Is the focus of community-based mental health treatment consistent with adolescent psychiatric diagnoses?

Abstract

Most adolescents do not receive effective mental health services. This may stem in part from infrequent use of evidence-based and multi-informant diagnostic assessments to guide clinical care. The primary purpose of the present study was to examine whether adolescent mental health diagnoses and suicidality, derived via evidence-based diagnostic interviews and assessments, correspond with reported “reason for treatment” received by adolescents. Secondarily, we examined the potential association between socio-economic status and the match between youth diagnoses and reasons for treatment. The influence of parent-adolescent agreement on diagnoses and reasons for treatment on findings was also explored. Using chi-square analyses, a significant association was found between youth diagnoses of mood disorders, disruptive behavior disorders, and suicidality, respectively, and a focus of treatment on these conditions per combined parent-adolescent report. The same was not true for youth anxiety, attention-deficit hyperactivity, or substance abuse disorders. Results of exploratory analyses suggest that these results are driven by adolescent, but not parent report. With regard to socio-economic status, there was a trend for those with higher incomes to report a treatment focus consistent with youth diagnoses, per combined parent-adolescent report. Results suggest that focus of mental health treatment received by adolescents in standard community-based care may not uniformly address all current disorders. Efforts are needed to disseminate multi-informant evidence-based assessments to enhance the quality and effectiveness of care.

1. Introduction

Extant literature indicates that most adolescents do not receive mental health services when in need. One in every 4–5 adolescents in the United States meets criteria for a mental health disorder, and about half of those adolescents experience severe distress and impairment due to their disorder (Kessler et al., 2009; Merikangas et al., 2010). However, on average, there is a nine-year delay between the onset of youth mental health issues and linkage with mental health treatment (Kessler et al., 2009) and nearly half of adolescents with mental health issues do not receive treatment services at all (Center for Behavioral Health Statistics and Quality, 2013). Moreover, even when mental health services are received, only 2% of youth receive evidence-based services, or treatment supported by research (Merikangas et al., 2010). This includes the use of evidence-based diagnostic assessments which have been shown to influence youth treatment outcomes (Jensen-Doss & Weisz, 2008; Hoagwood et al., 2008). When youth are not properly diagnosed, they may fail to receive services consistent with their needs.

Evidence based assessment, or the use of standardized assessment tools with adequate psychometric properties (Jensen-Doss & Hawley, 2010), are an important component of evidence-based practice. However, the standard practice in community-based mental health settings is to use an unstructured clinical interview to derive diagnoses. Unstructured clinical interviews, relative to structured diagnostic interviews, have been shown to be less accurate, less likely to identify co-occurring conditions, have limited empirical support, and are subject to biases that decrease assessment accuracy (Anderson & Paulosky, 2004; Cashel, 2002; Garb, 2005; Mash & Hunsley, 2007; Sheehan et al., 1998). Yet, due in part to practical obstacles (e.g., training needs, time for administration, insurance reimbursement, etc.) (Camara et al., 2000; Mash & Hunsley, 2007; Youngstrom et al., 2015) as well as a preference for practicing assessment in a clinically intuitive manner consistent with previous training (Garland et al. 2003), validated semi-structured/structured diagnostic interviews are rarely used.

Similar to the use of standardized diagnostic assessment, the degree to which both parent and adolescent report are incorporated into diagnostic formulation may also vary in clinical practice. Yet, the need for multi-informant assessment when deriving youth diagnoses has been well established (De Los Reyes, 2016; De Los Reyes et al., 2017). Relying on one informant’s report can lead to different diagnostic conclusions than using a consensus rating from multiple informants (De Los Reyes & Kazdin, 2005). Indeed, parent and child ratings of social, emotional, and behavioral issues are often discrepant (Achenbach et al., 1987; De Los Reyes & Kazdin, 2005), and this gap only increases from childhood through adolescence (Achenbach et al., 1987). This is also true of high-risk behaviors associated with psychiatric diagnoses, such as suicidality (Klaus et al., 2009; Lewis et al., 2014). In general, adolescents tend to report more internalizing symptoms (i.e., mood and anxiety) than their parents, in part, because these symptoms may go unnoticed (Rothen et al., 2009). In contrast, parents may be more likely to pick up on externalizing problems (i.e., conduct, hyperactive-impulsive, etc.) which are outwardly expressed. Thus, it is not surprising that parent-child agreement tends to be low for ratings of internalizing disorders, but moderate for externalizing disorders (Salbach-Andrae et al., 2009). With regard to substances, research suggests that parents are often knowledgeable about their adolescent’s mild use of substances (i.e., infrequent use of substances that does not meet criteria for a diagnosis), but not about more serious substance-related problems, such as alcohol and marijuana abuse and dependence (i.e., meets diagnostic criteria) (Fisher et al., 2006).

