About 1 out of every 5 children and adolescents has a significant mental health problem (Offord, Boyle, Fleming, Blum, & Grant, 1989; Offord, 1998; Pavuluri, Luk, & McGee, 1996; Costello et al., 1996). Yet only about 20% of children with mental health problems receive some form of help and just 5% receive specialized mental health services (Offord et al., 1987; Pavuluri et al., 1996; Verhulst & Van der, 1997; Leaf et al., 1996; Angold et al., 1998).

Classic models of help-seeking assert that a person must first recognize a problem, decide to seek help, and then select a source for help (Gurin, Veroff, & Feld, 1960; Fischer, Weiner, & Abramowitz (1983). Rogler and Cortes (1993) introduced the concept of a help-seeking pathway to describe how people navigate mental health systems. They defined pathway as “the sequence of contacts with individuals and organizations prompted by distressed person’s efforts, and those of his or her significant others, to seek help as well as the help that is supplied in response to such efforts” (p. 555). They suggested pathways have a linear direction (i.e., individuals contact organizations in a sequential manner), and specific duration (i.e., time lapses between help-seeking efforts).

Recent help-seeking models have been adapted to describe a more detailed progression of the help-seeking process. Srebnik, Cauce, and Baydar (1996) presented a model that described how youth utilized mental health services incorporating factors, originally proposed by Anderson and colleagues (Andersen & Newman, 1973), that impact the help-seeking process including subjective need for services, predisposing characteristics, and enabling and inhibiting factors. Logan and King (2001) proposed an elaborate linear progression of help-seeking for adolescents with mental health problems. They suggested that parents must first gain awareness of their adolescent’s distress, recognize the problem as psychological in nature, consider possible courses of action, develop the intention to seek mental health services, make an active attempt to seek services, and obtain mental health services. Progression through these stages is facilitated or derailed by social, cultural, and systemic factors. For example, physician referral increases the likelihood of obtaining specialized mental health care (Sayal, Taylor, Beecham, & Byrne, 2002).

Various elements in models of how parents access mental health care for their children have empirical support. Parental problem recognition is related to help seeking (Pavuluri et al., 1996; Rawlinson & Williams, 2000) and problems with transportation and costs lead to decreased access (Benway, Hamrin, & McMahon, 2003). However, some data are inconsistent with a linear progression through specific stages. Pavuluri, Luk & McGee (1996) identified 42 of 320 children with behaviour problems from community preschools. Consistent with a linear progression, the number of parents who recognized a problem (n = 19) and sought help for the problem (n = 17) was much greater than the number who finally obtained help (n = 8). Interestingly, 50% of parents who sought help (n = 4) “did not believe there was a problem but sought help all the same” (p. 219). It would seem that half of the parents who obtained help bypassed the problem recognition stage, suggesting that the help-seeking process is not always linear but may involve multiple pathways.

Alternatives to a linear progression of help-seeking include the family network-based model of access to children’s mental health services (Costello, Pescosolido, Angold and Burns, 1998), and the related gateway provider model (Stiffman, Pescosolido, & Cabassa, 2004). These models propose that multiple influences dynamically affect how children obtain mental health services. Parents are seen as key “gateway providers” in terms of recognizing their child’s need for help, and taking steps to obtain help. These models include treatment system factors including organizational constraints related to accessibility (Costello, Pescosolido, Angold, & Burns, 1998). The multiplicity of entry points into mental health care and the mix of services available have been identified in previous studies (Farmer, Burns, Phillips, Angold, & Costello (2003). Less is known about how parents navigate the systems that provide mental health care for children and in particular, how parents respond to problems in accessing services.

The present study examined the process of parental help-seeking among families who were actively seeking help for their child. Four issues were examined with a focus on exploring whether the help-seeking process is linear. (1) Do parents identify one single problem or multiple co-morbid problems? A linear help-seeking process would suggest that parents recognize a problem and seek help for that particular problem. (2) How do parents choose the agencies they contact? (3) Do parents simultaneously seek help from multiple agencies, and receive services from multiple agencies? A linear process would imply that parents seek help from one location, and receive help from the same location without accessing multiple services. (4) When treatment is offered, do parents want what is offered to them? Beyond commonly cited barriers for accessing services (e.g., transportation, cost, etc.), we explored parents’ agreement with the treatments they were offered and whether they accepted the help offered to them.

