MATERNAL CUSTODY STATUS AND LIVING ARRANGEMENTS OF CHILDREN OF WOMEN WITH SEVERE MENTAL ILLNESS

Historically, psychiatric treatment of women patients in public hospitals was consistent with Test and Berlin's (1981) assertion that mental health professionals tend to regard individuals with severe or serious mental illness (SMI) as "almost genderless" (p.136). Hospital administrators and staff denied or ignored the sexuality of the women patients and quietly managed the births of forbidden children by placing them in foster homes or arranging for their adoption (Apfel & Handel, 1993; Sands, 1995).

The transformation of mental health care delivery from public psychiatric hospitals to community-based services and advancements in psychopharmacology, have increased opportunities for women with SMI to pursue normal life developmental tasks such as sexual relationships, childbirth, and parenthood (Miller, 1997; Nicholson & Blanch, 1994; Oyserman, Mowbray, & Zemencuk, 1994). In fact, fertility rates for women with a severe mental illness are estimated to equal those of women in the general population (Apfel & Handel, 1993; Miller; Nimgaonkar, Ward, Agarde, Weston, ScGanguli, 1997; Saugstad, 1989).

The changing times challenge social work practitioners to increase their understanding of women with SMI who are parents. This article, based on data from a pilot study of 20 women with SMI who have at least one biological minor child, aspires to contribute to knowledge about patterns of care for the offspring of SMI mothers. Echoing earlier strains of ignorance and denial about the sexuality of women with a persistent mental disability, today's mothers with SMI frequently receive little or no support from community mental health providers for pregnancy, postpartum, and parenting concerns (Mowbray, Oyserman, Zemencuk, & Ross, 1995; Oyserman et al., 1994). This lack of assistance is crucial because mothers with SMI are also likely to be single parents, to experience poverty, and to be victimized (Lam & Rosenheck, 1998; Mowbray et al., 2000; Mowbray, Oyserman, & Ross, 1995; Oyserman et al.).

Given that the psychiatric community has been late in incorporating parenting as a mental health issue, the welfare system has been the default social services resource for mothers with SMI and their children when a crisis occurs. (Blanch, Nicholson, & Purcell, 1994; Nicholson & Blanch, 1994; Nicholson, Geller, & Fisher, 1996). Thus, the children of mothers with SMI often enter emergency foster placements when a psychiatric episode requires the mother's hospitalization or when there are suspicions of child neglect or abuse (Blanch et al., Nicholson et al., 1996).

Because of this breach in mental health services, women with SMI frequently lose custody of, and in some cases all contact with, some or all of their children (Nicholson, Geller, Fisher, & Dion, 1993; Nicholson et al., 1996). This pattern commands even more attention from social workers because of nationwide changes in child welfare administration ushered in by the Adoption and Safe Families Act of 1997 (RL. 105-89). By mandating that states exert greater efforts on adoption for children in foster placement, this law, in many cases, would expedite the filing of actions to terminate parental rights (Minuchin, Colapinto, & Minuchin, 1998). As the research discussed in this article demonstrates, children of mothers with SMI, like other children, live in a wide range of settings. The indepth mapping of the 20 separate families about which this article reports offers a vivid illustration of the complexity of the family constellations and suggests important implications for social work practice with mothers with SMI.

The increasing recognition of the needs and challenges of mothers with SMI and the context created by changes in the child welfare laws suggest that practitioners working with women with SMI, their children, or caregivers for the children could benefit from a better understanding of the size and shape of these families.

MATERNAL CHILDREARING ARRANGEMENTS AND CHILD CUSTODY STATUS

Although the married couple-headed family is the predominant type of U.S. household (Fields & Casper, 2001), mother-headed single-parent households have been increasing and are the most prevalent family type among African American families (Sands & Nuccio, 1989; Taylor, Tucker, Chatters, & Jayakody, 1997). Single-parent families encompass a variety of household living arrangements (Aquilino, 1996; Dressier & HaworthHoeppner, 1996; Manning & Smock, 1997), including those headed by single fathers, grandparents, and other relatives.