Relatedly, parents and their children fail to agree on a single problem to target during the child’s treatment more than half the time (Yeh & Weisz, 2001). These discrepancies between informants may speak to larger issues and yield important information to consider in treatment planning (De Los Reyes & Kazdin, 2005). The primary purpose of the present study is to examine the degree to which reported focus of mental health received in the community is consistent with adolescent psychiatric diagnoses, derived via the use of standardized diagnostic assessments completed by adolescents and parents.

Receipt of quality care, defined in the present study as treatment that is tailored to a patient’s diagnosis, may also be affected by socio-economic status. A fair amount of research suggests that health insurance coverage and SES may negatively affect parental ability to gain access to high quality mental health services for their children (Kataoka et al., 2002; Newacheck et al., 2003; Wang et al., 2000). Families of adolescents from low SES backgrounds and those with limited/no health insurance coverage may not be able to afford any, let alone quality, mental health care (Flisher et al., 1997; Kataoka et al., 2002; Newacheck et al., 2003). Rather, they tend to receive services that are inadequate or inappropriate for their needs (Bui et al., 1992; Newacheck et al., 2003; Wilk et al., 2005) and face challenges in accessing those services (e.g. therapists with less training and higher caseloads, disparities in wait times, inaccessible services, cultural and language barriers) (Bui et al., 1992, Bisgaier & Rhodes, 2011). A secondary aim of the present study is to examine the degree to which these socio-economic factors are associated with receipt of quality care, or in this case, treatment reportedly focused on youth diagnoses.

In summary, this is the first study to explore correspondence between parent and adolescent report of psychiatric diagnosis and focus of treatment in community-based mental health care. Based on extant literature, it was hypothesized that per combined adolescent and parent report: 1) there will be a significant association between the presence of psychiatric disorders most likely to be brought to the attention of clinicians by parents (e.g., disorders associated with externalizing behavior) and focus of treatment for these disorders, but not for disorders associated with internalizing behavior or substance abuse; and 2) there will be a positive association between the presence of psychiatric diagnosis-focus of treatment match (i.e., diagnosis corresponds with focus of treatment) and income level as well as private (vs. private) insurance. Parent-adolescent agreement on psychiatric diagnoses and reasons for treatment are also reported here, as well as exploratory analyses conducted to examine whether study findings vary as a result of reporter status (parent and adolescent). In the changing landscape of insurance coverage, public assistance for mental health treatment, and socio-demographics in the United States, it is particularly timely and important to examine whether focus of care corresponds with existing mental health problems among our most vulnerable youth.

2. Method

2.1. Participants

Participants were drawn from a parent study that tested an adjunctive substance abuse, suicide, and HIV prevention program for youth in mental health treatment (please see REMOVED FOR BLINDING, 2017 for additional details on the parent study). Participants included 81 adolescents and a parent/guardian who resided with the adolescent. They ranged in age from 13–17 years old (M_age = 15.4, SD = 1.4) and were predominantly female (58%). The sample was approximately 42.0% White, 37.0% Black/African-American, and 81.5% non-Hispanic. Participants were receiving mental healthcare in the community at the time of study enrollment and were recruited via clinical referral. The majority of parents/guardians who participated in the study were female (90.1%), biological mothers (74.1%), and over the age of 40 (M_age = 43.9, SD = 7.5). The yearly household income ranged from less than $20,000 per year (11%) to greater than $100,000/year (33.3%) (median = 50,001–60,000). Adolescents were eligible for the study if they were: 1) 13–17 years old; 2) receiving mental health treatment in the community (therapy or medication); 3) residing with a parent/guardian who could also participate in the study; and 4) spoke English as their primary language. Participants were ineligible if they were unable to provide assent due to cognitive limitations, psychotic, homicidal, or substance dependent at the time of study entry. Data were collected from 2010–2014. This study was approved by the affiliated University Human Subjects Review Board.

2.2. Measures

Standard demographic data were collected, including age, sex, race, ethnicity, family income, and insurance type.

