Consensus Panel Recommendations for the Pharmacological Management of Pregnant Women with Depressive Disorders

Introduction: The initiative of a consensus on the topic of antidepressant and anxiolytic drug use in pregnancy is developing in an area of clinical uncertainty. Although many studies have been published in recent years, there is still a paucity of authoritative evidence-based indications useful for guiding the prescription of these drugs during pregnancy, and the data from the literature are complex and require expert judgment to draw clear conclusions. Methods: For the elaboration of the consensus, we have involved the scientific societies of the sector, namely, the Italian Society of Toxicology, the Italian Society of Neuropsychopharmacology, the Italian Society of Psychiatry, the Italian Society of Obstetrics and Gynecology, the Italian Society of Drug Addiction and the Italian Society of Addiction Pathology. An interdisciplinary team of experts from different medical specialties (toxicologists, pharmacologists, psychiatrists, gynecologists, neonatologists) was first established to identify the needs underlying the consensus. The team, in its definitive structure, includes all the representatives of the aforementioned scientific societies; the task of the team was the evaluation of the most accredited international literature as well as using the methodology of the “Nominal Group Technique” with the help of a systematic review of the literature and with various discussion meetings, to arrive at the drafting and final approval of the document. Results: The following five areas of investigation were identified: (1) The importance of management of anxiety and depressive disorders in pregnancy, identifying the risks associated with untreated maternal depression in pregnancy. (2) The assessment of the overall risk of malformations with the antidepressant and anxiolytic drugs used in pregnancy. (3) The evaluation of neonatal adaptation disorders in the offspring of pregnant antidepressant/anxiolytic-treated women. (4) The long-term outcome of infants’ cognitive development or behavior after in utero exposure to antidepressant/anxiolytic medicines. (5) The evaluation of pharmacological treatment of opioid-abusing pregnant women with depressive disorders. Conclusions: Considering the state of the art, it is therefore necessary in the first instance to frame the issue of pharmacological choices in pregnant women who need treatment with antidepressant and anxiolytic drugs on the basis of data currently available in the literature. Particular attention must be paid to the evaluation of the risk/benefit ratio, understood both in terms of therapeutic benefit with respect to the potential risks of the treatment on the pregnancy and on the fetal outcome, and of the comparative risk between the treatment and the absence of treatment; in the choice prescription, the specialist needs to be aware of both the potential risks of pharmacological treatment and the equally important risks of an untreated or undertreated disorder.


Introduction
The initiative of a consensus on the topic of antidepressant and anxiolytic drug use in pregnancy is developing in an area of clinical uncertainty. Although many studies have been published in recent years, there is still a paucity of authoritative evidencebased indications useful for guiding the prescription of these drugs during pregnancy; moreover, the data from the literature are complex and require expert judgment to draw clear conclusions. Antidepressant and anxiolytic drugs in pregnancy may constitute a potential risk to the developing fetus but are also a clear benefit for patients with acute symptoms or for those with a high risk of relapses or recurrences. On the other hand, clinicians, patients and their partners should always keep in mind that women with untreated depressive symptoms or major depressive disorder (MDD) may have several pregnancy complications (miscarriage, preterm birth, placental abruption, intrauterine hemorrhages) and may deliver newborns with an increased risk of preterm birth and small for gestational age status. Moreover, abrupt discontinuation of antidepressants during pregnancy may be associated with relapses/recurrences.
The art of medicine is exemplified by balancing the risks of untreated psychiatric disorders in general, depressive symptoms or MDD for example, during pregnancy and the risks of adverse outcomes associated with the exposure to medications (antidepressants in this case). There is the following common concern in prescribing medications, such as selective serotonin reuptake inhibitors (SSRIs), serotonin and norepinephrine reuptake inhibitors (SNRIs), tricyclic antidepressants (TCAs), atypical antidepressants, benzodiazepines (BDZs) and non-benzodiazepines hypnotics drugs (hypnotic benzodiazepine receptor agonists-HBRAs), during pregnancy: the possibility that prenatal psychopharmacologic treatment may prevent adverse outcomes during human pregnancy is less commonly considered than the possibility of fetal harm. This has led, at least in the past, to not prescribing medications to patients in need of them and/or at risk for suicide or self-harm or to immediately discontinue antidepressants for patients already on treatment, with potential adverse outcomes.
The objective of this consensus is to provide a concise and specific guide to the therapeutic choices to be implemented in pregnant women affected by depressive disorders.

Materials and Methods
For the elaboration of the consensus, it was considered appropriate to involve the scientific societies of the sector, namely, the Italian Society of Toxicology (SITOX), the Italian Society of Neuropsychopharmacology (SINPF), the Italian Society of Psychiatry (SIP), the Italian Society of Obstetrics and Gynecology (SIGO), the Italian Society of Drug Addiction (SITD) and the Italian Society of Addiction Pathology (SIPaD).
The chairperson and the coordinators of the consensus were members of the Italian Society of Toxicology and the Italian Society of Neuropsychopharmacology. They defined the objectives, scope, users and sections of the document, as well as the specific areas of the investigation. They selected a multidisciplinary group consisting of experts in research and management of anxiety and depressive disorders in pregnancy. They identified 8 leading Italian experts in women's mental health, in neonatal complications and on the use of drugs in pregnancy and lactation. Selection of the expert group was based upon their practical input in diagnosis and treatment of major depression in pregnancy and their professional expertise in the field of the consensus. None of the experts had any potential conflicts of interest or funding sources in order to ensure the experts' transparency and credibility. Our guidelines are not financially sponsored. All experts had long-term experience and reviewed publications in the field.
This interdisciplinary team was composed of experts from the following different medical specialties: 6 toxicologists and pharmacologists, 4 psychiatrists, 2 gynecologists, 1 neonatologist and included representatives of all the aforementioned scientific societies. The role of the toxicologists was the evaluation of the risk of malformations after in utero use of the medicines; the interest of the gynecologists and the neonatologist was the complications in pregnancy and the neonatal abstinence syndrome, respectively, and the psychiatrists evaluated the risks due to the psychiatric disease of the patients involved in all published articles.
Such team first identified the needs underlying the consensus using the methodology of the "Nominal Group Technique" (NGT). The NGT approach was chosen as it is a faceto-face group meeting process, generates a large number of ideas, allows for immediate feedback and is facilitated by experienced leaders. The NGT is a commonly used consensus method in medical research, which uses a panel of specialists to discuss and provide prompt results for researchers. NGT is highly structured and involves generation, clarification and sharing of ideas and voting.
In our study, from September 2022 to June 2023, several NGT meetings were conducted in order to reach a consensus statement on the pharmacological management of pregnant women with depressive disorders.
The first face-to-face NGT was held in Bergamo, [8][9][10] September 2022, during the 33rd Conference of the European Network of the Teratology Information Services, followed by adjustment of the first draft of consensus based on expert feedback via e-mail. Then, the team evaluated the most accredited international literature by a systematic review of the literature. Finally, after various online meetings, the research team reviewed all the materials and defined a list of consensus statements and through several discussion meetings, participants reached consensus on the draft and the final document.
The following six areas of investigation were identified (the order of presentation of the topics is intended to guide the clinical approach to women with a psychiatric disorder who intend to plan a pregnancy): 1. Assessment of the risks associated with untreated maternal depression (depressive symptoms or MDD) during pregnancy; 2. Assessment of the overall risk of malformations associated with antidepressant and anxiolytic drug use during pregnancy; 3. Assessment of the risk of maternal complications during pregnancy and at delivery related to pharmacological treatment, such as gestational hypertension, pre-eclampsia, gestational diabetes, intrauterine growth retardation, spontaneous abortion, preterm birth and postpartum hemorrhage; 4. Evaluation of neonatal adaptation disorders and persistent pulmonary hypertension of the newborn in the offspring of women exposed to antidepressants and/or anxiolytics during pregnancy; 5. Long-term developmental outcomes of infants' cognitive development or behavior after in utero exposure to antidepressant/anxiolytic medications; 6. Evaluation of pharmacological treatment of opioid abuse pregnant women with depressive disorders.

