Children of adults with severe mental illness: mental health, help seeking and service use

AIMS AND METHOD 

We aimed to determine the prevalence of childhood mental health problems in children of parents registered with an Australian area mental health service, and to study the parents’ help-seeking and service use for their children. Parents were recruited through their case managers, and asked to complete the Strengths and Difficulties Questionnaire (SDQ), the Service Utilisation Questionnaire and the Help-seeking Questionnaire.

RESULTS 

A quarter of the children were in the clinical range of the SDQ total scores, with high sub-scale scores. However, 63% of the parents reported reluctance to seek help, and 19% reported not using services.

CLINICAL IMPLICATIONS 

Children of parents with mental illness are at higher risk of childhood psychiatric disorders. Assessment of the child at the time of referral of the parent is an opportunity for problem identification, parental education, and early intervention.


Procedure
Participants were clients of an area mental health service in Victoria, Australia, who were parents of dependent children up to 18 years old. The mental health service area includes rural, semi-rural and suburban localities, and contains over 8% (352 000) of the state population. Families with children comprise at least 57% of households in this area, compared with a proportion of 50% for the state. Generally, the proportion of households receiving low to average incomes is the same as that reported for the state, with some variation among localities within the service area. Around 45% of the population in the area covered by the mental health service is aged 18-50 years (Department of Infrastructure, 1998).
In Victoria adults attend a regional mental health service when the severity or impact of their illness cannot be managed by primary mental health services, such as general practitioners.
Participants for the project were identified in two ways. Case managers in the community treatment teams were asked to inform clients with children about the project; clients of the in-patient unit were approached by the research worker (V.C.). In the period the data were collected, the service teams had 846 registered active clients: 136 were identified as parents of dependent children and were asked to participate in the study; 23 parents could not be contacted, and 17 were not included because they were too unwell, were not seeing their children, were out of the service area, or had recent or current involvement with child protection services. Thirtyfive parents did not give consent, a refusal rate of 36%.

Demographic data
Information was collected about the parents' age, gender, education level, ethnicity, language spoken, employment status, marital status, household, number of children, type of housing, change of accommodation, income level and relationship to the child. Parents recorded the age and gender of each child under 18 years old.

Psychiatric disorder
The Strengths and Difficulties Questionnaire (SDQ) parent form (Goodman, 1997(Goodman, , 1999) covers a range of major symptoms, and the extended form for parents includes questions on subjective distress, interference with home, school, peer and leisure activities, and impact on others. The SDQ has been shown to be reliable in identifying psychiatric disorder in community samples (Goodman et al, 1999(Goodman et al, , 2000Koskelainen et al, 2000).

Service use
The Service Utilisation Questionnaire was adapted from Hornblow et al (1990) and asked parents if they had sought help in the past year, and if so the number and type of agencies contacted.

Help-seeking
The Help-seeking Questionnaire (Hornblow et al, 1990) assessed the presence and type of perceived barriers to seeking help.

Results
Descriptive statistics are shown for the SDQ results for 101 children aged 4-16 years, and help-seeking, barriers to accessing services and service use are recorded for all children in the study.
Sixty-one parents were recruited, with 128 children between the ages of 2 months and 18 years. The parents' age range was 25-56 years (mean 38 years, s.d.=7). There were 10 men and 51 women. Forty-eight parents recorded their diagnosis (including nine who reported multiple diagnoses). Fourteen parents recorded schizophrenia, 11 bipolar disorder, 18 depressive disorder, 3 personality disorder, one acute psychosis and one postnatal depression. Thirteen patients did not know. Seven parents lived alone, 37 were not married, 31 parents lived in their own flat and 13 had moved their living place four or more times in the past 5 years; 19 parents did not complete secondary school, and 49 were not working. Of the total 128 children, parents recorded being the legal guardian for 117 children.

Prevalence of children with mental health problems
The extended SDQ is designed for children in the age range 4-16 years, so data on 101 children were analysed; the children had a mean age of 9.4 years (s.d.=3.4), and 50 were boys. The highest proportion of children was in the 4-8 years age band (45%); Table 1 presents the number of children in each of three age bands with the means and standard deviations for their total and subscale scores.
The total SDQ scores and the sub-scale scores are presented in Table 2. Cut-off scores for the clinical range had been adjusted so that 10% of the population were above that point. Children in our study had 2.5 times the rate of mental health problems compared with the community norm.

Difficulty, impact and burden
Parents reported that 62 children (61%) had difficulties in mental health: 38 had minor difficulties, 17 definite and 7 severe. Most difficulties had been present for more than a year for 45 of the children. In 60% of the children with difficulties, the problems were rated as having a significant impact on the family and the child: this impact included interference with friendships (36%), classroom learning (38%), home life (65%) and leisure activities (17%). The burden on the family was rated as 'quite a lot' in 34% and as 'a great deal' in 16% of families.