When neither parent can provide full-time nurturing and protection for children, other significant adults may step in. The term kinship refers to an adaptation of family relations in which close relatives or nonrelatives are treated as significant family members; hence, kin are relatives by blood or marriage and any other person with close ties to the biological family (Hegar & Scannapieco, 1995). Kinship care may be negotiated informally between the parent or parents and relatives or formally by child protection agencies. Regardless of how the arrangements are made, maternal grandmothers and aunts are the principal kin who are primary caregivers of children (Hegar & Scannapieco).

If no relative or emotionally equivalent parent substitute is available, the child may be placed with an unrelated foster family. Such a placement may be voluntary, as may be the case when a parent experiences a sudden illness, or involuntary, when the child protective services agency initiates an investigation based on a report of child neglect or abuse. In the latter case, foster care placement may occur if the child protection agency determines, on investigation, that the allegations are supported and a subsequent hearing before a dependency court judge finds that the child lacks proper care or control and should enter state custody (Child Welfare League of America, 1992; Support Center for Child Advocates, 1998). Even then, alternatives to out-of-home and nonkin placements must be considered. Moreover, with the enactment of the Adoption and Safe Families Act of 1997 (P.L. 10589), child protection agencies received a mandate from the federal government to act more quickly than they had in the past to terminate parental rights in the cases of some children.

Not all children in state custody reside with foster families. Some children are overseen by the juvenile justice system, which may place youths in residential programs (Martin, Peters, & Glisson, 1998). Others, particularly those identified as having mental health, developmental, or substance abuse problems, may be placed in residential programs or group homes that offer specialized treatment services (Martin et al.). As suggested, placement of children in state custody in foster homes, kinship care, group homes, and residential treatment involves parents in a complex web of legal and social welfare systems.

Research has found that children of SMI mothers who do not live with a parent or parents reside primarily in residential alternatives. If the children are living with a parent, the predominant family form is the mother-headed single-parent family (Mowbray et al., 2000; Mowbray, Oyserman, & Zemuncuk, & Ross, 1995; Rogosch, Mowbray, & Bogat, 1992; Sands, 1995; Zemencuk, Rogosch, & Mowbray, 1995). A survey of the 22 mothers hospitalized in an inner-city psychiatric facility, about whom data were available or known, found that only five mothers had full custody of their children (Joseph, Joshi, Lewin, & Abrams, 1999). The children of the other 17 were in kinship foster care (n = 5), nonkin foster care (n = 4), and various arrangements for different children or across time (n = 8). Among 87 children of 42 SMI mothers who participated in a focus group study, about 64 percent lived with their mother, father, or mother and partner, with the rest either in foster care, with adoptive parents, with relatives, or living independently (Nicholson, Sweeney, & Geller, 1998). More than 48 percent of the 46 women with a schizophrenic disorder who participated in Miller and Finnerty's (1996) study had a child in foster care, and more than 36 percent had a child informally raised by someone else. The children of SMI parents whose relatives were interviewed in a study by Gamache, Tessler, and Nicholson (1995) were living predominantly with the well parent or other relatives but also lived with the ill parent, both parents, and nonkin.

THE PRESENT STUDY

The present study builds on this earlier research by systematically asking each participant to speak as a mother about her children and to describe where each is residing. Whereas other research has uncovered the prevalence of single parenthood (Mowbray, Oyserman, & Zemencuk, & Ross, 1995; Rogosch et al., 1992; Sands, 1995; Zemencuk et al., 1995) and the range of diversity of child rearing arrangements (Gamache et al., 1995; Joseph et al., 1999; Miller & Finnerty, 1996; Nicholson et al., 1998), our research strove to map with greater specificity who was caring for each child and to achieve an understanding about the legal relationship of the child to his or her current caregiver.

Based on data from a pilot study of 20 women with SMI, we provide descriptive information about the current living arrangements and custody status of the children of 20 women with SMI. We identified the primary residence of the child and the person or institution that had day-to-day decisionmaking authority in meeting the child's needs. For example, when a child resides in his or her mother's home without other adults present, the mother is the primary custodian. If a child is placed in foster care, the foster home is the primary residence with the state granting the foster parents day-to-day decision-making authority. In such a case, the biological mother retains decision-making authority in discrete areas such as health emergencies. We also distinguish between full and partial physical custody on the part of mothers. Mothers with full custody have their children living with them and under their authority on a daily basis. Partial custody is the right to have the child completely in one parent's dominion and control part of the time without having any other caretaker present. A common version of partial custody is for a parent to have a child for weekends or overnight visits.