Methods

Participants

Parents who contacted a centralized intake serving three publicly funded Children’s Mental Health Centers in London, Ontario (population about 337,000, metropolitan area about 432,000) for help with their child’s problems were recruited. The three centers specialize in treating behavioural and emotional problems for children and adolescents up to age 18 in the city and surrounding areas. Families seeking treatment from these centers tend to have children with more severe externalizing problems than internalizing problems. The mean T-score (age- and sex-based norms based on a community population sample) for externalizing problems of all children receiving treatment from the centers at the time of the study (M = 72.4, SD = 12.2) was significantly higher than the mean for internalizing problems [M = 64.1, SD = 14.7; t(748) = 8.48, P < .01].Footnote 1 Parents do not require a formal referral to obtain services from the agencies.

Inclusion criteria were: (a) Parent or legal guardian contacted the center regarding help for their child, (b) child age 4–17 years old. Exclusion criteria were: (a) crisis situations (i.e., an adverse event that was imminent or had recently occurred and the family was given immediate service by the center), (b) parents unable to speak or read English, (c) parents communicated with the center only through answering machines, (d) doctors or other health/mental health care professionals contacted the center for the parent, (e) parents who were not the custodial guardian of the child. The study was approved by the University of Western Ontario’s Institutional Review Board. There were no study incentives for participants.

During the 7 months of data collection, 487 parents called for help with their child and 144 (30%) of these parents agreed to be contacted about the study; the remainder either did not meet study criteria or intake workers neglected to ask them to participate. Of the 144 eligible parents who were asked and who agreed to be contacted, 23 had more than a 2 month period between the intake call and research contact and 34 could not be contacted despite repeated telephone calls.

Of the 87 parents contacted, 60 (69%) agreed to participate in the study. The primary reasons for declining were lack of interest or not enough time. It was not possible to compare parents who agreed to be contacted with those who did not agree, nor was it possible to compare parents who released their contact information and who agreed to participate to those who declined participation because intake staff collect information only on families offered treatment and not all families seeking help are offered treatment.

Children of parents who participated in the study were 3–16 years old (M = 10.57, SD = 3.8); 40 (67%) were male. Parents were 22–50 years old (M = 37.55, SD = 7.02); there were 52 mothers and 8 fathers. Parents reported their annual income in one of seven categories. The modal annual income per family was CA $10,000–$19,999 (US equivalent approximately: $7,700–$15,500), which is in the low-income range for a Canadian family in an urban setting (Statistics Canada, 2004). Of parents who participated in the study, 62% went on to receive treatment from one of the three treatment centers, the other 38% were not offered treatment at one of these three centers. This is similar, but slightly higher, than the total number of parents who contacted the intake service and received treatment from treatment centers during the study period (52%). The severity of child problems among families who participated in the study was not significantly different from the severity of child problems in all families receiving treatment from the centers during the time the study was conducted [Externalizing problems: t(433) = 1.6, n.s.; Internalizing problems: t(433) = 0.4, n.s.].

Procedures

Intake workers completed a standard intake interview and then asked all eligible parents if they would be willing to be contacted about participating in a research study. Parents were mailed a letter of information about the study and were contacted for the first time by the researcher over the telephone approximately 1 week later. Parents who consented completed a structured interview lasting about one-hour that included the Brief Child and Family Phone Interview (Cunningham, Pettingill, & Boyle, 2000) to assess the severity of children’s behaviour problems, and questions designed for the present study to assess parents’ decision-making processes and experiences related to accessing mental health services. The average length of time between the parents contact with the intake worker and the completion of the interview was 2.3 weeks (Mdn = 2.1, SD = 1.5).

Measures

The Brief Child and Family Phone Interview (BCFPI)

