify the prognosis of such patients through observational studies in the future.
APO Adverse pregnancy outcomes
GEE Generalized estimating equations
JCR Japan College of Rheumatology
NICU Neonatal intensive care unit
SLE Systemic lupus erythematosus
Ethics approvals and consent to participate
The study was conducted after approval by the Ethics Committee of Showa University (approval number 3316). An explanation of this study was provided at the beginning of the survey, and consent was assumed to have been obtained when the response was returned. All study procedures were performed in accordance with the Declaration of Helsinki and Health Research Involving Human Subjects in Japan.
Availability of data and materials
The data are available from the corresponding author upon reasonable request.
Competing interests
The authors declare that there is no conflict of interest.
Funding
This work was supported by JSPS KAKENHI Grant Number JP21K02113．
Authors’ contributions
Sakiko Isojima: methodology, conceptualization, project administration, writing – original draft, writing – review & editing; Nobuyuki Yajima: methodology, conceptualization, project administration, writing – original draft, writing – review & editing; Ryo Yanai: methodology, conceptualization, project administration, writing – review & editing; Yoko Miura: methodology, conceptualization, project administration, writing – review & editing; Shingo Fukuma: methodology, conceptualization, project administration, supervision, writing – review & editing; Kayoko Kaneko: supervision, writing – review & editing; Kenji Oku: supervision, writing – review & editing; Masakazu Matsushita: supervision, writing – review & editing; Takako Miyamae: supervision, writing – review & editing; Keishi Fujio: supervision, writing – review & editing; Yuko Kaneko: supervision, writing – review & editing; Tsutomu Takeuchi: supervision, writing – review & editing; Yoshiya Tanaka: supervision, writing – review & editing; Takashi Wada: supervision, writing – review & editing; Atsuko Murashima: conceptualization, supervision, writing – review & editing.
Acknowledgments
We would like to express our gratitude to the JCR members who cooperated in filling out the questionnaire and to the doctors (Shogo Toyama, Yu Katayama, Koei Oh, Yoichi Toyoshima, Michihito Sato, Serina Furuto, Ayuko Takatani Tetsuya Nemoto, Hajime Ishikawa, Kazuki Kawamori, Ryutaro Gunji) who gave their valuable opinions in the pre-test.