Definitions
1. Major depressive disorder (MDD) is defined by the DSM-5 [1] as a common and potentially severe mood disorder (depressive disorder in DSM-5): diagnostic criteria require that the patient is experiencing, during the same 2-week period, five or more symptoms, at least one of which should be either (1) depressed mood or (2) markedly diminished interest or pleasure (see Box 1).
2. Organ malformation (teratogenicity) was considered drug-induced if fetal drug exposure happened during the first 12 weeks of gestation when organ development occurs; in particular, a medication was considered teratogenic when prenatal exposure was associated with a significant increase in the risk of congenital physical deformities over the baseline risk. The baseline incidence of major congenital malformations (MM) ranges from 2% to 4% and has been estimated to be more than 15% if minor malformations are included [2][3][4]. 3. The following gestational and postpartum complications were considered as defined by the American College of Obstetricians and Gynecologists' Committee on Practice Bulletins [5][6][7][8][9][10]: gestational hypertension, pre-eclampsia, gestational diabetes, intrauterine growth retardation, spontaneous abortion, preterm birth and postpartum hemorrhage (Box 2). 4. Neonatal withdrawal syndrome was considered according to the definition provided by Anbalagan & Mendez [11]. The syndrome includes a wide range of physical and behavioral symptoms, usually associated with the use of drugs during pregnancy, appears shortly after birth and symptoms are typically of limited duration. Persistent pulmonary hypertension of the newborn (PPHN) is defined, according to Mandell et al., as the failure of the normal circulatory transition that occurs after birth. It is a syndrome characterized by marked pulmonary hypertension that causes hypoxemia secondary to right-to-left shunting of blood at the foramen ovale and ductus arteriosus [12]. 5. Developmental toxicity refers to the potential for long-term neurobehavioral abnormalities in children following in utero exposure to medications. 6. Opioid use disorder is defined according to American Psychiatric Association DSM-5 criteria and includes a persistent desire or unsuccessful efforts to cut down or control opioid use despite negative social and/or professional consequences.
1. Depressed mood, most of the day, nearly every day.

2.
Markedly diminished interest or pleasure in all, or almost all, activities most of the day, nearly every day. 3.
Significant weight loss when not dieting or weight gain, or decrease or increase in appetite nearly every day.

4.
A slowing down of thought and a reduction of physical movement (observable by others, not merely subjective feelings of restlessness or being slowed down).

5.
Fatigue or loss of energy nearly every day. 6.
Feelings of worthlessness or excessive or inappropriate guilt nearly every day. 7.
Diminished ability to think or concentrate, or indecisiveness, nearly every day. 8.
Recurrent thoughts of death, recurrent suicidal ideation without a specific plan, or a suicide attempt or a specific plan for committing suicide.

Box 2. Definition of common gestational and postpartum complications.
Gestational hypertension is defined as systolic blood pressure ≥ 140 mmHg or diastolic blood pressure ≥ 90 mmHg, or both, on two occasions at least 4 h apart after 20 weeks of gestation, in woman with previously normal blood pressure.
Pre-eclampsia is defined as the new-onset hypertension and proteinuria in pregnancy or new-onset hypertension with the new onset of any of the following: thrombocytopenia with platelets < 100 × 10 9 /L or renal insufficiency with serum creatinine > 1.1 mg/dL or impaired liver function with elevated blood concentrations of liver transaminases to twice normal concentration or pulmonary edema or unexplained new-onset headache unresponsive to medication-without an alternative diagnosis-or visual symptoms. Gestational diabetes mellitus is defined as any degree of glucose intolerance with onset or first recognition during pregnancy. Postpartum hemorrhage is defined as cumulative blood loss of greater than or equal to 1000 mL or blood loss accompanied by signs or symptoms of hypovolemia within 24 h after the birth. Intrauterine growth retardation describes fetuses with an estimated fetal weight or abdominal circumference that is less than the 10th percentile for gestational age. Preterm birth is defined as a delivery occurring at or after 20 (+0/7 d) weeks of gestation and before 37 (+0/7 d) weeks of gestation. Spontaneous abortion or early pregnancy loss is defined as a nonviable, intrauterine pregnancy with either an empty gestational sac or a gestational sac containing an embryo or fetus without fetal heart activity within the first 12 weeks of gestation.

Search Strategy
A systematic review of the literature was performed to assess the risks associated with untreated maternal depression in pregnancy and the risk of malformations with the antidepressant and anxiolytic drugs used in pregnancy. The search strategies for the PubMed, EMBASE and Cochrane library databases were defined using a list of combined generic keywords and medical subject headings. In order to be included, studies had to fulfill the following criteria: (1) they had to include pregnant patients with depression; (2) they had to provide data on pregnancy and delivery and on drug exposure during pregnancy; and (3) they had to be a meta-analysis, systematic review of the literature or observational studies.
The search for relevant studies using generic keywords and MeSH terms was the following: "depression during pregnancy" AND "malformations" OR "teratogenicity" OR "pregnancy outcomes" OR "perinatal outcome" OR "neonatal outcome" AND "selective serotonin reuptake inhibitors" OR "serotonin and norepinephrine reuptake inhibitors" OR "tricyclic antidepressants" OR "atypical antidepressants" OR "benzodiazepines" OR "hypnotic benzodiazepine receptor agonists" OR "opioids" AND "during pregnancy". Namely, the following compounds were considered: citalopram, escitalopram, fluoxetine, fluvoxamine, paroxetine and sertraline for the SSRIs; venlafaxine and duloxetine for the SNRIs; vortioxetine for the Serotonin Modulator and Stimulator antidepressants (SMS); bupropion, mirtazapine, reboxetine and trazodone for the atypical antidepressants; amitriptyline and clomipramine for the TCAs.
Periods between 1 January 1995 and 1 January 2023 were addressed. This information enabled the group to develop a series of recommendations.