The BCFPI is a 30-minute standardized telephone interview with 81 forced-choice questions (Cunningham et al., 2000). It is the mandated intake measure used by all children’s mental health centres in the province of Ontario. The BCFPI is based on the Ontario Child Health Study scales - Revised version (OCHS-R) (Boyle et al., 1993). Parents report the frequency of each child’s behaviour as “never”, “sometimes”, or “often”. The present study used three composite scales, which were based on nine factor analytically derived subscales: (a) externalizing (i.e., regulation of attention and activity; cooperation; conduct), (b) internalizing (i.e., separation from parents, managing anxiety and managing mood), and (c) child functioning/impairment (i.e., social participation, quality of child’s social relationships, school participation and achievement). Norms and reliability were derived from community and clinic data from the OCHS. Internal consistency for eight of the nine BCFPI subscales in the community sample ranged from .75 to .86 (Cronbach’s alpha) (Cunningham et al., 2000). Internal consistency for the conduct subscale was .56. The authors of the BCFPI, however, considered this subscale to be too infrequent in a community sample to produce reliable results. Correlations between the BCFPI subscales and the OCHS-R full-length scales range from .78 to .96. For the present study, T-scores were computed using the age and sex based norms from the community sample.

Help-seeking and Health Care Utilization

Parents were asked the number of times they contacted a list of 17 agencies/ professionals Footnote 2 in the past year; this included 10 agencies from the OCHS-R health care utilization schedule (e.g., physician, children’s mental health centre) (Offord et al., 1989) and seven additional agencies providing children’s mental health care in the surrounding community. For each agency contacted, parents were asked the date of their first and last contact, the number of times they contacted the agency, how they chose the agency, what problem they wanted to address, what kind of help was offered, whether it was the help that they wanted, how long they had to wait for the help, and if they were referred anywhere. These questions were repeated for each agency the parent encountered over the past year. In addition, parents reported the impact of the child’s problem upon themselves and their family.

Variables derived from the interview are outlined below. Open-ended questions were coded for each agency using a standardized coding manual. A research assistant was trained to use the coding manual to a kappa of at least .75 using data from 10 parents (who together contacted 36 agencies). Disagreements in the coding were resolved by discussion for these 10 interviews. Inter-rater reliability between the research assistant and the interviewer (D.S.) was calculated on data from 20% of the remaining interviews. All kappa values were in the excellent range; across eight codes kappa values ranged from 0.82 to 0.93 (M and Mdn = 0.89).

Type of Problem

Parents’ statements about the problems they wanted to address with each agency were coded into 26 symptom categories that were consolidated into eight problem areas: aggression/defiance, family functioning, anxious/depressed, learning difficulties, attention/concentration, physical symptoms, social functioning, other. See Table 1 for the definitions of each problem area.

Table 1 Problem areas that parents identified when seeking help for their childa
Full size table

Agencies Contacted

Two variables related to mental health agency contacts were developed. (1) Number of agency contacts past year. The total number of agencies parents contacted in the past year for their child’s mental health problem was computed. Some parents had contacts with agencies prior to 1 year ago; only contacts within the past year were discussed with parents. (2) First agency contacted. Parents’ recollections of whom they chose to first speak with about their child’s problems were documented. This represented parents’ very first contact with a professional, and may have been prior to 1 year ago. Responses were aggregated into the following categories: (a) physician, (b) school, (c) child welfare, (d) children’s mental health centre, (e) psychologist, (f) psychiatrist, and (g) police.

Treatment History

The length of time parents had been accessing services was documented by comparing when parents contacted an agency for the first time and the date of the interview. The length of time between their first contact with an agency and the interview date was coded as either (1) began accessing services more than 1 year ago, or (2) began accessing services within the past year.

Choice of Agency

Parents were asked why they chose to contact each agency. Responses were coded into seven categories: (a) referred by professional, (b) referred by non-professional, (c) personal knowledge of agency, (d) current or past involvement with the agency, (e) agency contacted the parent, (f) agency was most accessible, or (g) other.

Services Offered

Parents were asked what services each agency had offered them. Responses were coded and then categorized into (a) treatment or (b) non-treatment services. Treatments were coded in 13 categories and involved any form of intervention aimed at resolving the child’s problem (e.g., individual counselling with parent or child, parent group, medication). Non-treatment services were coded into nine categories and then aggregated as referrals or other services (e.g., intake appointments, assessment, listening or providing support, parent respite). Referrals included formal requisitions for specialized services by family physicians, as is typically the case in the Canadian medical care system (i.e., family physicians refer patients to medical specialists) and recommendations by agency staff (professionals or non-professionals) that parents should seek services elsewhere.

Treatment Received

Parents were asked about treatment from each agency. If treatment had not been received from the agency it was coded as (a) waiting or (b) other (e.g., professional cancelled the treatment).

