Discussion
With this study, we report that on average 1:1000 women with structurally normal hearts admitted in labour will experience an arrhythmic syndrome—namely SVT, AF or VA—with SVT being the most common (1 in 2000 admissions). More than half of patients with SVT had events prior to admission with only a minority of them being on medical therapy or offered an ablation procedure (28%). History of untreated SVT carried a 10-fold increased risk in developing SVT while in labour. The majority of AF and VA cases were diagnosed denovo during hospitalisation for delivery. Presence of an arrhythmic syndrome was associated with lengthier hospital stay and almost doubled the risk for C-section and PTL. The associated risks were mostly driven by presence of SVT and AF, while VAs were not associated with increased obstetrical adverse outcomes. The three outcomes examined in our study are interrelated, as the increased incidence of CS would predispose women to increased lengths of stay.
The prevalence of maternal SVT, AF and VA in our patient population was higher compared with prior studies. A study by Li et al reviewed a similar number of pregnancy admissions via a retrospective analysis of discharge ICD-9 codes from the period between 1992 and 2000. Their study found the rate of SVT, AF and VA to be 0.03%.4 Similarly, Vaidya et al’s review of discharge diagnoses from 2000 to 2012 found the rate of SVT, AF and VA to be 0.071%.3 There are several potential explanations for this discrepancy. First, there has been a general increase in maternal age and chronic disease in the general population in the time since the publication of those earlier trials, which likely has contributed to an increased burden of maternal arrhythmia.2 Also, these previous studies relied on a review of discharge diagnoses to identify the frequency of arrhythmias in pregnancy-related hospitalisations. Unlike in our study, additional sources such as the medical notes, ECGs and telemetry tracings of patients were not reviewed. The robust methodology used in our study increased our ability to identify cases of arrhythmia that otherwise may have been missed if solely relying on an administrative database. Therefore, the thorough review of our common EMR using this rigorous method likely improved the sensitivity in identifying arrhythmic events, even among patients with a structurally normal heart.
Arrhythmias in pregnancy have previously been described as a risk factor for adverse maternal and fetal outcomes, including increased mortality.5 However, it is unclear whether this association is due to underlying heart disease that typically accompanies serious arrhythmia syndromes. In our study, we examined the clinical significance of arrhythmias that are considered ‘benign’ when encountered in the general population with a demographic and clinical profile similar to that of our cohort’s. The independent association of these arrhythmic syndromes with adverse obstetric outcomes suggests that even ‘benign’ dysrhythmias carry different prognoses when occurring during pregnancy, and particularly labour. The aetiology of higher morbidity is multifold and includes pronounced haemodynamic consequences of arrhythmias during pregnancy, limited therapeutic options for pregnant women, lack of experience and clear guidelines in management and also the need for a multidisciplinary approach in the management of these patients. It was determined that all patients in the arrhythmia cohort had a cardiology consult during their admission.
The presence of SVT influenced the decision to perform CS on an urgent basis in 7 women due to haemodynamic compromise and five due to fetal distress. Presence of AF determined the decision to proceed with CS in two cases due to fetal distress. We acknowledge that the presence of recurrent arrhythmias likely increased the practice of placing women on frequent or continuous heart monitoring, potentially increasing the yield of diagnosing fetal decelerations. The remaining CS performed in the arrhythmia cohort were performed in a semielective manner after collaborative decision between the obstetric, anaesthesia and cardiology consultants with the intention to minimise perilabour maternal and fetal complication risk.
Our study also outlines the importance of early diagnosis and management of arrhythmic syndromes in women planning to become pregnant. While most AF and VAs were newly diagnosed, more than half of patients with SVT carried this diagnosis in their medical history. In addition, more than two thirds of patients with a prior SVT diagnosis were not offered preventive treatment at all and were exposed to a 10-fold higher risk of developing acute SVT during labour. According to ACC/HRS, catheter ablation is considered a class I indication for patients in the general population with symptomatic SVT.6 However, observation is also a recommended option for the same population of patients, which may lead to an underutilisation of ablation. A more aggressive approach might be appropriate for women planning to become pregnant,3 as advocated by the European Society of Cardiology .6 7 In addition, for the acute management of SVT, prior concerns over radiation exposure to the fetus due to fluoroscopy during catheter ablation have been addressed by the rise of non-fluoroscopic tools such as electroanatomic mapping systems and intracardiac echocardiography.8 It is now an effective, mature therapeutic modality for the treatment of SVTs, even in pregnancy.6 9 Given the likelihood of arrhythmia recurrence during pregnancy with its resulting adverse outcomes, patients with a history of arrhythmias should involve their cardiologists or electrophysiologists when planning their pregnancy to address this treatable condition.
In contrast to SVT cases, the majority of AF cases were newly diagnosed in our cohort. This could be explained by the fact that SVT is typically diagnosed in women of childbearing age, while AF is diagnosed later in life.5 The higher incidence of newly diagnosed AF during labour might support the notion that neurohormonal and haemodynamic changes during pregnancy have an atrial proarrhythmic effect.10 Further studies are needed to examine the long-term prognosis of these women and determine whether this relationship between labour and AF is temporary or long lasting, similar to conditions such as gestational hypertension and gestational diabetes.
Finally, we report that the occurrence of VAs with benign characteristics triggered the most response in acute medical management with the majority being initiated on beta blockers. However, these VAs were not associated with adverse obstetrical management. This finding belies the general perception that ventricular ectopy is associated with worse cardiovascular prognosis. Rather, our study supports that both in the general population and during pregnancy, the occurrence of ventricular ectopy is often triggered by neurohormonal changes that carry a benign prognosis in patients with structurally normal heart.
Study limitation
This study is subjected to the inherent limitations of a case–control study and relies heavily on the accuracy of written documentation. However, the common EMR of our health system allowed for an effective review of a large volume of medical records from multiple hospitals. We were not able to adjust for prior diagnosis of arrhythmias or other major comorbidities in any of the models, as the number was too low to include as effects in the models. Nevertheless, patients with structurally abnormal hearts were excluded from the analysis. Structural heart disease was excluded first through ICD codes and further through manual review of the final cohort. Due to the retrospective nature of this study, we reviewed all available documentation and diagnostic testing to rule out the presence of structural heart disease. However, given the clinical profile of the cohort, not all women had an echocardiogram performed on admission. We cannot rule out the presence of undiagnosed structural heart disease structural heart disease was identified as a confounder in our study design and the cohort was restricted to those without structural heart disease. Given the small sample size we employed this methodology rather than adjusting the models in our analysis.
Prior diagnosis of SVT was identified after manual adjudication in the arrhythmia cohort but only by absence of specific diagnoses and key phrases in the control group. It is likely that the previous medical history of arrhythmias was under-reported in the control group, especially for women with infrequent symptomatic episodes and for those who underwent a curative catheter ablation years prior to becoming pregnant. The precise burden of arrhythmia in each patient could not be determined due to the retrospective nature of the analysis. In addition, it is plausible that there was a degree of underdocumentation of history of benign arrhythmias by the obstetric services. The above might have underestimated the importance of medical history on arrhythmia recurrence during labour as well as the importance of early arrhythmia management prior to pregnancy. With the available clinical documentation, we could not ascertain the primary reason for extended LOS in those who underwent caesarean section; however, the main findings of our paper are the increased rate of caesarean section among this population.