Results
We have identified 620 records about the use of antidepressants during pregnancy, and after the exclusion of other languages other than English and experimental studies, we found 417 eligible studies. Searching only the full-text articles, we found 398 records. From them, we excluded 61 papers concerning case reports and case series, and 43 records were excluded for other reasons. From the remaining records, 30 reports were not retrieved. The research strategy allowed the retrieval of 264 references, and a total of 6 recommendations were generated. The literature results about the risk of congenital malformations are summarized in tables below.
The group of SSRIs accounted in 17 meta-analyses and 51 cohort and case-control studies (32 retrospective, 17 prospective and 2 population-based cohort studies); the SNRIs one was composed of 2 meta-analyses, 8 retrospective, 3 prospective and 2 populationbased cohort studies. The tricyclic antidepressant and the atypical antidepressant group were composed of 1 meta-analysis, 7 retrospective, 1 prospective and 1 population-based cohort studies and 1 meta-analysis, 5 retrospective, 5 prospective and 1 population-based cohort studies, respectively. Eventually, the BDZs and HBRAs accounted in 4 meta-analyses, 9 retrospective and 2 prospective cohort studies.
Key findings: MM and CM aggregate risk results are contradictory. When the underlying psychiatric condition is considered, aggregate MM or CM risk results as non-significant.
We found 10 studies that evaluated malformation risk after exposure to TCA antidepressants during pregnancy ( Table 4). Most of these studies did not control for possible indication bias as follows: three studies evaluated MM risk; one resulted in significant risk [49] while two resulted in non-significant MM risk [32,44]; six studies evaluated CM risk, three resulted in significant risk [41,49,64] while three resulted in non-significant CM risk [32,44,56]. After controlling for indication, or at least partially accounting for an underlying psychiatric condition, no study resulted in a significant risk of MM, while four resulted in non-significant MM risk [31,32,65,69], only one resulted in significant risk for CM [64] while three resulted in non-significant CM risk [32,56,65].
We also found 12 studies that evaluated malformation risk after exposure to multimodal antidepressants during pregnancy (Table 5). Of these, the majority reported results not controlled for possible indication bias as follows: seven studies evaluated MM risk; one resulted in significant risk [85] while six resulted in non-significant MM risk [49,[87][88][89][90][91]; five studies evaluated CM risk, one resulted in significant risk [64] while four resulted in non-significant CM risk [49,56,59,92]. After controlling for indication, or at least partially accounting for an underlying psychiatric condition, no study resulted in significant risk for MM, while two resulted in non-significant MM risk [85,93]; only one study reported significant CM risk [64] while three studies resulted in non-significant CM risk [56,59,93].
We found the following four meta-analyses that summarized malformation risk after exposure to BDZ during pregnancy: one [94], not controlling for possible confounding by underlying psychiatric condition, reported significant pooled MM risk for case-control studies but non-significant pooled risk for cohort studies; two meta-analyses [95,96] reported non-significant MM or CM risk overall, even after controlling for possible indication bias [96]. We found 11 studies that evaluated malformation risk after exposure to BDZ or HBRA during pregnancy (Table 6). Of these, six reported data not controlled for possible indication bias, eight studies evaluated MM risk, one study resulted in significant risk [97] while seven resulted in non-significant MM risk [53,[97][98][99][100][101][102]; four studies evaluated CM risk, no study resulted in significant risk while all of them resulted in non-significant CM risk [32,53,99,101]. After controlling for indication, or at least partially accounting for an underlying psychiatric condition, one study resulted in significant risk [103] while five resulted in non-significant MM risk [62,98,100,102,104]; one study resulted in significant risk for CM [103] while two resulted in non-significant CM risk [62,102].
Key findings: MM and CM risk results tended toward overall non-significant risk (124 vs. 37 results). When an underlying psychiatric condition is considered, MM or CM risk results as non-significant.

Complications in Pregnancy and Delivery
Complications in pregnancy and delivery, such as gestational hypertension, preeclampsia, gestational diabetes, intrauterine growth retardation, spontaneous abortion, preterm birth and postpartum hemorrhage associated with antenatal antidepressant or anxiolytic treatment, do not seem to overcome the baseline risk of the general population or the risks associated to the illness (Table 7).
Compared to the general population, a significant risk of gestational hypertension (OR 2.49, 95% CI 1.62-3.89) [107] and pre-eclampsia (OR 1.6, 95% CI 1.06-2.39) [108] in the SSRIs treated patients was reported in two [107,109] and one studies, respectively [108], while in other three studies [43,110], the risk was not significant. When considering any antidepressant use, significant gestational hypertension and pre-eclampsia risk was found in one [109] and three [49,80,108] studies, respectively. When antidepressant exposure was controlled for underlying condition (at least partially), a non-significant risk for SSRI was found in five studies for pre-eclampsia [80,107,[111][112][113] and in two studies for gestational hypertension [68,109], while the risk remained significant for antidepressant use in three studies [108,109,111].
Key findings: The results are contradictory, but when controlled for depression in the mother, gestational hypertension or pre-eclampsia risk tended toward non-significant.
A possible risk of intrauterine growth retardation has been reported [114], but results are not clear, with another study reporting a non-significant small for gestational age risk [115]. Considering the gestational diabetes risk, exposure to antidepressants resulted in significant gestational diabetes risk in two studies [49,116] and a non-significant risk in one study [43]. A non-significant gestational diabetes risk was reported for any antidepressants [117,118] and for SSRI [116] when controlling for underlying depression in three studies overall.
Key findings: Gestational diabetes risk was reported mainly as non-significant. A significant spontaneous abortion risk has been reported for women exposed to antidepressants compared to the general population in two meta-analyses [19,119] and seven observational studies [43,52,88,[120][121][122][123], but another 14 studies did not confirm these results [46,49,54,75,84,87,89,117,124], in particular for SSRI use [38,[71][72][73]81]. When an underlying psychiatric condition was considered, the results were not clear. Two studies, one of them a large population-based study, reported non-significant spontaneous abortion risk in women exposed to antidepressants when stratified for depression lifetime [84,122], but one study with a similar design reported significant spontaneous abortion risk [120].
Key findings: Spontaneous abortion or early pregnancy loss after antidepressant exposure are mainly reported as non-significant.
Key findings: Results for preterm birth risk are contradictory, even after controlling (at least partially) for an underlying condition.
The risk of postpartum hemorrhage after antidepressant exposure was reported as significant in four studies [49,80,150,151] and non-significant in another four studies [68,[151][152][153] and seems to be very small to none [151]. After controlling (at least partially) for an underlying psychiatric condition, PPH risk was reported as non-significant in one study [68] and significant for SSRI [151] and duloxetine [80] in one and one study, respectively.