Help-seeking and barriers to treatment
Thirty-seven of the 61 parents reported reluctance to seek help for their children's behaviour problems in the previous 12 months, although it might have been helpful. Correlation tests showed no significant relationship between parent's help-seeking response and age, gender, education level, income level, marital status, family size or number of places lived in the past 5 years. Common reasons presented by parents for not seeking help were that they thought they should be able to handle problems alone (59%), they did not know where to go for help (41%), and that they did not think anyone could help (38%). Other reasons included embarrassment, fear of what others would think and fear of the treatment that might be given to the child.

Service use
All parents were asked to report on the types of services they had used for their children during the past 6 months (Table 3). A scale of service use was formed by combining the services used by each child, and the relationship between this scale and the child's mental health was examined by correlating the scale with the SDQ total problem score using a non-parametric test. A significant relationship was found between the level of service use and the SDQ total problem score: r=0.41, P50.0001. Cowling

Clinical implications
The prevalence rate of mental health problems in a sample of children of parents with mental illness, according to the SDQ total score, is 2.5 times the norm. This supports existing findings that children of parents with mental illness are at higher risk of mental health problems. Alongside this finding, we need to acknowledge the effect that parental mental illness itself might have on the reporting of child behaviour problems, with anxious or depressed mothers in one study reporting more cases of child behaviour problems than mentally healthy parents, or children themselves (Najman et al, 2000). Compared with the community norms reported by Goodman (1997), the parents reported higher scores on all the SDQ sub-scales, particularly emotional problems, peer relations problems and conduct problems (at the time this study was conducted norms for the Australian population were not available). In addition, parents in our study reported more difficulties, impact and burden in relation to their children than might be expected from the actual SDQ symptom scores. We compared our data with Goodman (1999) and found that children in our study were rated higher than a community sample but lower than a clinic sample (difficulties 36%, 61%, 96%; impact 7%, 36%, 83%; burden 9%, 31%, 76%, for community, study and clinic samples respectively). This result further supports the need to target children of parents with mental illness in the planning of mental health service delivery, particularly prevention and early intervention.
Almost all the children with mental health problems (defined by the SDQ as being within that clinical range) in this study (21 of 22) attended at least one service to seek help, compared with half of those (defined by the Child Behaviour Checklist as being within that clinical range) reported in an Australian survey by Sawyer et al (2000). A diverse group of services was used. However, 63% of the parents reported reluctance to seek help in the preceding 12 months for their child's behaviour, compared with a 12.5% rate in a community study of parents (Pavuluri et al, 1995). This suggests that case managers of adult patients need to help those who are parents by providing information about child mental health problems and their treatment.
The high rate of service use was not predicted by the high level of self-reported reluctance to seek help. The Help-seeking Questionnaire asked parents a global question: was there a time you did not seek help in the past 12 months for your child/children? The question, and the 16 possible reasons for not seeking help, might have reminded parents about times they had not sought help for one or more of the reasons given (even though help was needed). Reported reluctance to seek help could also reflect parents' need to be seen as coping. On the other hand, the Service Utilization Questionnaire concerned the individual child and listed specific services that might have been attended, and so required less subjective responses. The issue of help-seeking and service use requires exploration in future studies.
The SDQ proved to be useful and acceptable to parents and may allow ready identification of children at greatest risk. All patients with dependent children should be identified by adult mental health services, and the SDQ can be offered to these parents as a way of checking whether their child needs help.

Limitations
The study had several limitations. Data on the proportion of clients who are parents are not routinely collected by many mental health services in Australia, so we do not know whether our sample is representative of the adult mental health service population. Because public sector services treat people with major mental illness, it is likely that the groups will be similar to those in the UK (Oates, 1997), although aspects of service delivery may differ. Parents' reported diagnoses were not checked against their records as we had not sought permission for this, being aware of parental sensitivity in the area of care and parental competence. Information about the child's mental health was obtained only from a single sourceagain, we did not want to arouse parents' anxiety by contacting other people about their children's mental health. Similarly, we excluded parents currently or recently involved with the Child Protection Service, to avoid raising the anxiety of unwell parents. In the state of Victoria, the Child Protection Service is a statutory authority whose task is to receive notifications of child abuse and neglect from the community. The direct assocation between the service and the possible removal of children leads to a negative perception of this organisation by parents. In the future, exclusion of certain categories of parents and/or children from research studies should be unnecessary, with growing awareness of the need to help children whose parents have mental health problems.
Cowling et al Children of adults with mental illness original papers

Declaration of interest
None. This project was supported by a grant from the Australian Rotary Health Research Fund.