Neonatal admission after lithium use in pregnant women with bipolar disorders 113 Strengths and limitations Our focus on neonatal admissions to a ward with monitoring after lithium exposure in solely women with BD is novel. Many studies have focused primarily on congenital malformations or birth outcomes instead of the necessity of monitoring the neonate for 24 h on a neonatal ward, which is an important clinical outcome that affects parents and neonates. Moreover, inclusion of a control group consisting of neonates born to women who were diagnosed with BD decreased confounding bias. This is important, as mental disorders itself, including BD, have shown to increase the risk of pregnancy, obstetric and neonatal adverse outcomes6,17. However, this study has potential limitations. First and foremost, analyses were based on a limited sample of neonates, leading to a lack of statistical power when discussing neonatal outcomes, and reason for (re)admission, with low prevalence. This is also applicable to the primary outcome neonatal admission after lithium exposure. Within this sample, lithium-exposed versus non-exposed neonates differed regarding type of BD, which we were unable to correct for due to the limited sample size. Moreover, as 93% of neonates were already exposed to lithium in the first trimester, we were unable to create different research groups to investigate possible differences in first, second or third trimester lithium use. Lithium levels of these women were monitored externally and could not be accessed. Due to the careful monitoring and information of lithium dosage, it is probably safe to assume lithium levels were within therapeutic range. However, other factors such as hydration status could impact lithium levels in neonates. In future studies it is advised to include lithium levels. Also, we found low numbers of women diagnosed with BD recorded in our hospital’s charts in this sample compared to the lifetime prevalence of BD18. Possibly, women with BD were underrepresented or underdiagnosed in OLVG hospital. This would be worrisome, as women with severe mental illness deserve and require tailored obstetric care which is available at the specialised clinic for pregnancy and psychiatric vulnerability in OLVG hospital. Another noteworthy limitation is the common use of psychotropic medication other than lithium in both groups. Although we have adjusted for other psychotropic use in our logistic regression to understand the effect of lithium aside from other psychotropic drugs, overall the other psychotropic medicine (such as antipsychotics (typical and atypical), various types of antidepressants, benzodiazepines, and anticonvulsants could have influenced birth outcomes in both groups, and therefore admission to a neonatal ward with monitoring6. Given this influence of psychotropic medication, we have no information on a causal association between lithium exposure and neonatal admission to a ward with monitoring. On the other hand, the diversity in medication use will make our results more generalizable to the clinical population, as many patients with BD use multiple psychotropic medications. As lithium-exposed neonates were not standardly monitored, not all anomalies might have been discovered. This
RkJQdWJsaXNoZXIy MTk4NDMw