Persistent Pulmonary Hypertension of the Newborn
The association between the use of SSRIs in pregnancy and PPHN cannot be confirmed (Table 8). Two meta-analyses concluded for significant risk, especially for SSRI or SNRI exposure late in the gestational period [154,155], but the risk was non-significant for anytime exposure [155]., Risk has been evaluated in 11 studies and resulted significant in 7 [52,54,[156][157][158][159][160] and non-significant in 3 studies [63,161,162]. However, the absolute risk is very low, ranging from 0.2% to 0.3% of the cases. In a large population-based cohort study of 29,822 live-born singleton children born to mothers who used antidepressants, the risk was not-significant when controlled for psychiatric history and antidepressant use before pregnancy (OR = 1.29; 95% CI 0.95-1.74) [156]. Also, three studies reported non-significant PPHN risk for antidepressant exposure after controlling for underlying conditions [86,156,159], while only one reported significant risk [54]. Key findings: the results remain not clear.

Neonatal Abstinence Syndrome
Neonatal complications, such as neonatal abstinence syndrome, are the main consequence of antidepressants and anxiolytic therapy during pregnancy, with an incidence of 30% of the neonates or much more, depending on the epidemiological study; paroxetine is generally the SSRI most frequently implicated, followed by fluoxetine. Thirty-nine studies and one meta-analysis (Table 9) have been evaluated. The only meta-analysis evaluated reported a non-significant risk of poor adaptation symptoms in the newborn [130]. However, the risk of NAS after in utero use of antidepressants was significantly increased in 26 studies [42,44,46,49,54,68,73,76,115,135,140,142,143,[163][164][165][166][167][168][169][170][171][172][173][174][175], and only 7 studies reported a non-significant risk for NAS symptoms or intensive care need [43,126,133,139,166,176,177]. It has been reported an incidence from 13.8% to 71.4%, depending on what signs and/or symptoms of NAS have been evaluated. Near-term exposure to antidepressants increased the risk of NAS, from 2-fold to 10-fold, depending on the comparison group. After controlling for an underlying condition in the mother, eight studies still reported significant NAS risk [54,68,72,86,115,140,143,173], while four reported non-significant risk for intensive care need [126,141,142,172]. The neonatal abstinence syndrome symptoms risk is significantly increased after antidepressant exposure during pregnancy.
It is more difficult to assess the incidence and severity of the neonatal withdrawal syndrome related to BDZ because, in the published studies, this potential adverse effect on the newborn has been assessed in pregnant women undergoing polypharmacological treatment (e.g., BDZ and antidepressants) [166,178] or exposed to other substances such as opiates. Overall, in these situations, neonatal withdrawal syndrome was observed in one out of three newborns. Concerning BDZ use in late pregnancy, the risk of NAS is as high as in the antidepressant group. The assessment of the NAS due to BDZ exposure during pregnancy is difficult, as most of the studies include other comedications (e.g., antidepressants) or they have collected opioid-exposed mothers. It has been reported that from 20% to 40% of the neonates with NAS were exposed to benzodiazepines.
Eventually, there is strong evidence of the association between opioid abuse with anxiety and depression, and the pharmacotherapy for substance abuse and addiction with methadone or buprenorphine treatment does not increase the risk of congenital malformations, while the risk of neonatal abstinence syndrome can overcome the 60% of the cases.    Of all infants exposed to an SSRI in utero, 20%-77% develop symptoms of NAS, "most studies reported percentages around 30%". "SSRI half-life seems to influence the risk of withdrawal and toxicity". TCAs: "20-50% develops PNA" Venlafaxine PNA risk comparable to SSRI, mirtazapine one case report reported increased risk Not controlled Retrospective cross-sectional study SSRI/SNRI and BDZ 13 newborns (50%) exposed in utero presented with NAS, 9 (69%) were exposed to SSRI/SNRI + BDZ and 4 (31%) were exposed only to SSRI/SNRI (p < 0.05   66/191 (34.6%) in the SSRI/SNRI-exposed group. Most common signs were difficulties in breathing and feeding.

Current Guidelines for the Treatment of Depression in Pregnancy
Many countries have developed guideline recommendations for the treatment of depression during pregnancy (Table 10). All but one of these guidelines advise psychotherapy as initial treatment for mild to moderate depression. Only the American College of Obstetricians and Gynecologists guideline recommends antidepressants as the preferred initial therapy instead of psychotherapy. Three guidelines advise continuing antidepressants with the drug to which the patient has responded prior to pregnancy.

Discussion
The optimal choice for treating anxiety and depressive disorders during pregnancy should, as a first step, start considering the risks correlated with untreated illnesses. Many studies have found that a mental illness can affect a mother's ability to obtain prenatal care and her ability to avoid unhealthy behavior. Moreover, these women have a poor quality of life, and they are more likely to smoke or use alcohol or other substances, which may worsen the pregnancy outcome. For these reasons, patient and health care provider preconception counseling can reduce maternal and neonatal risks by increasing maternal adherence to the antidepressant/anxiolytic treatment. In fact, while antidepressant medications have clear treatment benefits for the general population, in maternity, we must balance the risks of depression in pregnant women with the potential risks of medication passing through the placenta or breast milk to the fetus or newborn. Since ethical considerations preclude prospective, randomized controlled trials that would pose a danger to the fetus or newborn, physicians must rely on observational studies to make treatment decisions in this area.
Concerning untreated depression, many authors have associated it with adverse events in pregnancy, such as spontaneous abortion [189], bleeding during gestation [190], increased uterine artery resistance [191], gestational hypertension and subsequent preeclampsia [192,193], and also in the neonatal period as preterm deliveries [194], spontaneous early labor [195,196], low birth weight in babies [197], low Apgar scores [198], admission to a neonatal care unit [199], perinatal and birth complications and NAS-like symptoms [190,192,200] and neonatal growth retardation [201,202]. Furthermore, the discontinuation of therapy in the first trimester of pregnancy, or even the simple reduction of antidepressant therapy, entails a risk of exacerbation of the affective disorder in the former three months of gestation [203][204][205].

