Reproductive Decision-Making Capacity in Women with Psychiatric Illness: A Systematic Review

Abstract

Objective:

This paper aims to characterize existing literature about capacity evaluations in women with psychiatric illness making reproductive decisions.

Methods:

We conducted a systematic review searching PubMed, EMBASE, Ovid PsycINFO, and ClinicalTrials.gov databases through July 2020. Articles were evaluated for relevance to inclusion criteria, first by title and abstract screening then by reading the full text of articles. Our inclusion criteria were patient reports and studies that involved women of childbearing age with psychiatric illness making obstetrical decisions for whom decision-making capacity was evaluated. We qualitatively analyzed our data by examining themes within the studies, such as the circumstances of the referral and characteristics of reproductive decision-making. We also collected information about the clinical circumstances, such as the clinical setting (e.g. inpatient or outpatient, in primarily psychiatric or obstetric care) and who made the determination of capacity.

Results:

We identified 18 articles, which included 22 distinct patient cases and 27 distinct obstetrical decisions. Decisions about termination of pregnancy were most common, which were 10 out of 27 decisions. Decisions about timing and mode of delivery accounted for seven and five decisions respectively. The most common psychiatric diagnosis reported was schizophrenia, which was present in eight patients. Major depression and bipolar disorder were also frequently reported, present in three and five patients, respectively.

Conclusion:

Patients who make healthcare decisions must have decision-making capacity. A patient with psychiatric illness does not inherently lack capacity and in fact most patients with psychiatric illness have decision-making capacity. Psychiatric illness, however, can add complexity to obstetrical medical-decision making. Reasons for this complexity include the involvement of many stakeholders, the often time-sensitive nature of these decisions, and the potentially unpredictable course of psychiatric illness. Successful management of these cases involves multidisciplinary collaboration, the use of preventative ethics through advanced care planning, and identification of a guardian or surrogate decision maker when a patient is determined to lack decision-making capacity or for a patient who has capacity through advanced care planning.

PROSPERO Registration Number:

CRD42020143434

Introduction

Medical decision-making can be complicated by psychiatric illness. Compared to women without psychiatric illness, women with psychiatric illness are at a higher risk of poor compliance with contraceptive methods, risky sexual behaviors, coercive sexual encounters, unplanned and unwanted pregnancies, violence and victimization during pregnancy, substance use, medical comorbidities, and pregnancy complications.^1–5 Pregnancy itself can worsen psychiatric symptoms for many reasons, including hormonal and physiologic changes that can affect medication efficacy as well as inherently increase women’s vulnerability to psychiatric illness.^6,7 Many women discontinue their psychiatric medications upon learning they are pregnant, which often worsens psychiatric illness.^8,9 Many clinicians also advise their patients to stop taking or to lower their psychiatric medications upon learning that a woman is pregnant, due to misperceptions about medication safety in pregnancy.¹⁰ Poorly treated or undertreated psychiatric illness can impair a patient’s ability to provide informed consent by impairing an individual’s ability to understand the components of a medical decision.

Physicians legally and ethically must obtain informed consent from their patients.^11,12 Informed consent consists of three components: adequate information, voluntariness, and decision-making capacity.¹¹ Adequate information means that the physician should provide patients with enough information about the recommend treatment and risks of this treatment, treatment alternatives and their risks, and the risks of no treatment, so that patients can make informed decisions.^11,12 Voluntariness means that the decision must be made without coercion, or excessive manipulation or influence from external sources.^11,12 Finally, the patient must have decision-making capacity, or simply “capacity.”¹¹ Capacity is the ability to understand and reason through a decision and its consequences.^11,12 Although at times capacity is used interchangeably with competency, competence refers to a legal determination of a patient’s global decision-making ability.^11,12 In contrast, capacity is clinician-determined and refers to the ability to make a specific decision rather than decision-making ability in general.^11,12

