A majority of children and adolescents in the United States in need of mental health services do not receive them (1). Untreated child mental health problems result in significant morbidity, poor academic outcomes, increased criminal behavior, and other psychosocial dysfunction well into adulthood (2). Urban, low-income children have higher rates of treatment dropout and are less likely than their higher-income counterparts to receive needed mental health services. These socioeconomic disparities are prevalent even among children with attention-deficit hyperactivity disorder (ADHD), a chronic disorder with a well-established evidence-based medication treatment.
Extensive research has been conducted to better understand service disparities for ADHD treatment. Socioeconomic factors, health insurance status, and ethnic minority status explain some of the variation in the use of mental health services (3–6). However, less tangible factors, such as stigma, level of social support, and perceptions about mental health treatment, may have an equally powerful impact on families' use of mental health services and treatments (3,7–9). In addition, attitudinal and cultural factors influence ADHD treatment utilization specifically, for example, in medication compliance and appointment follow-up (5,10–12).
Although cultural and perceptual issues are known to be important in determining who gets care, it is not clear how these factors influence treatment retention. The primary aim of this study was to examine the experiences prior to the child's treatment of a group of parents and other caregivers whose child was diagnosed as having ADHD. The study explored correlations between the caregivers' experiences and subsequent treatment retention over a 12-month period.
Study design and recruitment
The study employed a mixed-methods, prospective design using qualitative data from in-depth interviews and quantitative data on ADHD treatment (13–15). Participants were recruited between November 2003 and December 2006 from pediatric primary care and specialty mental health clinics affiliated with a large teaching hospital in Baltimore. Participants were eligible if their child was six to 18 years old and newly diagnosed as having ADHD. The study procedures are described in detail elsewhere (13–15). Clinicians identified eligible participants and referred them to the investigators, who explained the study protocol and obtained written consent from the adult and assent from the child. Information about demographic characteristics, family psychiatric history, and treatment at three points (at the time of study recruitment and at six- and 12-month follow-ups) were obtained from the child's medical chart. Institutional review board approval was obtained.
Of the 69 individuals approached, 48 (70%) agreed to participate and were interviewed. Most (N=39, 89%) were recruited from primary care clinics, and nine (19%) were recruited from specialty mental health clinics, a distribution that reflects the fact that most initial diagnoses of ADHD occur in primary care settings.
Semistructured interviews ranging from 30 minutes to more than an hour were conducted by one author (SDR) and trained research assistants. Interviewers followed a field guide containing questions about how the child came to be diagnosed as having ADHD, caregivers' perceptions of ADHD before seeking treatment, their expectations for treatment, and their perception of their involvement in the child's mental health care. Participants described in detail their experiences from the time they first became concerned or were first told about their child's behaviors to the time of the initial diagnosis. All interviews were transcribed verbatim.
A manual for identifying themes underlying the caregivers' experiences with their child's ADHD before they sought care was developed on the basis of an analysis of the interview transcripts. Following the tradition of grounded theory methods, the analysis reflected caregivers' perspectives and not those of the researchers. Thus the analysis was not approached from a predetermined theory to avoid introducing researcher bias.
Initially, researchers independently read the first five interviews to identify passages relevant to what led caregivers to seek care for their child. The key to the analysis was to identify the meaning of each passage and to capture all other passages in the interview that related to the same theme. The research team discussed the theoretical dimensions of each theme and through a consensus process developed a coding manual that defined each thematic code. Next, another set of five interviews was analyzed using the coding manual. Any new codes that emerged in the second set of interviews were added to the manual on the basis of consensus of the researchers. Saturation was reached when no new codes were identified with each subsequent interview.
In this manner, themes were theoretically derived from the data, and each coder achieved a clear understanding of the theme. A coding report was generated for each interview by each research team member. Agreement in coding was assessed by comparing the reports to determine whether the same passages were identified. Through the consensus process and coding agreement discussions, 100% agreement was achieved.
Data were visually displayed to assess the different ways in which themes were endorsed by the caregivers as individuals and as compared with each other. The research team identified a set of patterns in the way themes related to each other and in the way individuals endorsed them, called “threshold patterns.” The team discussed each pattern to distinguish among the defining characteristics. Where disagreements occurred, salient aspects of each pattern were revisited, clarified, and redefined as appropriate. To ensure trustworthiness of the data, the themes and patterns were reviewed by a clinician not directly involved with the data, who then provided feedback on the face validity and credibility of the findings.
Fisher's exact tests were used to examine associations between characteristics of the sample and threshold patterns. Dichotomous measures (“yes” or “no” responses) were used to determine treatment retention—the participant was still attending clinic appointments—and medication use—a prescribed medication was noted in the child's medical chart. NVivo was used to code the data and organize reports of all coded text; SAS version 9.2 was used for all statistical analyses.