Simultaneous Agency Contacts

The most recent agency contacts were grouped by each month prior to the interview. Simultaneous agency contact was defined as having contact with more than one agency within the same month. The total number of agencies that parents were simultaneously in contact with in each of the 12 months prior to the interview was computed; simultaneously contacts 6–12 months prior to the interview were averaged due to low frequency of contact per month for this period. This method was modeled after Farmer, Stangl, Burns, Costello, & Angold (1999) who aggregated contacts over a 3-month period.

Agreement with Treatment

Parents who were offered treatment were asked if they wanted the treatment that was offered. Responses were coded as (1) Treatment wanted, (2) Treatment not wanted because of barriers (e.g., transportation, cost, lengthy waiting, child refusal to attend treatment), or (3) Treatment not wanted. Treatment not wanted refers to times when parents disagreed with the type of treatment offered; that is, they did not want the type treatment offered either because they felt that they or their child would not benefit from the treatment (e.g., offered parenting classes but the parents felt they had taken enough parenting classes already), or they felt that there was a different treatment that was more suitable (e.g., offered an anger management group for the child, but the parent wanted medication for the child).

Impact on Parent

The impact of the child’s problems on various aspects of the parent’s life was measured using six questions from the Child and Adolescent Impact Assessment (CAIA) (Angold, Costello, Farmer, Burns, & Erkanli, 1999) and one question adapted from the Child Health Questionnaire (CHQ) (Landgraf, Abetz, & Ware, 1996). Questions asked parents to rate on a scale from 1 to 5 (1 = not at all, 5 = very much so) the extent to which the child’s problems had affected the parents’ emotions, energy, time for personal needs, and physical health during the past 4 weeks.

Data Analyses

To calculate the reasons parents chose an agency as well as the services offered to parents, data were aggregated across parents or across agencies. This allowed for an examination of the services delivered to children from a systems perspective that could not be examined from only analyzing the data at a parent-by-parent level. T-tests were used to examine group differences within the sample.

Results

Help-Seeking History

Physicians were the most common first contact for help with their child’s problem (42% of parents) and schools were the next most common (22%). Parents first sought help for their child an average of 3.1 years ago (SD = 3.2; range = 0.1–12.8 years ago), when their child was an average of 7.5 years old (SD = 4.3; range = 1.2–15.5).

Thirty-seven percent of parents participating in the study began contacting professionals for their child’s mental health problem within the past year (n = 22); 63% had been in contact with at least one agency more than 1 year before the interview. Parents who began contacting agencies more than 1 year ago had children with more severe externalizing problems [t(58) = 2.34, P < .05], compared to those who began looking for help within the past year; there were no significant difference in terms of child age [t(58) = .42, n.s.], socio-economic status [t(58) = .22, n.s.], child internalizing problems [t(58) = .72, n.s.], or the impact of the problem on the parent [t(58) = .03, n.s.].

Parents’ Help-Seeking Experience: Past Year

Parents sought help for an average of two different problems (SD = 1.15, Mdn = 2, range 1–6). Table 1 presents the percentage of parents seeking help for each problem area. Parents contacted an average of 4.9 different agencies during the previous year for help with their child’s problems (SD = 1.91, Mdn = 5, range = 1–11; this includes contact with the agency from which the parents were recruited). Families had contact with all service sectors including medical (e.g., family physician, pediatrician, psychiatrist, emergency room), school (e.g., school psychologist, guidance counsellor, school board, learning disability association), mental health (e.g., child and adult mental health center, private psychologist), child protection agencies, justice (e.g., police, probation, courts), and community and social services (i.e., local community agencies).

“Referral” (by professionals and non-professionals) was the most common reason parents contacted an agency (52% of all agency contacts, n = 159/303), current or previous involvement with the agency and personal knowledge of the agency (i.e., knew about the agency from advertisements or other sources, such as a relative employed there) were the next most common reasons (see Table 2).

Table 2 Reasons parents chose to contact agencies
Full size table

Simultaneous Agency Contacts

The vast majority of parents (87%) were simultaneously in contact with more than one agency at some point within the past year (see Fig. 1). Parents were involved with increasingly more agencies simultaneously over time. In the 6–12 months prior to the interview parents were simultaneously in contact with an average of less than one agency (M = 0.05) versus 2.6 agencies in the month prior to the interview. As one parent stated, “I didn’t talk to any other places while I was on the waiting list for the psychologist because I was sure that the psychologist could help. He didn’t help one bit, now I talk to more than one place at a time or I never get anywhere”.