The presence of psychiatric illness does not mean a patient lacks decision-making capacity. In fact, most patients with psychiatric illness retain decision-making capacity.¹³ However, psychiatric illness can impact decision-making capacity when the symptoms of psychiatric illness impair an individual’s ability to understand why a medical decision is being proposed. For example, a patient experiencing paranoid delusions may feel paranoid about his or her physicians and may decline a medical recommendation because of this delusion, not because of a rational reason for declining this recommendation. Psychiatric illness that occurs during pregnancy can complicate obstetrical decision-making by potentially impairing capacity. To our knowledge, there has been no published review of decision-making capacity in women with psychiatric illness who are making reproductive decisions, which we define as decisions about pregnancy and future reproductive capacity (such as decisions about birth control). We sought to evaluate the literature for reports of patients with psychiatric illness making reproductive decisions for whom capacity was evaluated. Our goals were to characterize these cases and their challenges, with the ultimate goal of developing guidelines for physicians who treat these patients.

Materials and Methods

The PRISMA system was used to develop and implement the systematic review. The search was conducted through July 2020. The review protocol was registered prospectively in PROSPERO. The following electronic databases were searched: PubMed, EMBASE, Ovid PsycINFO, and ClinicalTrials.gov. Duplicates were removed. Search terms were developed using MeSH, EMTREE, the Thesaurus of Psychological Index Terms, the Diagnostic and Statistical Manual of Mental Disorders Version 5, and keywords for the concepts of decision-making capacity, pregnant women, and psychiatric conditions for use in the PubMed query (see Table 1). Search strings for each database were developed by a health sciences librarian, using controlled vocabulary (Medical Subject Headings, Emtree terms, Thesaurus of Psychological Index Terms) and natural language words or phrases to reflect the concepts of “pregnant women,” “psychiatric conditions,” and “decision making.” The PubMed search string is shown in (Table 1). We then iteratively tested different combinations of the search terms and made a decision to include or exclude a term depending upon its utility in capturing relevant articles. Terms that retrieved a large set of irrelevant records were excluded from the final search. This process was repeated until all search terms in the search pool were tested. These terms were then adapted for the EMBASE, PsycINFO and ClinicalTrials.gov queries. Searches excluded non-English language articles and animal studies. We did not limit studies by date range and considered any relevant article published at the time of our search.

Table 1:

Search Terms for PubMed Query

Psychiatric	ADHD or adjustment disorder or agoraphobia or alcohol or alcoholism or alcoholics or alcoholic or amphetamine or anorexia or anorexic or anorexics or antisocial or anxiety disorders or Asperger or attention deficit or autism or autistic or binge drink or binge eating or bipolar disorder or borderline personality or bulimia or Capgras or cocaine or compulsive or conduct disorder or depressed or depression or depressive or developmental disabilities or feeding and eating disorders or female athlete triad syndrome or learning disorders or morphine or neonatal abstinence or neurodevelopmental disorders or neurotic or opioids or opioid or substance abuse or panic or paranoid or personality disorders or phobic or post-traumatic or psychological trauma or psychotic or psychotic or schizophrenia or schizophrenia spectrum and other psychotic disorder or substance-related disorders or tobacco or traumatic stress or trauma and stressor related disorders
Reproductive	Abortion or abortion, induced or birth or births or birthing or contraception or fertilization or maternal-fetal relations or parity or preconception or pre-conception or pregnancies or pregnancy or pregnant or pregnant women or prenatal or prenatal care or reproductive physiological phenomena or stillbirth or sterilization
Medicolegal	Choice behavior or coercion or competence or consent or decide or decision or decision making or informed consent or involuntary or mental capacity or legal capacity or refusal to participate

Two authors independently performed the review process, in order to reduce the risk of bias in the selection process and create redundancy to ensure relevant articles were not inadvertently excluded. These authors first reviewed all titles and abstracts independently for potential relevance to inclusion criteria. Disagreements were resolved by discussion. After this screening process, both authors read all full text of articles included in the next phase to determine if these full articles met inclusion criteria.