METHOD

Participants

A sample of 20 mothers with SMI was recruited from community mental health centers, residential treatment centers for addiction, homeless shelters, and a family and children's agency. Some potential participants were referred by service providers. Others volunteered after hearing about the research from the interviewers, who met with client groups to describe the study. Screening interviews were used to determine whether individuals who expressed an interest in participating met the criteria for the study.

We defined serious mental illness as a long-term, persistent psychiatric disorder, such as schizophrenia, schizoaffective disorder, major depression, or bipolar disorder, which impedes an individual's capacity to carry out primary aspects of daily life (Mowbray, Oyserman, Zemencuk, & Ross, 1995), and sought women who had a history of treatment for these disorders and had at least one biological child under the age of 18.

The participants' diagnosis was based on selfreport. The sample included 11 women from one of five residential substance abuse programs; two from an outpatient substance abuse program; five from one of three community mental health programs; and two from a family and children's agency.

Procedure

We conducted face-to-face interviews with the participants in their own homes or in a private space at the recruitment site. The interview included questions about sociodemographic characteristics, mental health history, pregnancy, and childbirths and an assessment of substance abuse using the drug and alcohol scales of the Addiction Severity Index (ASI) (McLellan, Luborsky, Woody, & O'Brien, 1980). Questions were also asked about the age, sex, and the current residence of each of the mother's living children, and whether the child lived with the mother or someone else or split his or her time between households. In addition, mothers were asked whether there was a court hearing in relation to each noncustodial child and who had the right to make decisions on each child's behalf. These questions required considerable probing by the interviewers, both of whom were lawyers as well as social work researchers. The interviewers took qualitative notes in conjunction with these interviews. Women were paid 10 dollars for their participation. During data collection, the participating mothers resided in the community and interacted with a variety of mental health and human services agencies.

FINDINGS

Characteristics of the Mothers

The sample consisted of 20 mothers whose age range was from 24 to 46 (M = 35.3 years, SD = 4.87). Seventeen women were African American; the rest were white (n = 2) or other (n = 1). The monthly family income ranged from $201 to $ 1,500, with a mean income of $575. The years of completed schooling ranged from 7 to 14, with a mean of 10.9 (SD = 2.06). Twenty-five percent, however, had some college. Their religious orientations were diverse.

Participants reported that they were hospitalized for a psychiatric problem from O to 20 times (M = 4.8, SD = 4.92). During the preceding year they were hospitalized from 0 to 6 times (M = 1.2, SD = 1.61). The 17 participants who reported a history of psychiatric hospitalization said that they were from 11 to 43 years of age when they had their first episode (M = 22.6, SD = 9.41). When asked their diagnosis, eight reported depression, three schizophrenia, three bipolar disorder, one dual diagnosis, three a combination, and two unsure (although their history of hospitalization and medications indicated SMI). From what we could determine from responses to the Alcohol Severity Index and their placement in dual diagnosis or substance abuse treatment programs, it appears that six had SMI only, and 14 had both SMI and a substance abuse disorder.

Pregnancy History

The age of the women at first pregnancy ranged from 14 to 34, with the mean at 18.8 years (SD = 5.52) and the median at 16.5 years. Sixty-five percent of the participants had their first pregnancy between 14 and 17 years of age. The mean age at which they had their first child was 19.3 (SD = 5.39), with the median at 17 years. Fifty-five percent had their first child between ages 14 and 17. The participants were pregnant from 1 to 12 times, with a mean of 5.5 (SD = 2.91). They reported a range of one to nine births and had an average of 3.8 births per mother (SD = 1.98).

Maternal Custodial Status and Children's Living Arrangement

The 20 mothers described the living arrangements of 76 children, of whom 41 were boys and 35 were girls. Thirteen were between newborn and age two; 30 were between ages three and 10; 17 were between ages 11 and 17; and 10 were 18 and older. The ages of six children were unknown.

The mothers had full custody of only 22 (28.9 percent) of the 76 children and partial custody of 7 (9.2 percent) (Table 1). The children of mothers with partial custody were living primarily with the children's biological father or a family friend.