Fig. 1
figure 1

Number of times a parent was simultaneouslya in contact with more than one agency in the previous yearb. aSimultaneous agency contact was defined as having contact with more than one agency within the same month. Plots show the M as a dark bar, SD as shaded box, and range as error bar. b N = 60 parents

Full size image

Services Offered

During the previous year, parents sought help from up to 11 different agencies and each agency offered up to three services for each family. In total, 329 services were offered to the 60 families. Thirty-six percent of the services were treatments (n = 117), the remainder mainly included services such as referrals (27%), intake appointments (12%), listening/providing support but not treatment (e.g., at a physician’s office) (10%), and assessments (3%).

Services Received

Over two-thirds of parents reported having received treatment during the past year prior to contacting the children’s mental health centre from which they were recruited. In other words, over two-thirds of parents had already received treatment and were seeking additional treatment. Fifteen percent of families had not yet received any treatments at the time of the interview. Of the 85% who received at least one treatment, parents contacted an average of 2.3 agencies prior to receiving that treatment (SD = 1.7, Mdn = 2, range = 1–11), and they were placed on a waiting list 50% of the time. During the previous year, families received treatment from up to seven different agencies (M = 2.0, SD = 1.6, Mdn = 2). Within agencies, parents received up to three different treatments at each agency (M = 1.3, SD = 0.6 Mdn = 1).

Agreement with Treatment

Parents wanted the majority (66%) of treatment services offered to them (see Fig. 2). As would be expected, parents consented to participate in all treatments that they wanted for their child. Ten percent of treatments offered were not, or had not yet been, received (7.5% on a wait list, 2.5% agency cancelled treatment). A minority of treatments offered (8%) were declined due to barriers. About one quarter of the treatments offered (27%) were not wanted because parents disagreed with the treatment approach. Despite parents’ disagreement with the treatment, parents declined only 5% of these treatment services. Thus, out of all treatment services offered to these families, 21% of the time parents agreed to treatments that they did not want. In each of these cases (20/25 treatments) families participated in the treatments; for one treatment the family was still on a wait list and in four cases the agency cancelled the treatment.

Fig. 2
figure 2

The distribution of services offered to parents. Services offered to parents were divided into treatment services and non-treatment services. Parents reported whether the treatments offered were wanted or not wanted. The percentage of treatments that parents consented to and the percentage that parents received were documented. The total N for this figure (N = 329) is greater than the number of agencies contacted by parent (Table 2: N = 303) because some agencies offered more than one service to each parent

Full size image

We examined whether a small subset of parents accounted for not wanting any of the treatments offered. Of the 51 parents who were offered one or more treatments during the past year, 55% (n = 28) wanted all the treatments offered to them, 26% (n = 13) did not want some treatments, 14% (n = 7) did not want any, and 6% (n = 3) encountered barriers to all treatments. Parents who did not want at least one of the treatments (39%) did not differ from parents who wanted all treatments in terms of child age [t(46) = 0.6, n.s.], socio-economic status [t(46) = 1.6, n.s.], child problem severity [internalizing: t(46) = 1.4, n.s.; externalizing t(46) = 1.0, n.s.], family having accessed services more than 1 year ago versus within the past year [χ 2(1) = 1.0, n.s.], or the impact of the problem on the parent [t(46) = 0.4, n.s.].

We also examined whether there was a specific type of treatment that accounted for the treatment services that parents did not want because they disagreed with the treatment approach (see Table 3). Parenting classes, counselling for parents and medication were the most common treatments with which parents disagreed. In total, 11 different types of treatment were not wanted by parents.

Table 3 Treatments to which parents consented, but did not want because of disagreement with the treatment approacha
Full size table

Discussion

Most families involved in professional services for their child during the past year looked for help for multiple problems in multiple places. The results of this study support models of help-seeking such as the network-based model of access to children’s mental health services (Costello et al., 1998) and the gateway provider model (Stiffman, Pescosolido, & Cabassa, 2004). The results do not demonstrate strong support for linear progressions of help-seeking, which suggest that parents first recognize a problem, then contact organizations in a sequential manner for help with that problem, and consequently receive help (Rogler & Cortes, 1993).

