Fetal–neonatal and maternal outcomes in women with Sjögren syndrome: a population-based registry linkage study

Abstract

Objective

To provide better preconceptional and prenatal counselling to patients with sjögren syndrome (SS).

Methods

In total, 2 100 143 pregnancies between 2004 and 2014 were identified in the Taiwan National Health Insurance database and birth registry. The maternal history of SS was ascertained, and data were compared between pregnant women with and without SS. We assessed the odds ratios and 95% CIs of fetal–neonatal and maternal outcomes.

Results

There were 449 pregnancies in women with SS and 2 099 694 pregnancies in women without SS. The risks of still birth [odds ratio (OR) = 2.14, 95% CI = 1.01, 4.55], low birth weight (<2500 g, OR = 2.53, 95% CI = 1.92, 3.33), small for gestational age (OR = 2.03, 95% CI = 1.57, 2.03) and fetal distress (OR = 1.72, 95% CI = 1.2, 2.45) as well as maternal risks of pulmonary oedema (OR = 11.64, 95% CI = 1.62, 83.48), shock (OR = 6.07, 95% CI = 1.51, 24.3) and respiratory distress (OR = 5.61, 95% CI = 1.39, 22.6) were higher in the SS group than in the non-SS group.

Conclusion

Women with SS have significant risks of adverse fetal–neonatal and maternal outcomes and must undergo prenatal counselling to understand the risks involved before conception.

Introduction

Sjögren syndrome (SS) is an autoimmune disease that can develop alone (primary SS) or in conjunction with an underlying connective tissue disease (secondary SS), most commonly SLE or RA [1]. The spectrum of the clinical presentation of SS extends from dryness of the primary mucosal surfaces to systemic involvement (extraglandular manifestations). Mucosal surface dryness occurs because of secretory gland dysfunction caused by immune-mediated inflammation [2]. The sicca features of SS primarily affect patients’ quality of life, whereas the prognosis of SS is affected by systemic involvement [3]. SS is known to occur predominantly in women, and affected women are likely to experience more complicated pregnancies than women without SS [4, 5].

The effect of autoimmune diseases on pregnancy differs according to the maternal disease, disease activity, severity of organ damage, antibody profile and drug treatment [6]. Available data on pregnancy outcomes in SS are scarce. Only a few studies have evaluated pregnancy and fetal outcomes in patients with SS. Some studies have linked SS to adverse maternal/neonatal outcomes, but their sample sizes tended to be small, with few outcomes examined [4, 7–11]. This prompted us to analyse population-based registry data and perform a review of the literature to examine the effect of SS on pregnancy and fetal outcomes compared with those in the general obstetric population.

Patients and methods

Ethics approval

This study was approved by the Institutional Review Board of Chang Gung Memorial Hospital (104-8080B) Taiwan’s National Health Insurance (NHI) and National Birth Registry. The patients’ data in this study were completely anonymized.

Study design and data sources

The data utilized in this population-based cohort study was acquired from Taiwan’s NHI database that is linked to the National Birth Registry and National Death Registry. Taiwan’s NHI database, established in 1996, contains the health data of 99.5% of the population in Taiwan. Almost all participants of this study were covered by the Taiwan’s NHI; this is considering the fact that patients diagnosed with SS at the hospital should apply for a catastrophic illness certification. Therefore, the number of cases included was considerably high. Patients’ sex, date of birth, place of residence, insurance details, family relationships, dates of inpatient and outpatient visits, medical diagnoses, medical expenditures, prescription details, examinations, operations, procedures and fees incurred are included in the NHI database. Due to privacy concerns and the patient's protection, all beneficiaries are given a unique identification which allows internal linkage of data in the NHI database. This identification was encrypted to ensure confidentiality but the uniqueness of the encrypted identification is retained to ensure valid internal linkage. In addition, all researchers need to access the NHI data using workstations provided by the Ministry of Health and Welfare. Therefore, patient consents were waived as all data are fully de-identified (https://dep.mohw.gov.tw/dos/lp-5146-113.html).