Inclusion and exclusion criteria were defined and reviewed by additional co-authors prior to this review process. We included any study that discussed actual patient scenarios in case reports or higher levels of evidence. These studies had to involve adult patients with psychiatric illnesses who were making obstetrical decisions and for whom capacity was evaluated. With regard to the obstetrical decisions, we included pre-pregnancy decisions such as contraception, both reversible and irreversible forms. We also included decisions made during pregnancy about obstetrical care. We did not include post-pregnancy decisions, such as breast-feeding, meconium sampling, and child custody. We also did not include general medical issues that were treated during pregnancy, given our research aim of exploring specifically reproductive decisions. We excluded studies involving child or adolescent patients, because individuals under the age of eighteen who make obstetrical decisions have very different ethical and legal considerations than adults. We excluded articles where the legal system determined an individual to be incompetent, because this determination is very different from the physician determination of capacity to make a single medical decision.

We reviewed these full text articles that met inclusion criteria and extracted our data from these articles. If this decision involved pregnancy management, we collected information about the gestational date when capacity was evaluated. For contraception decisions, we extracted information about the type of contraception, differentiating reversible methods such as oral contraceptive pills and intrauterine devices from irreversible methods such as tubal ligation or other forms of sterilization. We collected information about the patients’ primary psychiatric diagnoses as per the Diagnostic and Statistical Manual of Mental Disorders, 5th Edition (DSM-5.¹⁴ If a DSM-5 diagnosis was not specified or the study used a diagnosis present in an earlier version of the DSM, we referred to it as the most relevant DSM-5 diagnosis based on available criteria. We included diagnoses of Intellectual Disability because of its potential impact on decision-making capacity, despite the fact that Intellectual Disability may have many etiologies including medical. We also included Substance Use Disorders, because these are DSM-5 diagnoses that could impact capacity and are a frequent comorbidity with serious mental illness.

Results

We identified distinct 8345 articles from our initial query (see Figure 1) and reviewed their titles and abstracts for relevance to inclusion criteria. We then reviewed the full text for 62 articles, identifying 19 articles that met our inclusion criteria.^15–30 Two of these articles discussed the same patient scenario, and therefore they were considered as a single data point.^28,29 This generated a total of 18 articles, 22 distinct patients, and 27 distinct obstetrical decisions for our final analysis.^15–30 A majority of these articles were from patients in the United States but several were from outside of the United States (Table 2). We qualitatively examined this data identifying themes within identified studies. A qualitative analysis was necessary due to the nature of our findings, which were case studies and case series.

Figure 1 –

Flowchart summarizing literature search and selection process

Table 2:

Included Cases and Their Characteristics

Obstetrical Decision	First Author (Year)	Primary Psychiatric Diagnoses	Gestational Age	Evaluation Setting	Specialty of Capacity Evaluator	Did Patient Have Capacity?	Decision Maker	Country
IVF *
	Doghor (2017)	OCD†	NA‡	Outpatient – Psychiatry	Psychiatry	Yes	Patient	United States
Contraception Reversible
	Acera Pozzi (2014)	MDD§; OCD	NA	Inpatient – Psychiatry	Multidisciplinary team	Yes	Patient	United States
Irreversible	O’Hara (1989)	Intellectual disability	NA	Inpatient – Psychiatry	Not specified	No	Unspecified relatives	United States
Termination
	Altinoz (2018)	Bipolar I disorder, most recent episode manic; Borderline intellectual functioning	19 weeks	Outpatient - Psychiatry	Psychiatry	Yes	Patient	Turkey
	Altinoz (2018)	MDD with psychotic features	9 weeks	Outpatient – OB/GYN	Psychiatry	No	Spouse	Turkey
	Brody (2016)	MDD	23 weeks	Inpatient – Psychiatry	Psychiatry (2)	Initially no, capacity restored	Patient	United States
	Cohen (1993)	Benzodiazepine intoxication; unspecified mania	20 weeks	Inpatient – Psychiatry	Psychiatry	Yes	Patient	United States
	Coverdale (2004)	Schizophrenia	10 weeks	Inpatient – Psychiatry	Not specified	No	Unspecified surrogate	United States
	de Nesnera (2000)	Schizophrenia	17 weeks	Inpatient – Psychiatry	Psychiatry	Yes		United States
	Desai (2009)	Schizophrenia	2nd trimester	Outpatient – Psychiatry	Psychiatry	Initially no, capacity restored	Patient	India
	Dudzinski (2004, 2006)	Schizophrenia; Substance use disorder (multiple)	14 weeks	Outpatient – OB/GYN	Psychiatry	No	Father	United States
	Mahowald (1985)	Schizophrenia	12 weeks	Inpatient – Psychiatry	Psychiatry	No	Mother	United States
	O’Hara (1989)	Intellectual disability	24 weeks	Inpatient – Psychiatry	Multidisciplinary team	No	Unspecified relatives	England
Unspecified Reproductive Decision-making
	Nau (2011)	Schizophrenia	Throughout pregnancy	Inpatient – Psychiatry	Psychiatry	No	Legal system	United States
	Slayton (1981)	Unspecified schizophrenia spectrum disorder	38 weeks	Inpatient – Psychiatry	Psychiatry	No	Not specified	United States
Delivery Method
	Acera Pozzi (2014)	MDD; OCD	27–35 weeks	Inpatient – OB/GYN	Psychiatry	Initially no, capacity restored	Patient	United States
	Chaudhry (2016)	Bipolar I disorder, most recent episode manic	36 weeks	Inpatient – Psychiatry	Multidisciplinary Team	No	Sister (father of baby declined)	United States
	Chaudhry (2016)	Bipolar I disorder, most recent episode manic	38 weeks	Inpatient – Psychiatry	Multidisciplinary Team	Yes	Patient	United States
	Chaudhry (2016)	Bipolar I disorder, most recent episode depressed	39 weeks	Inpatient – Psychiatry	Multidisciplinary Team	Yes	Patient	United States
	O’Hara (1989)	Intellectual disability; Trichotillomania	38 weeks	Inpatient – Psychiatry	Not specified	No	Unspecified relatives	England
	Spada (2019)	Anorexia nervosa, restricting type	32 weeks	Inpatient – OB/GYN	Psychiatry	Initially no, capacity restored	Patient	United States
	Teng (2016)	Borderline personality disorder	31 weeks	Inpatient – OB/GYN	Psychiatry	Yes	Patient	Singapore
Timing
	Acera Pozzi (2014)	MDD; OCD	27–35 weeks	Inpatient – OB/GYN	Psychiatry	Initially no, capacity restored	Patient	United States
	Spada (2019)	Anxiety; Posttraumatic stress disorder	38 weeks	Inpatient – OB/GYN	Psychiatry	Yes	Patient	United States
	Spada (2019)	Anorexia nervosa, restricting type	32 weeks	Inpatient – OB/GYN	Psychiatry	Initially no, capacity restored	Patient	United States
	Spada (2019)	Schizophrenia	37–41 weeks	Inpatient – OB/GYN	Psychiatry	No	No identifiable surrogate	United States
	Teng (2016)	Borderline personality disorder	31 weeks	Inpatient – OB/GYN	Psychiatry	Yes	Patient	Singapore

^*IVF = in vitro fertilization

^†OCD = obsessive compulsive disorder

^‡NA = not applicable (patient not pregnant)

^§MDD = major depressive disorder

The most common type of obstetrical decision reported in the literature was pregnancy termination (see Table 2), which accounted for 10 decisions. Mode of delivery and timing of delivery accounted for the second most common decisions, occurring seven and five times respectively. Far fewer decisions were about contraception. Only one case report evaluated the decision to have a reversible contraceptive method, specifically an intrauterine device.¹⁵ Similarly only one case evaluated the decision to undergo non-reversible contraception, with the specific method not specified.¹⁸ One case report evaluated capacity to decide to pursue in vitro fertilization (IVF). ²⁷ Finally, two of these publications evaluated capacity to receive general obstetric care without more specific obstetric decisions discussed.^17,19