Congenital Malformations
SSRI. Selective serotonin reuptake inhibitors are the most common antidepressants used to treat anxiety and depressive disorders during pregnancy. All SSRIs cross the placenta. Most of the studies did not find an increased total rate of malformations associated with SSRIs as a grouped exposure [17,34,44,51,[61][62][63]. Currently, the scientific data published on the relationship between the use of SSRIs and congenital anomalies suggest a low absolute risk. According to published studies and meta-analysis, sertraline appears to be the safest among SSRIs, while paroxetine seems to be the most reported drug associated with fetal heart malformations. Actually, cardiovascular anomalies are the main defect that could be significantly higher and the reason for growing concerns regarding the use of SSRIs, and paroxetine particularly, during pregnancy. Nevertheless, the current evidence of SSRI teratogenicity may be affected by various methodological weaknesses, lack of investigations using control groups of untreated depression, confounding by indication and recall bias 86,206].
SNRI. The existing literature on the safety of the SNRI medicines during pregnancy suggests that venlafaxine and duloxetine are relatively safe during pregnancy, in particular as far as major malformations are concerned. Bellantuono et al., in their review of 29 case-reports and case-series, suggested a lack of association between venlafaxine and an increased risk of major congenital malformations [207]. The same result was reported by Lassen et al., where the cumulative data of 3186 infants exposed to venlafaxine in pregnancy and 107 major malformations showed that venlafaxine was not associated with an increased risk of major congenital malformations [82].
Eventually, the position paper of the Canadian Pediatric Society in 2021 concludes that women who use an SSRI or SNRI during pregnancy should be counselled that the overall risk for congenital malformations or persistent pulmonary hypertension is low [208]. In conclusion, it remains difficult to determine whether SSRIs or SNRIs are associated with increased risk for teratogenic effects because, while some studies appear to demonstrate an association, others clearly do not. Importantly, however, when an association has been found, the absolute risk was low [69,[79][80][81][82][83][84].
Tricyclic antidepressants. Although this class of antidepressants is much older than the SSRIs, there is a little experience about the potential teratogenicity. The risk of major congenital malformations seems to be not significant in all the epidemiological studies [32,44,65,133]. However, a slightly elevated risk of cardiac defects was identified on the basis of Swedish health databases, which include about 1600 pregnant women who were prescribed TCA in the first trimester [49]; in another register-based cohort study, the risk was not significant [69].
Atypical antidepressants. Almost all the epidemiological studies about the use of atypical antidepressants in pregnancy showed a non-significant risk for congenital malformations, but the number of patients evaluated till this year is very small, so it is difficult a clear assessment of the risk for this class of antidepressants [27,85,[87][88][89][90][91][92]206].
Benzodiazepines. In benzodiazepines exposure, meta-analysis data from 11 cohort studies reported no association between fetal exposure to benzodiazepines and major malformations [66]. The same result was confirmed by an update to this meta-analysis by Enato et al., where the authors did not detect any statistically significant association between benzodiazepine exposure and major malformations [95]. To the same conclusions came a recent meta-analysis [96] and many other cohort studies [53,62,[97][98][99][100][101][102]106]. Only one study reported an increased risk for cardiac malformations at the higher daily dose of BDZ during the first trimester of pregnancy [103].

Pregnancy Complications
Although the main doubt to treat or not treat women with depression is the risk of congenital malformations, there are other drug-induced maternal and fetal or neonatal complications that should be evaluated in this consensus. Gestational hypertension, preeclampsia, gestational diabetes, intrauterine growth retardation, spontaneous abortion, preterm birth and postpartum hemorrhage are topics of major concern.

Antidepressants
Very few studies have focused on the potential obstetrical adverse effects of antidepressant use during pregnancy. Toh et al. reported that gestational hypertension was detected in 9.0% of the 5532 depressed women who were not treated with SSRIs and 19.1% of the 199 women who were treated with SSRIs [107]. Other authors have confirmed the significant risk of gestational hypertension in the SSRI-treated group [109]. On the other hand, Malm et al. found a non-significant risk of gestational hypertension in treatment with SSRI (4.5%) vs. untreated women with a psychiatric diagnosis (4.5%) [68]. It has been suggested that psychological conditions, such as stress, anxiety and depression [110,193], may trigger the pathogenic vascular processes that lead to this condition and need to be balanced against the effect induced by the pharmacological treatment. Some authors have supposed a possible link between pre-eclampsia and the use of SSRIs in pregnancy [108,111]. Other authors did not observe such an association [112], while others found diametrically opposite results, observing that the use of SSRIs in pregnancy was associated with decreased pre-eclampsia after controlling for clinical confounders, such as depression severity, chronic hypertension, diabetes, body mass index and age [113].
Intrauterine growth retardation has been associated with several factors such as maternal medical conditions (antiphospholipid syndrome, gestational hypertension), tobacco and use of substances such as cocaine, alcohol and narcotics, maternal malnutrition, multiple gestation, infections, genetics and placental disorders. Also, the use of antineoplastic medications (e.g., cyclophosphamide), antiepileptic drugs (e.g., valproic acid) and antithrombotic drugs (e.g., warfarin) has been associated with an increased risk of fetal growth restriction. Concerning SSRI use in pregnancy, only a few studies have investigated a possible association with a possible intrauterine growth retardation and the results vary from a very low to a not-at-all risk [115].
In Europe, gestational diabetes mellitus is significant in pregnant women, around 11% of all pregnancies [209]. It has been postulated that all antidepressant exposure in pregnancy may be associated with a moderate risk of gestational diabetes [49], while other authors did not find any statistical evidence [43]. Eventually, recent publications on this topic indicated that there is no association between the use of SSRIs during pregnancy and the onset of gestational diabetes [116][117][118].
Spontaneous abortion is one of the most important adverse pregnancy outcomes, and the potential for maternal antidepressant use to influence the risk of spontaneous abortion has been studied by several authors. Broy et al., as well as Nakhai-Pour et al., reported a possible significant risk only for paroxetine and venlafaxine [120,121], while in a population-based study of more than one million pregnancies among women with a diagnosis of depression, the risk for spontaneous abortion after any antidepressant exposure was not significant [122].
Preterm birth is among the most complex and important challenges in obstetrics. It has been reported that in the United States, the preterm birth rate was 10.2% [210]. Many factors have been associated with preterm birth, including maternal demographics and characteristics, social and economic factors, medical complications, obstetric history and conditions specific to the current pregnancy. Also, the use of SSRIs has previously been found to be associated with an increased risk of preterm birth [43,49,115,126,127], while Lattimore et al., in their meta-analysis, showed a non-significant risk for preterm birth (1.85, 95% CI: 0.79-4.29) [130]. On the other hand, it is not clear whether this is attributable to the underlying depression or to SSRI use [125,126]. In fact, in the study of Vlenterie et al., women with depression were also observed to have an increased and significant risk of preterm birth (OR 2.2, 95% CI 1.7-3.0), highlighting the risks of untreated depression in pregnancy and suggesting low absolute risks of SSRI use during pregnancy [134]. Moreover, the heterogeneity of the studies was high, and no significantly increased risk of preterm birth was observed for SSRI use in the general pregnant female population [211].
About the postpartum hemorrhage, Corbetta-Rastelli et al., in their repeated crosssectional study, analyzed delivery hospitalizations in the United States, and they reported that among an estimated number of 76.7 million delivery hospitalizations, 2.3 million (3.0%) were complicated by postpartum hemorrhage [152]. There are many causes of the maternal hemorrhage, and uterine atony is estimated to cause 70-80% of them. Moreover, a number of well-established risk factors, such as prolonged labor or chorioamnionitis, are associated with postpartum hemorrhage. The SSRI medications may interfere with normal platelet function and hemostasis by inhibiting serotonin reuptake in platelets, and they could be a risk factor as well. Several studies have observed a possible small increase in the risk of postpartum hemorrhage [150], while other studies reported no disproportionate risk of postpartum hemorrhage [153]. Studies reporting an increased risk of postpartum hemorrhage may be confounded by clinical management or other factors unrelated to pregnancy (i.e., treatment with NSAISs or a low dose of aspirin) or unrelated to the drug therapy (i.e., twin pregnancy). Lindqvist et al. did not exclude women with other bleeding disorders and cannot be certain that women taking SSRI continued to use the drug up to delivery [150], while Skalkidou et al. has identified an increased risk also for pregnant women with psychiatric illness but no medication exposure [151].