2.2.1. The Child and Adolescent Services Assessment (CASA-Parent and Adolescent Version; Farmer et al., 1994).

The CASA is a semi-structured interview that assesses mental health services utilized, type of treatment facility, professional discipline of providers, outpatient services, and reasons for treatment. Adequate inter-rater reliability for the administration of the CASA by trained study staff has been demonstrated (ICCs=.74-.76) (Farmer et al., 1994). Treatment received between study enrollment and the 6-month follow-up was assessed via independent adolescent and parent report. Variables for “services received” and “reason for treatment” codes (e.g. drug, alcohol, depression) were included in study analyses.

2.2.2. Diagnostic Interview Schedule for Children 4.0 (DISC-4.0 - Parent and Adolescent Report; Shaffer, 1991a, 1991b).

The DISC-4.0 is a computer-assisted structured diagnostic interview that is administered separately to youth and parents (Shaffer, Fisher, Lucas, Dulcan, & Schwab-Stone, 2000) and generates DSM-IV diagnoses (see information on specific diagnoses in section 2.4). Test-retest reliability is moderate to good for parent (kappa (κ) = 0.43–0.96) and adolescent (κ = 0.42–0.92, with the exception of social phobia, κ = 0.25) reports (Shaffer et al., 2000). Criterion validity has not been examined for the DISC-4.0. There are four response options on the DISC-4.0: “no,” “yes,” and “sometimes” or “somewhat.” The DISC-4.0 uses a computerized scoring program to designate whether enough DSM-IV diagnostic symptoms are present for a diagnosis, per parent and adolescent report, respectively (Shaffer et al., 1996).

2.2.3. Revised Self-Injurious Thoughts and Behaviors Interview 2.0 - Short Form (SITBI 2.0-SF; Nock et al., 2007).

The 90-item SITBI 2.0 - SF assesses suicidal ideation, plans, gestures, and attempts, as well as thoughts about non-suicidal self-injury. The SITBI has been shown to have excellent psychometric properties (Nock et al., 2007). The items that assess presence of suicidal ideation and suicide attempts, respectively, in the past month were used. Responses are dichotomous (0= no suicide ideation or attempt and 1 = the presence of suicidal ideation or an attempt).

2.2.4. Suicide Ideation Questionnaire (SIQ) (Reynolds, 1987).

The SIQ is a 30-item youth self-report instrument that assesses thoughts about suicide in the past month. Adolescent scores on the SIQ have demonstrated excellent internal consistency (α=.97) and test-retest reliability (r =.72) (Reynolds, 1987). The SIQ has been supported as a criterion-valid measure of suicidal ideation (Reynolds, 1998). Scores ≥ 41 denote clinically significant suicidal ideation (Reynolds, 1987).

2.3. Procedures

Adolescents were recruited into the parent study via clinical referral from community-based facilities (43% from a youth shelter, 27% community mental health centers, 10% private practices, 10% juvenile court diversion program, 2% school counselors, 3% unknown) as well as advertisements (10%). Specifically, staff at recruitment sites provided families with a brochure about the study. If interested, families provided their contact information on a consent to contact form that was shared with study staff. A research assistant then contacted the family and read from a script that included a study overview and eligibility screener. If the family agreed, they set up an appointment for consenting and baseline assessment. Those self-referred contacted study staff directly.

2.4. Data Analysis

Psychiatric diagnosis and suicidality were assessed at the baseline assessment. Mental health care received, and perceptions of the focus of this care, between the baseline assessment and the following 6 months was assessed at the 6-month follow-up. Data were available for 81 adolescents at baseline and 47 at 6 month-follow-up. Twenty-two adolescents were lost to follow-up and 12 reported no further mental health treatment post-baseline assessment and thus could not be included in analyses. There were no significant differences on socio-demographic variables or rates of diagnoses between those adolescents who were and were not retained in 6-month follow-up analyses (all p’s > .05).

Data were analyzed using descriptive statistics, Kappa, chi-square analyses, and point-biserial correlation tests in SPSS. The DISC-4.0 psychiatric diagnoses examined in the present study overlap with reasons for mental health treatment (outpatient and inpatient) coded in the CASA interview. Psychiatric diagnoses were collapsed into six treatment categories or codes included in the CASA: substance use (alcohol, marijuana, and other substance abuse), anxiety (panic, generalized anxiety, social phobia, posttraumatic stress disorders), depression/mania [referred to as mood disorder throughout] (major depressive, manic, and hypomanic episodes as well as dysthymia), attention-deficit hyperactivity disorder (ADHD), and conduct problems [referred to as Disruptive Behavior Disorder (DBD) throughout] (oppositional defiant and conduct disorders). Suicidal thoughts and suicide attempts over the prior month, assessed via the SIQ and SITBI, corresponded with the suicide code in the CASA. Each disorder as well as the presence of suicidality reported in the sample was evaluated for a match with a reported reason for treatment (or CASA code).