The patients in these articles had a variety of psychiatric diagnoses, including schizophrenia, mood disorders, substance use disorders, and anxiety disorders (see Table 2). Schizophrenia and related disorders were the most common psychiatric diagnoses, present in eight patients.^{17,19,20,25,26,28–31} An additional case report featured psychosis secondary to major depressive disorder (MDD).¹⁶ Mood disorders were also commonly observed. MDD was diagnosed in three patients and bipolar disorder in five patients.^{15,16,22–24} Only two case reports involved substance use disorders.^22,28,29 Two publications involved intellectual disability.^16,18 Two patients had obsessive compulsive disorder (OCD).^15,27 The other psychiatric diagnoses observed were posttraumatic stress disorder (PTSD), personality disorder (borderline personality disorder), and an eating disorder (anorexia nervosa), present in one patient each.^20,21

The number of decisions where patients were evaluated to have capacity varied across psychiatric diagnoses (see Table 2). Of the decisions made by patients with schizophrenia, the majority of patients lacked capacity to make the decision.^{17,19,20,25,26,28–31} In only one case in the literature did a patient with schizophrenia have capacity to make a pregnancy termination decision.²⁵ Only one patient with schizophrenia had capacity restored in order to make a decision about termination.²⁶ In this case reported by Desai et al, her mental illness was treated for about a month, at which point the patient decided to not terminate the pregnancy.²⁶ In the majority of cases including decisions about termination, the patients with schizophrenia were determined to lack capacity.¹⁵

In cases where patients lacked capacity and capacity was not restored in time to make a decision, the surrogate decision-maker was most often a family member, including a spouse, parent, or other family members. ^{16,18,23,28–30} Only one case specifically noted that the team tried but could not identify a surrogate.²⁰ The decision-maker in this case was not clearly specified but appears to have been members of the treatment team.²⁰ Predominantly, capacity evaluations occurred in inpatient settings, either inpatient psychiatry or obstetrical units. Only five decisions were made in outpatient settings and generally capacity evaluations were made by physicians and treatment teams who had more longitudinal relationships with these patients compared to patients seen in inpatient settings.^15–30

Discussion

As discussed earlier, physicians routinely obtain informed consent, and thus should routinely evaluate capacity to provide this informed consent. The cases that are published and thus are in our review are a subset of particularly challenging cases. Since capacity is routinely evaluated, the more complex cases are more likely to be material for a case result and therefore the cases observed are biased towards a higher level of complexity than the average case.

Challenges of these cases include decisions that prompted strong emotions, the involvement of multiple stakeholders, time-sensitivity, and variable and unpredictable manifestations of psychiatric illness. It is unsurprising that the majority of decisions in our review involved decisions that are not reversible, such as irreversible sterilization, termination of pregnancy, and mode and timing of delivery. The consequences of these decisions are longstanding and so it is reasonable capacity may be questioned more often.

The most common reproductive decision was termination. Termination can be a divisive decision, prompting strong emotions as well as ethical and legal debate. The frequency of termination decisions noted in our study may reflect that physicians possibly have a lower threshold for the evaluation of the capacity of a patient requesting termination. In our review, several studies noted involved people’s discomfort with pregnancy termination decisions, including psychiatry nursing staff, psychiatrists, and surrogate decision-makers.^18,25,28,29 For example, the case by De Nesnera et al featured a 43-year-old woman with schizophrenia determined to have capacity to terminate her pregnancy. This decision was made by evaluating the capacity of the patient, as well as discussion with family, consultation with the hospital ethics team, and meetings with hospital administrators and attorneys all of whom were supportive of the patient’s decision.²⁵ The authors noted that the decision itself prompted strong reactions from the treatment team, including empathy, discomfort, and anger.²⁵ The authors discussed that their own discomfort likely stemmed from their personal beliefs about pregnancy termination.²⁵ Despite this discomfort, they supported the patient’s decision and the patient terminated the pregnancy.²⁵

Many of these cases involved time-sensitive decisions. The cases involving termination often referred to and were impacted by legal constraints on termination timing. Several patient cases involved patients who lacked capacity to terminate their pregancies.^{16,18,23,28,29,31} In two of these cases, an acceptable surrogate decision-maker could not be identified in time to legally terminate the pregnancy and so both pregnancies were continued.^16,18 Decisions about delivery method and timing, such as about induction of labor, are also time-sensitive and delaying these decisions can have medical and psychiatric repercussions.²⁰ Not making a decision can be a decision itself, particularly when the issue at-hand is time-sensitive.