Benzodiazepines
Concerning benzodiazepines and HBRA exposure during pregnancy, the most adverse outcomes assessed were spontaneous abortion, preterm birth delivery, low birth weight, low Apgar score at 5 min and neonatal intensive care unit admission. Grigoriadis et al., in their meta-analysis of fourteen studies, found that antenatal BZD exposure was significantly associated with an increased risk of all the above gestational complications [105], and other studies reported the same result [99,100]. However, other authors did not confirm any increased risk of low birth weight or small for gestational age [212,213].

Neonatal Complications
The risk for persistent pulmonary hypertension, neonatal adaptation difficulties of the newborn and long-term effects of in utero exposure are the main adverse effects in the offspring of pregnant women treated with antidepressants and/or anxiolytics.

Persistent Pulmonary Hypertension
Persistent pulmonary hypertension has been associated with several conditions, while in the absence of cardiac or pulmonary disease, PPHN is referred to as idiopathic PPHN [214]. Concerning the association between the use of SSRI in pregnancy and PPHN, Chambers et al. were the first authors who reported an absolute risk of PPHN in SSRIexposed infants of 1-2% [157], while the incidence in the general population has been reported as 1.9 per 1000 live births [215]. Subsequent studies have shown a much lower risk of PPHN, ranging from 0.2% to 0.3% absolute risk [154,155,158,216]. Other researchers could not confirm an association of SSRI exposure with PPHN [156]. Moreover, maternal depression, obesity, smoking and surgical delivery are also risk factors for PPHN, a fact that raises the difficulty of the interpretation of these data [161,162,217,218].

Neonatal Abstinence Syndrome
Neonatal abstinence syndrome is a spectrum of clinical manifestations seen in neonates due to withdrawal from intrauterine psychotropic drug exposure. Neonatal symptoms, such as jitteriness, excitability, respiratory distress syndrome and, sporadically, seizures, were observed in newborns exposed in utero to antidepressants, benzodiazepines or opiates. Central nervous system hyperirritability leads to sleep disturbances and difficulty in breastfeeding. These symptoms are usually mild and transient [163,164]. The signs begin in the first days of life and continue in the typical infant for less than 2 weeks and for a month in very rare cases [165]. Paroxetine is generally the SSRI most frequently implicated, followed by fluoxetine [219]. The pathophysiology of this condition remains unknown. Due to different mechanisms of action, NAS from SSRI, TCA, BDZ and opioids may be attributable to a relative hyposerotoninergic state in the synaptic space, to a weak cholinergic transmission, to the reduced BDZ-induced GABAergic response, resulting in an increased glutamate excitatory activity and to an adenylyl cyclase super-activation [220] after tapering SSRI, TCA, BDZ and opioids, respectively. In addition to the receptor-specific differences between the SSRIs, the short half-life of paroxetine may account for more pronounced withdrawal symptoms. Interindividual differences are also possibly caused by genetic polymorphism of the metabolizing enzymes and the 5-HT transporter activity [166,221]. When withdrawal signs are severe, with a Finnegan score above eight, pharmacotherapy should be initiated. Many authors reported a prevalence of NAS of 30-35% to more than 70% of the neonates exposed in utero to SSRIs [135,[167][168][169][170]178,181,222,223]. A systematic review and meta-analysis of observational studies concerning NAS following late in utero exposure to SSRI reported that, depending on the symptoms of NAS, some manifestations could reach a high incidence, up to 77% of the SSRI-exposed newborns [171]. Eventually, other authors found that there were no differences seen between infants whose mothers reported discontinuation of the SSRI prior to the last month of pregnancy compared with infants whose mothers continued SSRI use through delivery [172,173].
Maternal use of benzodiazepines during late pregnancy has been associated with neonatal abstinence syndrome. The case reports of NAS in newborns exposed in utero to benzodiazepines have included descriptions of increased sedation, abnormal muscle tone, poor feeding, sleep problems, poor weight gain, tachypnea, irritability, loose stools, vomiting and vigorous sucking [97,136,148,224]. The incidence of NAS may vary from one to another in epidemiological studies; it has been reported that from 20% to 40% of the neonates with intrauterine exposure to benzodiazepines present symptoms with characteristics of neonatal abstinence syndrome [225,226]. The onset and duration of symptoms correlate with the pharmacokinetics and placental disposition of benzodiazepines [227]. In a large population-based retrospective cohort study, Sanlorenzo et al. reported an increased risk of developing NAS requiring pharmacological treatment in infants exposed antenatally to benzodiazepines [228]. For this reason, some authors suggest tapering the treatment of anxiolytics (or even antidepressants) before delivery in order to minimize the possibility of an appearance of NAS [97]; conversely, other studies conclude that reducing late third-trimester exposure to SSRIs would not substantially reduce the risk of the neonatal adverse outcomes [172].
About the pharmacologic characteristics of the BDZ (e.g., half-life) for a short-period treatment, the most used BDZ are the ones with a short half-life and no active metabolites like lorazepam. In the case of a BDZ with a long half-life and active metabolites like diazepam, it is expected an accumulation of the drug in adipose tissue and a prolonged transfer to the fetal compartment. On the other hand, the use of diazepam and other BDZs with long half-lives is less likely to be associated with NAS [229].
In addition, in newborns prenatally exposed to long-acting benzodiazepines or exposed to benzodiazepines shortly before delivery, it has been described a floppy infant syndrome (FIS) occurs. FIS symptoms described are hypotonia, inactivity, weak cry, lethargy, sucking difficulties, low Apgar score, hypothermia, apnea, cyanosis, hyperbilirubinemia, central nervous system depression that occur mainly within the first hours after labor and last up to 14 days [227].