Maternity care in Taiwan is entirely covered by the NHI. It includes antenatal health examinations, fetal ultrasound in each trimester, delivery, and infant and postpartum care. Because maternity care has universal access, there has been a sharp decrease in maternal mortality in Taiwan [12]. The National Birth Registry has been validated for use in epidemiological research [13], and it contains information on live births and stillbirths (>20 weeks old or >500 g in weight) occurring since 2004. The registry includes the demographic details of pregnant women and their spouses, pregnancy data, and data of newborns. The data accuracy of the National Birth Registry has been documented and accordingly, the registry has been confirmed as a valid information source for fetal–neonatal and maternal outcomes [13]. The birth registry accurately reports sex, birth order, gestational age, neonatal abnormalities, Apgar scores and birth weight. A kappa statistic of >0.85 was identified through a concordance between the National Birth Registry and a retrospective chart review for all selected variables. When birth weight and gestational age were treated as categorical variables, the kappa statistics ranged from 0.92 to 0.96 [13].

Study cohort

We excluded pregnancies in women who were aged <15 years or >45 years and those involving multiple births from the analysis. Using the NHI database and linked registries from 1 January 2004 to 31 December 2014, we identified a cohort of 2 100 143 singleton pregnancies that resulted in either live births or stillbirths (a gestational age of >20 weeks or fetal weight of >500 g). A total of 1 468 318 women were included in this study, and all of them were classified according to their SS diagnosis prior to pregnancy.

In Taiwan, people who experience long-term dry eye and dry mouth with suspected SS are referred to specialists for diagnosis and treatment. After confirming an SS diagnosis, which comprised all the patients included in this study, rheumatologists are required to submit a catastrophic illness application. Patients diagnosed with SS are eligible for a medical co-payment waiver. Before approving the waivers, commissioned experts from the insurance administration panel reviewed the application along with complete diagnostic information in compliance with the classification criteria [14]. The Registry for Catastrophic Illness Patients contains information on diagnosis, demographics, application date, diagnosing physician, hospital, and other administration data and has as a unique personal identification; this registry was used to identify women with SS.

Outcome classification

The International Classification of Diseases, Ninth Revision, Clinical Modification (ICD-9-CM) was utilized to establish diagnostic information. We examined both fetal–neonatal and maternal outcomes. Fetal–neonatal outcomes included stillbirth, low birth weight (<2500 g), premature birth (<37 weeks), and small for gestational age (SGA, <10th percentile for the same gestational age) or large for gestational age (LGA, >90th percentile for the same gestational age). Apgar scores at 1 and 5 min after birth, fetal distress, fetal abnormalities, central nervous system malformations, chromosomal abnormalities, suspected damage that is attributable to viral or other diseases in the mother, suspected damage that is attributable to drugs or radiation, decreased fetal movement, and other/unspecified abnormalities were also included. SGA and LGA were indicated by a nomogram summarizing all live births between 2004 and 2014 that were recorded in the National Birth Registry (Supplementary Fig. S1, available at Rheumatology online). Two categories for stillbirth—explained and unexplained—were categorized according to a previous study [15]. Fetal–neonatal and maternal outcomes were identified using codes (see Supplementary Table S1, available at Rheumatology online).

Maternal outcomes were classified into the following five major categories (with subcategories): death, cardiovascular complications, complications during delivery, complications during surgery, and others. Taiwan’s National Death Registry, which records the causes of death for all deceased citizens, was used to confirm maternal death. The accuracy of the coding has been previously validated [16, 17].

Covariate definitions

We assessed the following maternal covariates: country of origin, place of residence, age, income level, occupation, obstetric history and Charlson comorbidity index (CCI). The infant covariates included birth weight, sex and gestational age. The place of residence was selected among 369 towns or districts in Taiwan for each individual. The level of urbanization was indicated as urban, suburban or rural [18]. Patients’ occupations were categorized as follows: (i) civil servants, including teachers and military personnel/veterans; (ii) professionals and non-manual workers; (iii) manual workers; (iv) other; and (v) dependents. The income levels from employee payrolls and employers’ business income were estimated and categorized into gender-specific quintiles. We identified 17 comorbidities for CCI, including myocardial infarction, congestive heart failure, peripheral vascular disease, cerebrovascular disease, dementia, chronic pulmonary disease, rheumatic disease, peptic ulcer disease, mild liver disease, moderate or severe liver disease, diabetes mellitus (DM), DM with chronic complications, renal diseases, any malignancy, metastatic solid tumours and HIV infection. The category of rheumatic disease was not used for CCI. The validated version of the ICD-9-CM codes used in this study were developed by Deyo et al. [19].

