The Perinatal Treatment Conundrum

Psychiatric disorders are common during pregnancy. For those women with symptoms prior to pregnancy, symptoms can intensify, placing the mother at risk for continued illness during the postpartum as well as for other psychiatric issues. As there can be differing manifestations of various psychiatric disorders across the perinatal period (e.g., the finding of relative quiescence of schizophrenia-like disorders during pregnancy¹), treatment decisions regarding medication prior to pregnancy become more complex. Among the factors to consider, experts recommend weighing the potential adverse effects of the illness on the mother and baby against the potential adverse effects and benefits of medication on outcomes, as well as the effects of medication discontinuation when faced with treatment decisions. This is often difficult to do as our understanding of many relevant issues remains incomplete. The dilemma faced by a woman is that her choice does not appear to be without consequence for her, the infant, or family, regardless of the decision she may make. Given just under half of all pregnancies are unplanned, it is imperative to further our understanding.

In this issue, Leong et al.² provide a piece of the missing information, specifically the lacking population-based Canadian data on all psychotropic drug use during the perinatal period. They used administrative data located at the Manitoba Centre for Health Policy to examine psychotropic medication use across the 12 months prior to pregnancy, during pregnancy, and up to 3 months postpartum in all pregnancies occurring between April 1, 2000 and December 31, 2012 (N = 224,762). They report increased rates of psychotropic use from 2001 to 2013, consistent with other reports of increasing prescription rates overall in pregnant women.³ Moreover, the 1.5-fold increase over time in the 3 to 12 months prepregnancy is comparable to other Canadian data on secular trends in the rate of overall psychotropic prescriptions.⁴ Contributors to these rising rates need to be explored. Specifically, are we as health care providers doing a better job at detecting and treating or is prevalence increasing? Is there an increasing tendency to prescribe psychotropics or, perhaps, greater acceptance by patients? These are important questions, but the fact that the timing is around the perinatal period has further implications for practice.

Leong et al.² found that up to 16% of women used psychotropics in the 3 to 12 months prior to pregnancy; however, of those who used 3 months proximate to pregnancy, only 10% continued use during pregnancy while close to 40% and 20% discontinued prior to the first and second trimesters, respectively, and a further 30% demonstrated intermittent use throughout pregnancy. International studies have also reported high rates of discontinuation of psychotropic medications around pregnancy.^5,6 With the exception of sedatives/hypnotics, psychotropic medications are typically to be used daily, not intermittently. This highlights the need for increased knowledge of the reasons for discontinuation and/or intermittent use and the consequences of cessation to adequately address treatment needs. Risk perception is a critical component of the clinical decision-making process, and although there have been numerous reports of fear of adverse effects on the fetus as a factor for discontinuation, some have argued women can overestimate the risk of using drugs during pregnancy.⁷ Indeed, decisional conflict does appear to be at play,⁸ and risk tolerance may be an important factor as well, although both have not been well studied.

There has been a surge of research on the potential adverse effects of psychotropic use lately, particularly for selective serotonin reuptake inhibitors (SSRIs). Poor outcomes, including preterm birth and immediate neonatal outcomes, among others,^9–13 have been reported in association with prenatal use of SSRIs, benzodiazepines, antipsychotics, and anticonvulsants. However, with the exception of valproate, some of the clinical implications for many of the adverse effects of these drugs have been debated. The literature on the risks of untreated psychiatric illness continues to evolve as well and is confounded by the association between these disorders and a number of behaviours known to affect outcomes such as smoking, poor self-care, and nutrition. Nevertheless, antenatal depressive and anxiety disorders, bipolar disorder, and schizophrenia have each been associated with preterm birth and various other adverse obstetric and neonatal outcomes.^14–17 Of note, a recent UK study found that 1 of 25 women of early reproductive age (20-35 years) died by suicide perinatally.¹⁸ Clearly, there are major implications to having active psychiatric illness during pregnancy that must be treated and this is an area of continued active research. Further, the data on the efficacy of psychotropic medication are limited in pregnancy. It is also not clear if the treatment of the disorders with medications is associated with a reduction in the obstetrical/neonatal adverse effects reported for untreated illness. The literature on the effects of medication discontinuation is also not extensive and at times may seem contradictory. For example, some have found increased rates of relapse for depression¹⁹ while others have not,²⁰ although these differences may relate to illness severity. For bipolar disorder, women with known disease who discontinue medications are at much higher rate of relapse both during pregnancy and postpartum.²¹ The facets needing consideration in the decision process are multiple, with many parts still not clearly understood.

Another key finding reported by Leong et al.² is that the women who were more adherent to psychotropic drugs (at least 80% adherent and filled at least 2 prescriptions 3 months prior to pregnancy) in the prepregnancy stage were more likely to continue use throughout pregnancy. We see among this group that close to 57% continued to use medication throughout the perinatal period with about 16% and 15% discontinuing prior to the first and second trimesters. As the authors suggest, this may reflect continued use in pregnancy by those with the most severe, chronic, and/or recurrent mental disorders. This aligns with the message put forth by clinical experts where psychotropic use is recommended for women with moderate to severe illness and those who are at higher risk of relapse when discontinuing medications. Moreover, the fact that the authors examined adherence is notable as the administrative database research is weakened by the uncertainty of whether the medication was consumed. The authors examined not only repeat prescriptions but also those with an average medication possession ratio >80%. This is a significant contribution as we can have more confidence that the women who appear to need the medication most may be the most likely to take it.

Although there has been a surge of interest in the potential effects of medications, most of the data have been on antidepressants. Leong et al.² present data on psychotropics broadly, divided by class, including sedatives/hypnotics and stimulants, where these is even less data. While anxiolytics/sedatives and hypnotics were most used before pregnancy, antidepressants were most frequently used throughout (i.e., prior and during the perinatal period). Leong et al.² found that most of the prescriptions for anticonvulsants, followed by antipsychotics, were the most likely to be continued among adherent users (antidepressants and lithium followed antiepileptics when not limiting only to the adherent group). This is reassuring as these medications are used for bipolar disorder and psychosis, which are typically amongst the most severe of the psychiatric illnesses, although indication was not examined; it is possible that women with epilepsy confounded this (10% of continuous users had an epilepsy diagnosis in the 3 years prior to start of pregnancy, for example). At least 1 prior study found higher discontinuation rates for pregnant women with bipolar disorder using antiepileptic drugs than for those with epilepsy.²² Moreover, the pattern of use in the Leong et al.² study indicated that for those who did not discontinue medication, they would use the medication throughout the perinatal period. A final implication of this study relates to the need for assessment of medication adherence. The authors, however, do point out that they were not able to determine if the clinician was aware/supportive of discontinuation. As women receive information from several sources, we as health care providers have the responsibility to provide the most up-to-date evidence-based information so that necessary treatment decisions are made based on data.

In no other area of psychiatry is it as imperative to work collaboratively with the patient and her partner as any decision will affect the family unit. The decision whether to use psychotropic medication during pregnancy must be guided by an open, informed, and dynamic risk/benefit discussion between clinician and patient. There are consequences to using and not using medications for psychiatric illness perinatally. Psychosocial treatments are of course first-line options when appropriate. We continue to have more questions with an incomplete understanding of all end points of the branches in the decision tree but nevertheless must do our best with what we have to date, and it appears reassuring to think that at least those who were adherent to treatment continue with their chosen course. This study provides needed pieces of the puzzle from a Canadian perspective.