Long-Term Effects
Concerning the long-term impact of these medications on exposed offspring, there are few data, and this possible effect is still poorly understood [230]. Given the important role of the serotonergic system throughout neurodevelopment and the presence of serotonin transport protein in early brain development, coupled with the ability of SSRIs to cross the placenta, it is reasonable to speculate that SSRIs in utero exposure, by increasing serotonin levels in the offspring, could have a long-term impact on the neurobehavioral outcome. Indeed, the serotonergic system regulates a range of brain development processes, including neuronal proliferation, migration, differentiation and synaptogenesis [231]. However, it should also not be forgotten that exposure to maternal stress-related disorders can also alter the developing serotonergic system in the fetus, as observed in mice [232].
Recent reviews reported the possible effects of perinatal SSRI exposure on social behavior, neurodevelopmental disorders and anxiety. Nevertheless, conflicting findings exist, and such evidence is not conclusive [231,233,234]. Population-based and crosssectional studies showed that maternal depressive symptoms, rather than prenatal SSRI exposure, were related to higher levels of internalizing behavior [235,236] or higher levels of conduct and social problems in children [237]. In a similar study, maternal depressive symptoms without and with SSRIs were both associated with autistic traits in children [137]. However, in a comparative study, children prenatally exposed to venlafaxine, a selective serotonin and norepinephrine reuptake inhibitor, and an SSRI did not differ significantly from the children of mothers with untreated depression in any of the cognitive outcome or behavioral measures [238]. No data are available on other SNRI medications, such as duloxetine and vortioxetine or the atypical antidepressants. Limited and not updated data are available about tricyclic antidepressants. Indeed, these drugs have been largely replaced by SSRIs and SNRIs in depression management, mainly due to their adverse effects and lethality in overdose. Nonetheless, three studies investigating fetal exposure to TCA did not report any effects on the psychological and cognitive infant's development [237,239]. In a large retrospective cohort study of 1,580,629 Swedish offspring using multiple statistical and methodical approaches to adjust for confounding factors, first-trimester antidepressant exposure was not significantly associated with the risk of being born small for gestational age or later autism spectrum disorder or attention-deficit/hyperactivity disorder [138].
Anxiety affects up to 15% of pregnant individuals worldwide and may require pharmacological treatment with anxiolytic agents, such as benzodiazepines. Z-hypnotics are benzodiazepine-related drugs mainly prescribed to treat mild insomnia. Children exposed to benzodiazepines and z-drugs in utero did not show significant neurodevelopmental effects [240]. A large retrospective cohort study suggests that previously described adverse neurodevelopmental outcomes associated with benzodiazepine exposure during pregnancy were likely to be accounted for by maternal genetic confounding [241]. However, due to limited data on the topic, some authors are still currently not certain whether or not prenatal exposure to BZDs and/or z-drugs might be associated with neurodevelopmental outcomes in the offspring [242].

Abuse Substances
There is strong evidence of the association between opioid abuse with anxiety and depression, as well as between depression and current tobacco or cannabis use [243][244][245]. This association is most commonly explained either by a causal relationship or a shared etiologic factor underlying both disorders. Some authors also found that each drug abused followed the onset of depression, except for LSD, which coincided with the onset of depression [246]. There is also a possible pathophysiologic link between the polygenic risk for major depressive disorder and opioid use disorder and opioid system activity, as evidenced by Love et al. [247].
The pharmacotherapy for substance abuse and addiction has been reviewed in several guidelines [248], and in National Institute for Health and Care Excellence [182] as well as in the Cochrane review of improving pregnancy outcomes [249], and reviews of psychosocial interventions and maintenance programs in pregnant women pregnancy exposed to opioids [250]. Treatment in pregnancy is based on a substantial body of evidence on the management of illicit opioid use in the general population, and this treatment has improved maternal and fetal outcomes compared to those seen with untreated opioid abuse [251]. The use of methadone in pregnancy was associated with better antenatal care, reduced maternal morbidity and better neonatal outcomes [252]. There are fewer data regarding buprenorphine in pregnancy, although evidence is increasing [253,254]. In a cohort study of 11,272 pregnant women exposed to buprenorphine and 5056 to methadone in late pregnancy, the risk of adverse maternal outcomes was similar among the two groups and similar to the baseline risk of the general population [255]. Zedler et al., in their meta-analysis, reported that buprenorphine was associated with approximately half the risk of preterm delivery and with greater birth weight (approximately 10 g) than the methadone treatment. No differences were observed for fetal death, congenital anomalies and fetal growth [251].
Regarding the neonatal abstinence syndrome, the risk is high as follows: a Cochrane review found no difference between methadone and buprenorphine in the incidence of neonatal abstinence syndrome requiring drug treatment [256]. In another study, 65.8% of the neonates born from mothers treated with methadone were diagnosed with neonatal opioid withdrawal syndrome requiring pharmacological interventions compared with 38.7% in the buprenorphine/naloxone group [257]. Lastly, the MOTHER study found that infants born to mothers treated with buprenorphine had shorter treatment for NAS, required fewer medications, and had shorter hospital stays [253].
Eventually, there is a possible link between childhood attention-deficit/hyperactivity disorder (ADHD) and substance abuse and depression [258], and pharmacological treatments for psychiatric disorders appear to mitigate the development of substance use disorders [259]. Furthermore, many authors suggest that long-term methylphenidate may reduce depression and suicide risk in ADHD patients [260]. During pregnancy, several studies reported that methylphenidate does not seem to increase the risk for major malformations [261,262], while Koren et al., in their meta-analysis, found a possible increased risk of cardiac malformations [263]. Despite these contrasting data, all authors agree that the discontinuation of methylphenidate therapy may be harmful with re-emergence of symptoms in ADHD patients [264].