Statistical analysis

We compared the rate of each outcome between women with and without SS using a generalized estimating equation model to estimate the odds ratios (ORs) and corresponding 95% CIs. We also considered consecutive pregnancies in the same woman during the study period. An autoregressive model was used to illustrate the correlation structure and two models for a priori-defined analysis. Model 1 was adjusted for maternal age whereas Model 2 was further adjusted for potential confounders, including country of origin, residence, income, occupation, birth year, obstetric history, and CCI. For all statistical hypotheses, a two-sided test with a 5% level of significance was performed. All statistical analyses were performed using SAS version 9.3 (SAS Institute, Cary, NC, USA).

Results

Characteristics of patients with SS

A summary of the demographic, maternal and infant characteristics for women with and without SS is presented in Table 1. Of the 2 100 143 singleton pregnancies recorded between 2004 and 2014 involving 1 468 318 women, 449 infants were born to mothers with SS. The mean pregnancy age in women without SS was 30.2 years. Women with SS tended to conceive at an older age (P < 0.0001). Furthermore, smoking, drinking and place of residence were not significantly different between the groups (Table 1).

Effects of SS on fetal outcomes

High risks of stillbirth, prematurity and SGA were noted on fetal outcomes (Fig. 1). Compared with women without SS, those with SS had poor outcomes regarding stillbirth, unexplained stillbirth, premature birth, SGA, 1-min Apgar score, fetal distress and other unspecified abnormalities (all P < 0.05; Fig. 2, Table 2).

Effect of SS on maternal outcomes

Women with SS had increased risks of adult respiratory distress syndrome (adjusted OR = 5.61, 95% CI = 1.39, 22.60), disseminated intravascular coagulation (adjusted OR = 4.33, 95% CI = 1.09, 17.30), pulmonary oedema (adjusted OR = 11.64, 95% CI = 1.62, 83.48) and eclampsia (adjusted OR = 5.33, 95% CI = 1.34, 21.26) compared with those without SS after adjusting for age, infant sex, urban residence, income, occupation, birth year and CCI (Fig. 3, Table 3).

Most women with SS did not have increased risks of death and no cardiovascular complications during delivery. In addition, women with SS did not have an increased risk of surgical complications or increased odds of other important adverse maternal outcomes. However, women with SS had higher risks of shock (adjusted OR = 6.07, 95% CI 1.51, 24.30) and preterm labour (adjusted OR = 1.95, 95% CI = 1.39, 2.72; Fig. 3, Table 3).

Discussion

This study is the first large database study investigating pregnancy and fetal outcomes in Asian women with SS. Over the 11-year period, the data of 449 pregnancies in patients with SS were collected and analysed from the comprehensive national healthcare database in Taiwan. Elliott et al. analysed 1947 pregnancies in women with SS between 1999 and 2014 [20] and found that women with SS are at a higher risk of preeclampsia, premature membrane rupture, caesarean delivery, pulmonary embolism/deep vein thrombosis, and longer hospital stay than those without SS. Infants born to women with SS were more likely to be premature with intrauterine growth restriction and congenital malformations. Our study as well as that by Elliott et al. highlights the risk to the mother as well as the fetus due to SS.

Several studies have reported a high incidence of spontaneous abortion and fetal loss in women with SS [4, 7, 8, 21]. Furthermore, a significant increase in the number of preterm deliveries has been reported in pregnant women with SS [10, 22]. This might be due to the older age of patients at the time of conception and the possible involvement of immunological factors [21]. In our study, the mean neonatal birth weight was significantly lower in the offspring of women with SS. This might be due to intrauterine growth restriction and not due to the timing of delivery [8]. Elliott et al. [20] reported that women with SS have approximately a 2-fold greater risk of preterm delivery than those without SS, and these findings are in agreement with those of two other studies [8, 11]. However, a meta-analysis by Upala et al. found no significant association between SS and premature delivery and showed that women with SS have an increased risk of total fetal loss [23]. These discrepancies among the previous studies might be due to the small sample size. In our study, the incidence of low birth weight and premature birth was high, and the association between preterm birth and low birth weight is well-documented. Most infants with a low birth weight are premature; however, other in utero factors can also contribute to a birth weight of <2500 g in term infants [24].