The association between diagnosis and reason for treatment were examined using a series of chi-square analyses. Phi was used to explain the effect size for these outcomes (.10 = small; .30 = medium; and .50 = large effect sizes) (Cohen, 1988). The association between family income as well as insurance status (public insurance vs. private insurance) and match between diagnosis and reason for treatment (0 = no match, 1 = match) was examined using a point-biserial correlation and a chi square test, respectively. Of note, cases were coded as a match as long as at least one diagnosis matched a reported reason for treatment. Cases were coded as a non-match if no diagnosis matched a reported reason for treatment. Degree of agreement between informants (parent and adolescent) on psychiatric diagnoses and reasons for treatment were examined using the Kappa statistic ( < 0 = no; 0–0.20 = slight; 0.21–0.40 = fair; 0.41–0.60 = moderate; 0.61–0.80 = substantial; and 0.81–1 = almost perfect agreement; Landis & Koch, 1977).

3. Results

3.1. Descriptive Analyses

Rates of psychiatric diagnoses in the present sample can be found in Table 1. As can be seen, there were high rates of psychiatric comorbidity in the present sample. Twenty-two percent (n =18) of adolescents reported suicidal ideation and four percent reported a suicide attempt (n =3) in the prior month. Of the 47 adolescents who reported receipt of mental health care in the 6 months post-baseline, 85% reported outpatient treatment alone, 0% inpatient treatment alone, and 15% outpatient and inpatient treatment. Further, 51% of adolescents were taking psychiatric medication and all of these adolescents were receiving concurrent therapy.

Table 1.

Percentages of Diagnoses and Comorbid Conditions in Study Sample at Baseline (N=81)

Disorders	% with disorder	Comorbid DBD	Comorbid Mood	Comorbid Substance	Comorbid ADHD
Anxiety	27% (n = 22)	17% (n =14)	15% (n = 12)	6% (n = 5)	7% (n = 6)
DBD	53% (n = 43)	---	23% (n = 19)	15% (n = 12)	21% (n = 17)
Mood	42% (n = 34)	---	---	9% (n = 7)	11% (n = 9)
Substance	25% (n = 20)	---	---	---	9% (n = 7)
ADHD	26% (n = 21)	---	---	---	---

Note. DBD = Disruptive Behavior Disorders, ADHD = Attention-Deficit Hyperactivity Disorder

3.2. Parent and Adolescent Ratings of Diagnoses

Cross-tabulations between parent and adolescent report of adolescent diagnoses produced four possible categories, including: agreement on the presence of the disorder, agreement on the absence of the disorder, disagreement where only the parent endorsed the presence of the disorder, and disagreement where only the adolescent endorsed the presence of the disorder (see Table 2). The majority of parent and adolescent pairs agreed about the presence or absence of a mood disorder (71.6%; κ = .30, p = .01; fair agreement), a DBD (61.7%; κ = .22, p = .02; fair agreement), and a substance abuse disorder (81.3%; κ = .31, p = .01; fair agreement). There was not significant agreement found between adolescent and parent report of the presence or absence of an anxiety disorder (77.7%, κ = .18, p = .12) or ADHD (76.6%, κ = .09, p = .32).

Table 2.

Parent-Adolescent Agreement on Rates of Adolescent Psychiatric Disorders (N=81)

	Anxiety	DBD	Mood	Substance	ADHD
Agreement
Presence of diagnosis	4.9%	14.8%	13.6%	6.3%	2.5%
Absence of diagnosis	72.8%	46.9%	58.0%	75.0%	74.1%
Disagreement
Only parent endorsed	11.1%	33.3%	17.3%	2.5%	19.8%
Only adolescent endorsed	11.1%	4.9%	11.1%	16.3%	3.7%
Kappa	.18	.22	.30	.31	.09
p	.12	.02	.01	.01	.32
Association Size	Slight	Fair	Fair	Fair	Slight

DBD = Disruptive Behavior Disorders, ADHD = Attention-Deficit Hyperactivity Disorder

3.3. Parent and Adolescent Ratings of Reasons for Treatment

Cross-tabulations between parent and adolescent report of reasons for treatment produced four possible categories, including: agreement on the reason for treatment, agreement on the absence of the reason for treatment, disagreement where only the parent endorsed the reason for treatment, and disagreement where only the adolescent endorsed the reason for treatment (see Table 3).