Twenty of the SMI mothers'children (26.3 percent) lived in nonkinship care settings, mostly foster homes. Children were also in residential treatment programs and a group home. One set of foster parents who had been raising two siblings for many years legally adopted them. Based on the mother's long-standing relationship with the foster or adoptive parents, the parties informally agreed to an open adoption so that the SMI mother could maintain her relationship with her children.

A plurality of children (n = 33, 43.4 percent) lived with kin. Consistent with the literature, fathers, grandparents, and aunts were prominent as custodians, and all the grandparents except one were from the mothers' side of the family. In addition, some of the children lived with their older siblings, and one adult child lived independently.

From the perspective of the mothers (see Table 2), three had full custody of all their children and nine had custody of some children. Thus, 12 of the 20 SMI mothers had full custody of at least one of their children, but 8 mothers did not have full custody of any of their children. The children whose mothers had full custody were predominantly infants, preschoolers, and pradolescents. Among the five mothers who had partial custody, one had partial custody of all her children, with her two children living primarily with their father; two had full custody of one child and partial custody of another; and two had children living in a variety of settings. With one exception, the mothers with full or partial custody were responsible for one or two children. In the single case in which the mother had full custody of 4 children, the child protection agency was involved.

Remarkably, only three mothers had all their children with them. In another case, all the children were with their grandparent. The predominant pattern, seen in 16 of the 20 cases, was for the children to be scattered among a variety of placements-with the mother, kin, and nonkin in various combinations (Table 3).

Qualitative Findings

Almost uniformly, the participants commented on their enjoyment at being asked about their children, and some shared that this was the first time anyone had taken an interest in them as parents. The majority of women reported a salutary association with motherhood, whether or not they were caring for or residing with any of their children.

Not surprising, a number of the mothers appeared genuinely bewildered about the custodial arrangements of their children and articulated confusion about whether their children in longterm foster care had been adopted and whether they could do anything about the process.

Case Examples

Sherry, the mother of four children, under an agreement filed in family court by her youngest child's biological father, is entitled to partial custody, every other weekend, of their two-year-old daughter. Sherry also believes that if the weekend visits continue to go well and she completes her drug treatment program and finds employment, she will be awarded full custody. The interviewer had no way to assess the accuracy of her beliefs. Her three other children have resided in foster homes with nonrelatives for varying lengths of time. Sherry voluntarily placed her eldest son, now age 14, in foster care at his birth, and she is aware that the child protective services agency is moving to terminate her parental rights over him. Sherry wishes to oppose the termination, but she is unsure how to do this. She is also concerned that the state will initiate actions to terminate her parental rights to her other two children, and she does not know the foster care agency's position with regard to the other two children.

Kitty, the mother of two teenage boys, expressed regret that she did not have adequate guidance when she entered a psychiatric hospital, and her ex-husband filed for full custody in family court. Panicked about receiving legal papers while being hospitalized for depression, she asked her ex-husband what she should do about the impending hearing. He incorrectly advised her that the hearing concerned child support only, and he assured her that there was no need for her to attend or obtain counsel. As a result, the father received full custody of the boys. He blocked her from seeing her sons for a time, and she had to initiate a legal action to receive the right to overnight weekend visits.

Linda was the only study participant who became uncomfortable and highly agitated during the child custody portion of the interview. Linda is the mother of five children who range in age from four to 16. She resides in one side of a duplex with her eldest teenage daughter. The maternal grandmother, who lives on the other side of the double house, is the primary caretaker for Linda's four-year-old daughter. Linda's other three children have lived with the biological father since their birth, and he has gone to family court to secure his custody rights. Linda became agitated when the interviewer asked how often she sees the three children who live with their father. Apparently, she has not been permitted to visit for nearly a year, a situation that angered her. After Linda left the interview session, the interviewer met briefly with Linda's case manager, who commented on what a strong and positive support system Linda has in caring for her children. It was obvious that the case manager was not aware of Linda's lack of access to the three children in their father's care.

These case examples give texture to the numbers by illustrating how confounding it is for a mother to have children residing in different settings.