The 48 study participants included biological mothers (75%), biological aunts or grandmothers (10%), fathers (8%), and stepparents (7%). Most were single parents living in Baltimore City with at most a high school-level education. As shown in Table 1, children averaged nine years of age. Sixty-three percent were male, 81% were African American, and 56% were enrolled in the state medical assistance program. Fifty-six percent had at least one comorbid diagnosis, including adjustment (N=11, 23%), learning (N=8, 16%), disruptive (N=5, 10%), and depression or anxiety (N=7, 14%) disorders. About half (54%) had a family history of psychiatric illness. One-third had a relative with a substance use disorder (N=16, 33%), and 23% (N=11) had a relative with depression. Medication for ADHD was prescribed shortly after diagnosis for 83% (N=40) of the children.
Caregivers' experiences with their child's ADHD
Six themes underlying caregivers' experiences with their child's ADHD emerged from the data. They included caregiver strain, attribution of meaning, perception of responsibility, problem-solving approach, ADHD treatment beliefs, and response to societal influences. Multiple themes were endorsed in unique ways by each caregiver, but some themes were more predominant than others. The themes represent the breadth of experiences reported by the caregivers and could function as either catalysts or impediments to care seeking.
Caregiver strain was an expression of the distress resulting from the child's behaviors. One caregiver, noting the strain from constant calls from the school, said, “We go through this struggle every day and get these notes every day and these phone calls every day.” Others were not distressed, saying, “If you didn't come from a background with a large family you would be really stressed out .… I'm used to the boys fighting.” Changes in life circumstances, such as having additional children, affected caregivers' tolerance for their child's behaviors and symptoms.
Attribution of meaning referred to how caregivers explained the symptoms, reflecting a deeper insight into caregivers' initial receptivity to getting care for their child. Some explained the child's behavior as an innate characteristic by saying, for example, “Ms. T. recommended [clinic name] to see if he had attention deficit … but I'd always just thought that that was [child's name]!” or “I was just assuming that he was just … bad.” Some caregivers accepted the diagnosis, whereas others offered alternative explanations for behaviors by saying, for example, “I really think she just wants my attention because she is the only girl.”
Perception of responsibility reflected caregivers' view of what needed to be done to help their child. One caregiver stated, “I'm going to make sure she gets the best help she might need.” Others felt their responsibility was to protect the child from being medicated (“He can't speak because he's a child, so someone had to speak for him”). Some placed responsibility with “experts” such as school officials or physicians (“I don't know because I'm not an ADHD expert. … They should get a tutor for him”).
Problem-solving approach reflected how caregivers organized themselves to get help for their child. Some observed the problem, identified a potential solution, and sought help in a timely fashion. Others were problem avoidant, saying, for example, “I kind of ignore it to be honest with you,” or “he's the age where it's harder for me to contain him like my younger children … so I send him away to get a break.”
ADHD treatment beliefs were shaped by past experiences or prior knowledge of medication appropriateness, efficacy, and side effects. These beliefs either encouraged caregivers to seek treatment or discouraged them from seeking treatment. For example, one caregiver stated, “I know a few people whose kids are on [medication] that [are] real sullen and withdrawn, and I didn't want that to be my child.” One caregiver's past experience with her own use of medication (“Before I took medication I didn't believe in ‘head medicine‘ ”) influenced her decision to use medication for her child.
Response to societal influences was also prominent. Some caregivers relied on teachers to “basically just tell [them] what was going on.” Others were confused by conflicting influences. Even as clinicians suggested medication treatment for the child, one said, “Her godmother [told me], ‘It's nothing wrong with her. … You know doctors always say something is wrong with a child and … put [them] on medicine.' ”
Patterns of caregiver theme endorsement
Despite the caregivers' unique situations, the themes endorsed by individuals were classified into four distinct patterns that reflected individuals' experiences leading up to their decision to seek care. They were motivated by observation (N=13, 27%), motivated by strain (N=11, 23%), motivated by experience (N=9, 19%), and struggling with meaning (N=15; 31%).
The group described as motivated by observation clearly perceived their child's behavior as a problem requiring professional attention. The distinguishing characteristics of this group were their systematic problem-solving approaches, their strong sense of responsibility, and their relatively low strain. These caregivers took primary responsibility for the child's success and partnered with professionals to obtain the care the child needed.
The group described as motivated by strain was characterized by disruption, exhaustion, uncertainty, and frustration attributed to their child's behavior. They had a strong sense of responsibility for the child but were uncertain of how to get help and often received mixed advice from others. This situation was a major source of stress. When the diagnosis of ADHD was made, it was generally accepted and received with relief.
The group labeled “struggling with meaning” reflected caregivers who offered alternative explanations for the behaviors that were diagnosed as characteristic of ADHD. Although some had agreed to treatment, their alternative explanation for the child's behavior and adversarial interactions with schools and clinicians were prominent. In contrast to those motivated by strain, these families were strained by the pressure from teachers and doctors to use ADHD medication.
The group considered motivated by experience were those who had greater knowledge of ADHD and were more comfortable with medication and management of their child's symptoms. These families' knowledge allowed them to identify the behavior problems as secondary to the ADHD. Consequently they did not reach a point of strain or distress with the ADHD behaviors.