Do parents identify one single problem or multiple co-morbid problems? Parents did not seek services in response to a particular problem but rather, they sought services for multiple, co-morbid problems. Other studies of families presenting at children’s mental health centers have also found that parents report multiple presenting problems (Yeh & Weisz, 2001). A linear help-seeking process would suggest that parents often begin this process by first recognizing the problem and then seeking help for the problem. However, not all families need to recognize a problem to enter the mental health care system (Pavuluri et al., 1996). Consistent with the gateway provider model, families could access services because they were told to get help for their child by a professional such as the school, or because they were mandated to receive treatment by child protection agencies. The frequency of parents who are told or mandated to seek help was not examined in the present study; future studies should examine this.

How do parents choose the agencies they contact? Again, in support of the gateway provider model and consistent with studies from the United States (Burns et al., 1995; Farmer et al., 2003; Lavigne et al., 1998; Stiffman et al., 2004), the majority of parents in the present sample began to access help through physicians or school personnel. It should be noted that for the present study, first point of access was modelled after Farmer et al. (2003) who reported parents’ recollection of first contact based on lifetime use of services, rather than use of services for the current episode of help-seeking.

The high percentage of time parents utilized referrals with professionals and non-professionals, along with personal knowledge of agencies from friends or family, is consistent with social network theories suggesting that the process of help-seeking involves consultation with numerous formal and informal sources. Referrals may have been the primary choice for how to select an agency because parents may not have been aware of the various service options available to them or because parents thought they required a referral to access services, which is not the case for children’s mental health centers in Ontario. Parents may have simply preferred to follow the advice of professionals who have knowledge of the various service options in their community. Given the number of agency contacts prior to treatment (ranging from 1 to 11), these data suggest that agencies in the community may not be as knowledgeable about service options as parents would assume. Other studies have found that service provider knowledge is related to accessing mental health care [see (Stiffman et al., 2001)].

Do parents simultaneously seek help from multiple agencies, receive multiple treatments or receive no treatment at all? These results support a previous study that found parents to be simultaneously in contact with multiple agencies within the same time period (Farmer et al., 2003). Farmer et al. (2003) examined mental health care utilization across sectors. The present study included a further analysis of services received within agencies and found that parents were not only typically in contact with over two agencies in the month prior to the interview, but they were also receiving up to three different treatments at each agency. Examining services received both across and within agencies provides a clearer picture of the challenges and complexities related to seeking and receiving mental health care for children.

A linear help-seeking process implies that parents progress from seeking help to receiving help. However, it would appear that parents followed a more disorganized and varied pathway, rather than a linear help-seeking pathway. Almost all parents used multiple services from multiple sectors within the same time frame, and they often contacted a new agency before concluding their involvement with a previous agency. Parents did not necessarily stop seeking help once they received treatment. Thus, families were simultaneously at multiple stages of the linear help-seeking process both across and within agencies.

Given the data from this study, one might question whether the amount of effort parents expend looking for help is reasonable. Parents contacted between one and 11 different agencies before receiving help (with an average of two agencies). To our knowledge, this is first study to document the number of contacts parents make prior to receiving treatment. We do not currently have an understanding of the impact that the help-seeking process has on families (e.g., parents’ stress) or the systems that care for these children (e.g., costs of multiple intake calls). Further, the number of agencies contacted to receive treatment must be examined against system resources (Costello, Burns, Angold, & Leaf 1993; Costello et al., 1998). For families in rural areas, contacting the school, physician and one agency might represent contacting 100% of all resources, whereas this might reflect only a fraction of potential resources in urban areas. A related question is whether receiving simultaneous treatments from an average of nearly three different agencies during the previous month is excessive. Depending on the nature and complexity of the child/family’s problems, receiving multiple treatments from multiple agencies may be warranted. However if families are receiving treatments from more than one agency for the same or a similar problem, treatment effectiveness is likely decreased, and family burden and costs increased.