DISCUSSION AND IMPLICATIONS

This study found that although the SMI mothers in the sample became pregnant and bore children at what appeared to be relatively high rates, they were rearing only 29 percent of their children. The mothers in our sample simultaneously had some children living with them and others living in kinship and nonkinship arrangements, with few mothers having full or partial custody of all their children. The predominant out-of-home placements were with kin-particularly fathers and grandparents-but it was typical for one child to be with the biological mother, one with a relative, and another child in a nonkin foster home. Thus, we see the families of SMI mothers dispersed in a variety of living situations. This finding raises questions about describing families of SMI mothers and their children as one household (Hartman & Laird, 1983) and, like the other major findings, has implications for practice.

The finding that the families of SMI mothers and their children were dispersed across households needs to be considered in relation to the context of the interviews. First, qualitative findings confirm research observations that women with SMI derive great meaning and pride from being mothers (Mowbray, Oyserman, & Ross, 1995; Sands, 1995; Schwab, Clark, & Drake, 1991). Because motherhood for women with SMI appears to be part of their positive identity, we recommend that this status be fully explored and considered in working with these mothers. We acknowledge, however, that the meaning of motherhood to the women is a separate issue from the meaning of the mother to the child and whether maternal custody is in the child's best interest. Children are affected by the presence and absence of the biological mother in their lives. Although our research focuses on the mothers, we believe that decisions about custody and visitation should consider and balance the best interests of both.

Second, the vast complexity in the family configurations of the 20 women in our small sample suggests the utility for practitioners of "mapping" where each noncustodial minor child resides, ascertaining from the mother or from outside agencies the terms of the custodial arrangements and discovering the mother's feelings and expectations about each child custody plan (see Minuchin et al., 1998). Our data show that 16 of 20 mothers had children residing in at least two different settings, and it was not unusual for siblings to be spread out over three settings. Some mothers had to mediate two legal systems-the family court system and the child dependency system-in addition to managing interpersonal arrangements with kin and other caretakers.

Because the multiple custody settings and systems may be confounding to mothers with SMI, clinical social workers need to offer case management to help mothers maintain contact with their children. On a more urgent note, mothers with SMI with children in long-term, nonkinship foster placement are at risk of the state's initiating an action to terminate their parental rights and free their children for adoption (Roberts, 1999). Although adoption may be in the child's best interests in some instances, social work practitioners working with the mother have a responsibility to assist her themselves or link her with professionals who can help her, so that she can fully exercise and protect her parental rights. Beyond helping mothers mediate with various child custody systems to remain connected with their children, social workers' full appreciation of the complex custodial arrangements and the mother's feelings about them will enable practitioners to effectively help the mother cope with issues of grief around custody loss, separation, and an inability to visit.

A final implication for practice addresses findings about the mean birth rate of 3.8, the norm of having physical custody of only one or two children, and children's placements in multiple settings. It appears that it is difficult for women with SMI to care for more than one or two children at a time, unless the mothers have strong social supports. Beyond the assistance offered by families, formal social supports related to parenting would be a desirable component of a comprehensive treatment plan. Although a few parenting programs specific to this population exist (for example, Jenkins, 1996; Rubovits, 1996; Zeitz, 1995), they are few and far between. Many communities do, however, have other parenting programs that may be appropriate for mothers with SMI. Our research points to the need for parent education classes, parent assistants, and clinical work that will strengthen the women's ability to parent children in their care. If parenting with support continues to be problematic, mothers need help facing the prospect of losing custody and having a family that is dispersed among multiple households.

This study was limited by its small nonrepresentative sample in which low-income, African American mothers predominated. Findings about family patterns of children of SMI mothers may be related to the mother's low-income status or reflect a cultural pattern of relying on the African American extended family for caregiving. Research on a larger, more representative sample should clarify how generalizable the patterns we found are to the target population of women with SMI who have children. There are also limitations with selfreported diagnoses. Some women may not recall their diagnosis or may not know what it is. Had we relied on clinical diagnoses in the case records of those who were in psychiatric treatment, it is likely that there would have been some inaccuracies or inconsistencies across settings. Given the exploratory nature of this study, research diagnoses were not feasible.

This pilot study of the custodial arrangements of children of women with SMI found that mothers had full or partial custody of fewer than half of their children, and typically the children were dispersed in a variety of settings under the aegis of different legal and social welfare systems. Considering that motherhood is a vital source of meaning to the mothers, social workers face the challenge of helping these mothers sort out the systems, assert their custodial rights, and live with and without custody of some or all of their children. HSW