Demographic and treatment characteristics
As displayed in Table 1, the children of individuals motivated by experience were more likely to be girls (78%), compared with children of the other groups of caregivers (p<.007). The children of caregivers who were motivated by observation were the least likely to have comorbid psychiatric diagnoses (38%), compared with children of caregivers who were struggling with meaning (67%) and children of caregivers motivated by strain (64%) (p<.01). In addition, a family psychiatric history was more likely to be reported for children whose caregivers were motivated by observation (69%) and experience (78%) than for children whose caregivers were motivated by strain (36%) or were struggling with meaning (40%) (p<.01). The children's race, age, insurance status, living situation, and source of ADHD referral were similar across groups.
Table 2 shows correlations between the caregiver groups and clinic attendance and medication use. The groups were not different in terms of their medication use at baseline and six months and 12 months later. At 12 months, the children of caregivers struggling with meaning were least likely to be attending clinic visits (53%), compared with children of caregivers motivated by experience (89%), motivated by observation (77%), and motivated by strain (73%) (p<.017).
The parenting experiences recalled by caregivers of children with ADHD were correlated with whether they stayed in treatment. Caregivers in the threshold group labeled “struggling with meaning” constituted the largest proportion of the sample (31%), yet they were the most likely to drop out of treatment.
Other studies have demonstrated that problem recognition and perceived caregiver burden have greater importance in determining service utilization than do socioeconomic factors and the nature and severity of a child's symptoms (6,16–18). Prior research on the dynamic interplay of caregiver factors that lead to service use has been conducted (9). A better understanding of the heterogeneity of these factors across individual caregivers could lead to more effective implementation of evidence-based reatments.
The analysis supports the literature that child-specific factors alone, such as the presence of comorbidities or severity of symptoms, do not drive caregivers' utilization of services for their child. The children of caregivers who were struggling with meaning were on average two years older and were more likely to have comorbid psychiatric diagnoses compared with children in the other groups. Although there was no formal measure of mental illness severity, many of the caregivers in this group reported extreme behavior problems and significant psychosocial dysfunction of their children, such as multiple repeated grades, stealing, and significant aggression. In contrast, in the group motivated by observation, a lower rate of psychiatric comorbidity among the children (38%) corresponded with reports of more minor symptoms, such as attention problems and distractibility.
Despite the fact that children whose caregivers were struggling with meaning had higher levels of comorbid disorders, their clinic attendance (53%) and medication use (40%) were low after 12 months. In contrast, children whose caregivers were motivated by experience had both high rates of medication use (78%) and treatment attendance (89%) after 12 months. Thus interventions to enhance treatment retention may be improved by efforts to more deeply understand caregivers' thoughts and experiences as they relate to service use.
Stigma regarding ADHD and its treatment, for example, has been shown to be highly prevalent (4,8,19) and may influence whether families decide to remain in care. Notably a larger proportion of the caregivers motivated by observation (69%) and motivated by experience (78%) had a psychiatric family history, compared with caregivers who were struggling with meaning (40%) and motivated by strain (36%). Familiarity with mental health services may have reduced stigma in these groups, possibly facilitating service use.
The strain felt by caregivers emerged as an important factor as they described their experiences leading up to seeking care. The relationship between caregiver strain and service utilization is well documented (6,20–23). This study, however, addressed the various sources of caregiver strain, such as the direct effect of the child's behaviors, the frustration of dealing with school personnel or health care professionals, the distress associated with not knowing why the child is misbehaving, and the distress about how to access resources. Although high levels of strain were reported by caregivers who were motivated by strain and by those struggling with meaning, a larger proportion of the group motivated by strain remained in treatment (73% versus 53%). It is possible that certain types of strain may be a moderator of treatment retention. Examining the effect of different types of strain on retention and adherence to child mental health services is important area for future research.
This study had several limitations, including a sample size too small to permit quantitative inferences. The potential also exists for a selection bias, in that the beliefs and expectations in the sample may not generalize to caregivers in other sociocultural groups, geographic areas, or medical centers. The sample also did not represent caregivers who do not seek care for their children. Further research will help us to understand how caregivers from other socioeconomic or cultural backgrounds differ from the study participants in their experiences of strain, in their explanation for their child's symptoms, and in the degree to which others influence their decisions to seek care. Even though the individual experiences may not be generalizable, the conceptual framework of the study is likely to be transferable to other samples.
An additional weakness of the study was a lack of a direct measure of symptom severity to quantify the relationship between severity and the decision to seek care and to remain in treatment. Qualitative methods, however, by assessing the caregivers' capacity to manage their child's ADHD and their perception of symptom severity, provide a deeper understanding of what motivates caregivers to use services.
Despite these concerns, this study highlights the importance of caregiver-related factors in the retention of children in mental health treatment. In the initial evaluation of a child suspected of having ADHD, it may be important for clinicians to elicit the extent to which caregivers feel a sense of strain as well as their expectations for care and their explanation for their child's symptoms. These factors can be used to identify subgroups at high risk of treatment attrition and to develop targeted interventions early in care that may ultimately maximize treatment engagement. A few interventions conducted early on to increase adherence to treatment by exploring caregivers' concerns and expectations for treatment have reported a positive impact on future treatment retention (24,25). The identification of modifiable risk factors for minimizing treatment delay and attrition in high-risk subgroups may result in more effective care and reduced psychosocial morbidity for children and their families.