Complicating decision-making was the challenge of identifying a surrogate for patients who lacked capacity. Several pregnant patients in our sample either did not know or were not in contact with the biological father.^{18,20,28–30} In many of these studies, surrogate decision-makers did not have close relationships with the patient, either due to the lack of closer relatives, closer relatives declining to serve as decision-maker, or strained relationships.^23,24,28,29 The lack of a clearly identifiable surrogate decision-maker can delay decision-making.³² A surrogate who does not know the patient well may not know the patient’s preferences, necessitating the use of the best interests rather than substituted judgment standard.³² Finally, an identified surrogate may not be an appropriate decision-maker.^32,33 Surrogate decision-makers may themselves lack decision-making capacity or may have ulterior motives for making a certain decision.³³ One case report was of a woman with schizophrenia who was 12-weeks pregnant who expressed her desire to continue her pregnancy but presumably lacked capacity to make this decision. ³⁰ Her mother obtained legal guardianship and requested to terminate the patient’s pregnancy.³⁰ The final outcome was not described by the authors, but the case study noted that that the hospital’s attorney planned to challenge the decision made by the patient’s mother.³⁰ State guidelines vary on who should evaluate the appropriateness of an identified surrogate but in several states this can be done by clinicians.³³

Most cases required multidisciplinary approaches. The evaluation and management of most cases in our study involved obstetricians, psychiatrists, other physicians, nursing staff, social work, hospital administration, legal counsel, and ethics consultants. This multidisciplinary treatment was likely necessary for many reasons, including the complex and changing psychiatric and obstetrical conditions and the need for coordination of care across treatment settings.^{15,17–23,25,30,31} Psychiatrists rather than obstetricians often performed the actual capacity evaluations, but these evaluations required education from the obstetrical teams about the risks and benefits of the decisions themselves. Most cases took place in acute settings, including inpatient obstetrical and psychiatric units, which inherently involve more complex obstetrical and/or psychiatric issues.^{15,17–23,25,30,31}

Another element adding complexity to these cases is the varying manifestations and courses of psychiatric illness, for example, due to varying rates of treatment responsivity.^{17,19,20,25,28,29,31} Of note, the majority of the patient cases we identified were psychiatrically hospitalized and as such more psychiatrically ill than in a community sample. As noted previously, most patients with psychiatric illness have decision-making capacity.¹³ However, patients who require psychiatric hospitalization are more unwell and more likely to lack capacity. One patient with psychosis due to schizophrenia had capacity to make the obstetrical decision on initial evaluation.²⁵ Only one patient with psychosis successfully had her capacity restored in order to make the decision.²⁶ The majority of patients with psychotic symptoms lacked decision-making capacity when initially evaluated and were not reported as regaining capacity.^{17,19,20,28,29,31} These cases included psychosis due to schizophrenia as well as major depression with psychotic features and psychosis in conjunction with substance use.^{17,19,20,28,29,31} However, they should not be taken as a representative sample of women with psychosis, because no population studies were found, only case reports and case series. The variable courses of psychiatric illness can make restoration of capacity complicated, particularly if this restoration must occur in a limited time frame.

Many of these case studies advocated for incorporation of preventative ethics, working with the patient and her support system to identify potential ethical challenges that may arise and planning for these scenarios.^15,29 As discussed earlier, one way to do this is by obstetrical and psychiatric treatment teams discussing anticipated decisions with patients and their families. The case reported by Acera Pozzi et al presents a 37-year-old woman with OCD and MDD admitted to an intensive care unit (ICU) after a suicide attempt and ultimately transferred to an inpatient psychiatric hospital.¹⁵ In the ICU, the patient initially lacked capacity to decide on delivery method and timing, but capacity was restored through psychiatric treatment.¹⁵ After capacity was restored and prior to the patient’s delivery, the patient and her psychiatric and obstetrical teams discussed the patient’s preferences and formulated a plan to guide delivery decision-making in the event the patient’s capacity was impaired at the time of delivery.¹⁵ The process of developing anticipatory guidance when the patient has capacity can help preserve patient autonomy as well as enact the patient’s wishes should she someday lack capacity.^15,34