Common adverse fetal outcomes in pregnant women with SS include neonatal lupus and congenital heart block (CHB) [25]. CHB is the most severe fetal complication in women with SS and is supposedly due to atrioventricular node damage caused by anti-SS-A and/or anti-SS-B antibodies. However, because our study has not taken into account the presence or absence of anti-SS-A/SS-B antibodies, we are unable to comment on its actual impact.

In contrast to a previous study, we did not observe an increased incidence of caesarean delivery in patients with SS, although a higher risk of preeclampsia was noted in our study population after adjusting for maternal age, which is in accordance with previous studies [14]. However, the risk did not remain significant after adjusting for additional confounders. Based on our findings, we speculate that increased monitoring during pregnancy might benefit women with SS.

In the present study, the ORs for the risk of shock and pulmonary oedema, which is not commonly associated with SS, were 6.07 (95% CI = 1.51, 24.3) and 11.64 (95% CI = 1.62, 83.48), respectively. Both OR values were calculated based on two events. Due to the wide range of CIs, establishing a definite conclusion regarding these outcomes might be inappropriate. However, because SS might be associated with pulmonary hypertension in some cases, peripartum shock and pulmonary oedema cannot be ruled out.

Our study has limitations. The retrospective study design has its inherent drawbacks. Although we retrieved the data from the largest Asian database, our search was limited to Taiwan, thus, we analysed a small number of women with SS. Besides overall score and comorbidities, including hypertension, obesity, and diabetes, other medication information and laboratory data related to anti-SSA/SSB or antiphospholipid antibodies were not assessed in this study. Furthermore, the extremely high number of patients in the control group might have influenced our results. In Taiwan, neonates do not have their own NHI registration until they survive the first month. Before that, neonates are co-registered in their mothers’ records. Therefore, information on neonatal lupus and congenital heart block is not well-registered. Nevertheless, our report presents comprehensive and objective information along with an analysis model. Our control group included patients with other autoimmune diseases, such as SLE, APS and systemic sclerosis. Although the nature of these diseases is non-predictable and they have minimal effects on most people, they might affect patients with SS. Moreover, as there is an increased risk of delayed diagnosis of SS, some cases could have been overlooked in our study. Nonetheless, patients who have been accurately diagnosed were included. Furthermore, only information regarding the number of pregnancies, not the number of mothers, was included, therefore, some data on complications that are associated with a higher risk of neonatal lupus and preeclampsia recurrence could have been lost. Finally, although we obtained maternal and fetal outcome data, there were no individual disease activity records.

In conclusion, SS increases the risk of poor fetal–neonatal and maternal outcomes. Our findings highlight the need for preconception counselling in women with SS to control the disease. Because there are very few published studies on maternal outcomes in women with SS, our findings will be useful for pregnant women with SS. However, further cohort studies assessing disease activity among pregnant women with SS are necessary for further clinical assessment.

Supplementary material

Supplementary material is available at Rheumatology online.

Data availability

The data underlying this article cannot be shared publicly due to the data holder, Ministry of Health and Welfare in Taiwan, prohibiting any unauthorized public release of data to non-research staffs. Currently all data analysis is limited to the dedicated data centre provided by the Ministry. Therefore, we are not able to provide original raw data. However, data of original analysis results can be shared on reasonable request to the corresponding author.

Contribution statement

T.-M.C. and C.-F.K. were involved in drafting the manuscript. All authors were involved in revising it for critical intellectual content. All authors approve the publication of the final version.

Funding

This work was funded by the National Science Council of Taiwan (Project No.: 105–2314-B-182A-135-MY2) and the Chang Gung Memorial Hospital (Project Nos.: CMRPG3F0843, CORPG3G0231, CMRPG3E1962, CMRPD1F0251 and CGRPG3M0021) and supported by the University of Nottingham in terms of methodological assistance.

Disclosure statement: The authors have declared no conflicts of interest.

Acknowledgements

We would like to thank the Centre for Big Data Analytics and Statistics at Chang Gung Memorial Hospital for the statistical consultation. The sponsors of this study had no participation in the design and conduct of the study, including the collection, management, analysis and interpretation of the data, nor in the preparation, review and approval of the manuscript.

References