Parents did not want just over one quarter of the treatments offered to them due to reasons other than logistical barriers. Neither a sub-group of parents, nor a specific treatment, accounted for the high percentage of parents who did not want the treatments offered. Interestingly, the majority of parents who did not want treatments still accepted the treatment that was offered. Parents who accepted these treatments could be demonstrating flexibility in that they were prepared to try treatments they did not initially want, or they could have been so frustrated that they were willing to take whatever help was offered. Treatment engagement and compliance are likely compromised for the latter parents. Parents who accept a treatment regime they do not want would be less likely to comply and more likely to drop-out (Kazdin, Holland, & Crowley, 1997). They may also continue to seek alternative treatments. This is not to suggest that agencies should only offer what parents what. Parents often have their own theory about what is needed to ‘fix’ their child’s aggression, non-compliance, poor school achievement, etc. However, professional assessment and subsequent treatment planning should result in the most effective treatment for child problems. Such treatment planning needs to incorporate parental perceptions on what might resolve their child’s problem. These data highlight the challenge that many agencies face of helping parents to develop a more comprehensive understanding of the multiple influences on the development and resolution of their child’s problems and to encourage an active role in resolving these problems.

The basis of parents’ acceptance or rejection of treatment needs further study. Parent attributions about their child’s problems, and related expectations that treatment should be child focused, have been examined in relation to treatment engagement in a few studies [see (Morrissey-Kane & Prinz, 1999) for a review]. Our data suggest the need to examine the broader notion of how parents conceptualize the cause, development and maintenance of their child’s problem (e.g., biological factors, parent, family, peer influences) and its relation to their help-seeking strategies and treatment acceptance.

Limitations

There are a number of limitations to this study. First, parents were recruited from one intake service for three children’s mental health centers that tend to focus on externalizing behaviour problems. Therefore, children with externalizing problems are likely over-represented beyond the general tendency that these children receive treatment more often than those with internalizing problems.

Second, the sample was obtained from a mid-sized city in Ontario that has multiple service options. Results would vary for rural populations and areas with fewer service options. They would also differ in United States for families that have insurance that covers both medical and psychological treatment. In Canada, only families with private health insurance benefits or high incomes tend to access private psychologists, and extended health insurance often covers only a fraction of the actual costs (Hunsley, Lee, & Aubry, 1999).

Third, data were obtained only from parents. Studies examining issues such as presenting problems have compared child, parent and therapist ratings of presenting problems (Yeh et al., 2001). Similar data cannot be obtained in relation to help-seeking as many times parents have only limited contact with a specific agency or professional. For example, they call looking for help only to be told that the agency contacted does not offer the services that parents are looking for, the child/family does not meet the agency criteria due to child age, or because they live outside the agency’s catchment area. Similarly, relying on parent report only does not capture the added complexity of seeking help for adolescents. Although the majority of time parents are involved in seeking help for their teenage children, adolescents may self-refer and they can have greater input in accepting or rejecting the treatment options that parents identify. The focus on parents is, however, appropriate given that parental perceptions including factors such as perceived burden are the strongest predictions of help seeking (Angold et al., 1998).

Fourth, a minority of families who called the intake service during the study period participated. We were unable to compare our sample to entire sample of potentially eligible participants as intake staff only record descriptive information for parents who will be receiving help from the centers. It is possible that the demographics of non-respondents differed from the demographics of this sample.

Fifth, a number of analyses comparing subgroups within the sample would be informative (i.e., comparing the data from parents who have accessed services for more than 1 year with parents who have accessed services for less than 1 year; comparing parents of children with more severe problems with parents of children with less severe problems). Given the sample size of 60, the authors did not want to conduct analyses on subsets of the data that were too small for meaningful comparisons. It would be beneficial to examine such issues in future research with larger samples.

In conclusion, despite the limitations, the study supports the need to understand help-seeking pathways as part of our knowledge of mental health care utilization and ways to improve accessibility and effectiveness of mental health care (Rogler & Cortes, 1993). A study utilizing community-sampling (Farmer et al., 2003) would be better able to document service paths in general and to confirm a particular help-seeking model. However, our data suggests that linear help-seeking processes involving problem recognition, the decision to seek help, seeking help and then receiving help, disregard the iterative referral process that many parents face and do not fully describe how parents interact with services that provide mental health care to children. Accurately describing help-seeking pathways for children’s mental health care is a precursor to implementing and evaluating system change (Morrissey, 1982).

The network-based model of access to children’s mental health services (Costello et al., 1998) and the gateway provider model (Stiffman et al., 2004) have begun to capture the complexity of accessing mental health care for children. These models can now be extended to include not only how parents enter the children’s mental health system, but also how parents navigate between the various services. This would begin to capture the iterative referral process that many parents are experiencing.