Limitations to our study include the restriction of reports to English language only and to medical literature. A majority of articles were regarding patients in the United States (Table 2) although several cases were from outside of the United States. Although we did not exclude articles that discussed gender diverse persons, we did not specifically search for terms relevant to gender diverse persons and we do note that our findings are limited to biologically female individuals. We did not search legal literature or include articles where capacity was not determined by a medical professional, given our desire to have these findings most directly relevant to physicians’ clinical practice. We also limited our findings to reproductive decisions, because of the particular complexity that these decisions have in concert with patients with psychiatric illness. Expanding the scope of decisions to any medical decisions in the perinatal period made by patients with psychiatric illness could potentially shed further light on capacity evaluation within this particular patient population. We included a large number of search terms across psychiatric, reproductive, and medicolegal realms, which were identified and tested for relevance in a systematic way, we did not include every possible DSM-5 diagnosis and we did include several additional diagnoses not found in the DSM-5. Though we included all articles meeting our search criteria, no population studies were found, only case reports and case series. The lack of higher level evidence indicates a need for further investigation in this area. The qualitative nature of our data necessitated a qualitative, descriptive analysis rather than a quantitative analysis. Finally, we noted that the patients cases in our review were more complex than the average patient with psychiatric illness. This reflects publication bias, namely that more complex patient scenarios are more likely to be published.

A primary strength of our study is that this study is the first, to our knowledge, systematic review of the evaluation of capacity in patients with psychiatric illness making reproductive decisions. The themes we identified in our study present multiple potential clinical implications, including the importance of preventative ethics and of multidisciplinary collaboration. Preventative ethics are crucial to incorporate into clinical care when possible. As much as possible, physicians should incorporate preventative ethics by discussing potential obstetrical decisions with reproductive-age women and their support systems. This discussion can occur any time the patient has capacity and can be an ongoing one that occurs on a regular basis as preferences and medical conditions change.³⁴ This planning process can help physicians and potential surrogate decision-makers understand the patients’ preferences.³⁴ Should a patient lack capacity in the future, this discussion can help potential surrogate decision-makers feel confident in their decisions. Furthermore, having this discussion with a patient without a clearly identifiable surrogate decision-maker could help facilitate the identification of a surrogate decision-maker and avoid this task falling to someone who does not know the patient well, such as a distant relative or someone appointed by the legal system.^34,35 These discussions may be more conducive to outpatient settings where psychiatrists are able to have more longitudinal patient relationships. However, psychiatrists if at all possible should look out for the opportunity to have these discussions in advance of potential loss of capacity.

Finally, this review emphasizes the importance of a multidisciplinary approach. A hallmark of Consultation and Liaison Psychiatry is multidisciplinary collaboration, but perhaps this collaborative nature is particularly crucial with time-sensitive obstetric decisions particularly with patients with higher burden of psychiatric illness. Open dialogue and collaboration between obstetric and psychiatric teams as well as through other sources such as legal counsel, hospital administration, and ethics committees played a crucial role in many of these cases. Often, this communication occurred throughout the course of the decision-making. These cases often required complicated decision-making and management that required a wide breadth of expertise, one that is the product of partnership between many healthcare professionals. We recommend that physicians involved in the care of psychiatric patients, particularly patients with more acute and severe psychiatric illness, involve their colleagues in other relevant discipline in order to provide multidisciplinary, comprehensive care and evaluate for any capacity issue in a timely manner.

Conclusions

Our review highlights the challenges of reproductive decision-making in women with psychiatric issues, particularly psychosis. The cases described in this systematic review are more challenging than the average clinical case. Challenges included decisions prompted strong emotions, the involvement of multiple stakeholders, time-sensitive decisions, and the variable treatment presentation and treatment course of psychiatric illness. As discussed previously, these patients have a greater burden of psychiatric illness than others and as such reflect more complex treatment scenarios.