Combined Treatment Options: Drug Treatment and Psychotherapy
We found limited evidence for the combination of drug treatment with psychotherapy in pregnant women. A meta-analysis (analyzing among others [65,142,143]) reported increased Preterm Birth risk in depressed women exposed to SSRI compared to psychotherapy alone [131]. Similar results are reported in a cohort study [141]. NICU admission rates do not seem to be increased with the combination treatment compared to psychotherapy alone [141].

Limitations of the Study
These consensus recommendations are comprehensive but not exhaustive. It is possible that we have omitted some interesting topics on which we did not query the expert panel. There is also the possibility of missed studies in the review strategy. Furthermore, some recommendations in this report are based on expert opinion, clinical experience and the extensive literature research, but we found a low number of high-quality studies with large samples. Eventually, the guidelines can be viewed as an expert consultation, to be weighed in conjunction with other information and in the context of each individual patient-physician relationship.

Conclusions
Several guidelines exist in other countries, regarding the treatment of anxiety and depressive disorders in pregnancy. Interestingly, based on the same available data, different guidelines advise different approaches. Our consensus is the first one in Italy, and our results draw a more optimistic behavior of the health practitioner in respect of the patient with depression. First and foremost, an adherence, meaning collaboration between the patient and the healthcare professional that respects the beliefs and wishes of a patient in determining whether, when and how medicines need to be taken should be achieved. Furthermore, this multidisciplinary consensus guide was developed to assist specialists and primary healthcare providers to risk stratifying and managing pregnant women with depression. It provides an overview of issues that may be of importance to healthcare providers involved in the management of anxiety and depressive disorders in pregnancy.
In conclusion, the recommendations engendered were the following: Recommendation 1. Qualified preconception counseling should be offered to all patients of reproductive age affected by anxiety and depressive disorders so that women and their partners can be more aware of the risks of untreated illness and of pharmacological treatment. Anxiety and depressive disorders are not a contraindication for gestation, nor it is a pharmacological treatment. The Panel recommends that treatment should not be stopped abruptly in order to avoid relapse.
Recommendation 2. The Panel recommends continuing the medication to which the patient has responded well before pregnancy or switching to a safer drug if it can be equally effective (Table 11). When there is the need to start an antidepressant during pregnancy, drugs with a more favorable safety profile and more epidemiological data, such as the SSRI, should be preferred and prescribed at the lowest effective dose. Overall, most studies have not found an increased overall rate of malformations associated with SSRI exposure as a group. Moreover, studies that controlled for indication did not find an association with an increased rate of major or cardiac malformations. We can then conclude that it is the underlying condition and not the therapy indicated for it that is associated with an increased risk of MM or CM. Currently, the scientific data published on the relationship between the use of SSRIs and congenital anomalies suggest a low absolute risk. According to current studies and meta-analyses, sertraline appears to be the safest among SSRIs. The Panel recommends continuing the medication to which the patient has responded before pregnancy or switching to a safer drug if it can be equally effective.
Recommendation 3. For the treatment of anxiety symptoms and short-term treatment of sleep disturbances, benzodiazepines can be administered during pregnancy. In order to minimize the pharmacological effects of the drug on the fetus, it is preferable to use drugs with a short elimination half-life, such as lorazepam, oxazepam and brotizolam.
The Panel concluded that benzodiazepines can be used during pregnancy and recommends using short half-life drugs.
Recommendation 4. In opioid-dependent pregnant women, the use of methadone and buprenorphine is associated with better maternal and fetal outcomes compared with uncontrolled opioid misuse. There are no recommended pharmacological interventions for the abuse of psychoactive stimulants, cannabinoids or the new psychoactive substances in pregnancy. The Panel recommends that opioid misuse treatment should not be stopped.
Recommendation 5. As for all drug use during pregnancy, possible fetal risk cannot be excluded; the risks are limited but not zero. Patients with depression who require standard treatments to control their disease must be followed up closely by an obstetrician once the pregnancy is confirmed (Table 11). The Panel recommends that cooperation between specialists (e.g., psychiatrists, obstetrics, toxicologist) should be encouraged. Whenever possible, every attempt should be made to offer multidisciplinary care to the patient, involving gynecologists and psychiatrists experienced in high-risk pregnancies. Coordination with the general practitioner is also recommended. The Panel recommends offering multidisciplinary care when possible. Recommendation 6. Neonatal abstinence syndrome is common, although the symptoms are mild and usually resolve quickly. The patient should be prepared to face this possible event with the utmost serenity (Table 11). The Panel does not agree with previous suggestions to stop or gradually taper the use of antidepressants before delivery in order to minimize the risk of NAS, as this may be associated with a predictable relapse of the precedent-treated depression during the postpartum. The Panel recommends continuing pharmacological treatment till delivery.

Ethical Considerations
When facing a woman in need of advice on the management of depression or anxiety disorders, consider informing her clearly and with informative materials (such as official leaflets or scientific papers). In many countries, there is no strict need to make an informed consent, but doing so can be useful to ensure that a shared decision is made in the most informed way. Regular follow-ups should be offered, and it is preferable to ensure a multi-disciplinary approach. The evidence clearly suggests that treating underlying conditions during pregnancy is the most recommended choice. Drug treatment is mostly related to an enhanced risk of neonatal complications, but these complications usually resolve quickly and without sequelae. Women in need of advice should be informed clearly in order to ensure shared decision making. Regular multidisciplinary follow-up is recommended.

Future Research
More data are needed to clearly address indication bias. Most of the studies found do not control for underlying conditions such as depression or anxiety disorders. When this was considered, data were clearly in favor of treatment when considering malformation risk. More data are needed to clearly address the risk of complications for women during pregnancy and at the time of delivery. Limited data were found if not for spontaneous abortion and preterm birth risk, and even in these two cases, results are not clear even when indication bias was taken into account. Future research in the field should more clearly address long-term safety for children exposed to antidepressants or anxiolytics in utero, especially concerning neurodevelopmental outcomes. Table 11. Consensus panel recommendation. Colors are used to highlight risk of side effects for drug exposure both for pregnant women and for fetus.

Possible Complication to Address Recommendation Based on Evidence of Risk for Pharmacological Treatment
Major Malformation or Cardiac Malformation Treat underling condition, risk not increased.