Table 3.

Parent-Adolescent Agreement on Rating of Reasons for Treatment (N=47)

	Anxiety	DBD	Mood	Substance	ADHD	Suicide
Agreement
Presence of Reason for Tx	2.1%	6.4%	10.6%	4.3%	10.6%	8.7%
Absence of Reason for Tx	76.6%	72.3%	55.3%	85.1%	76.6%	60.9%
Disagreement
Only parent endorsed	6.4%	17.0%	31.9%	6.4%	10.6%	13.0%
Only adolescent endorsed	14.9%	4.3%	2.1%	4.3%	2.1%	17.4%
Kappa	.06	.27	.23	.39	.55	.16
p	.66	.04	.03	.01	.00	.26
Association Size	Slight	Fair	Fair	Fair	Moderate	Slight

DBD = Disruptive Behavior Disorders, ADHD = Attention-Deficit Hyperactivity Disorder

The majority of parent and adolescent pairs agreed that a mood disorder (65.9%; κ = .23, p = .03; fair agreement), DBD (78.7%; κ = .27, p = .04; fair agreement), substance abuse disorder (89.4%; κ = .39, p = .01; fair agreement), and ADHD (87.2%; κ = .55, p = .00; moderate agreement) was or was not a focus of treatment. Parent and adolescent agreement on whether an anxiety disorder (78.7%; κ = .06, p = .66) or suicidality (69.6%, κ = .16, p = .26) was or was not a focus of treatment was not significant.

3.4. Diagnosis and Reasons for Treatment - Combined Parent and Adolescent Report

Cross-tabulations between diagnoses and reason for treatment (per adolescent and/or parent report) produced four possible categories, including: presence of a diagnosis and matched reason for treatment; absence of a diagnosis and matched absence of reason for treatment; presence of a diagnosis and no matched reason for treatment; and absence of a diagnosis but reported focus of treatment was on the absent diagnosis (see Table 4). We examined these scenarios separately for each of the 5 disorders and suicidality.

Table 4.

Association Between Adolescent Diagnoses and Reason for Community-Based Treatment (Combined Parent-Adolescent Report) (N=47)

	Anxiety	DBD	Mood	Substance	ADHD
Agreement
Diagnosis (yes) / Focus (yes)	10.6%	21.3%	31.9%	6.4%	4.3%
Diagnosis (no) / Focus (no)	48.9%	40.4%	36.2%	68.1%	53.2%
Disagreement
Diagnosis (yes) / Focus (no)	27.7%	34.0%	19.1%	17.0%	23.4%
Diagnosis (no) / Focus (yes)	12.8%	4.3%	12.8%	8.5%	19.1%
χ2	0.31	5.12	6.30	1.74	0.66
P	.58	.02	.01	.19	.42
Phi	.08	.33	.37	.19	−.12
Effect Size	Small	Medium	Medium	Small	Small

DBD = Disruptive Behavior Disorders, ADHD =Attention-Deficit Hyperactivity Disorder

There was a significant association between diagnosis of mood disorders as well as DBDs and associated reasons for treatment. Specifically, 62.5% of adolescents with a mood disorder reported a treatment focus on a mood disorder whereas 73.9% of adolescents without a mood disorder did not report a treatment focus on a mood disorder, χ² (1) = 6.30, p = .01. Similarly, 38.5% of adolescents with a DBD reported a treatment focus on DBDs whereas 90.5% of adolescents without a diagnosis of DBD did not report a treatment focus on DBDs, χ² (1) = 5.12, p = .02. These effects were of medium size for mood disorder, Φ = .37, and DBD, Φ = .33, analyses. There was not a significant association found between diagnoses of anxiety disorders, χ² (1) = 0.31, p = .58, substance abuse disorders, χ² (1) = 1.74, p = .19, or ADHD, χ² (1) = 0.66, p = .42, and corresponding reasons for treatment (see Table 4). These effects were small for anxiety disorder, Φ = .08, substance abuse disorder, Φ = .19, and ADHD, Φ = −.12, analyses.

3.5. Exploratory Analysis: Diagnoses and Reasons for Treatment - Adolescent Report Only

There was a significant association between diagnosis of mood disorder as well as DBD and associated reasons for treatment. Specifically, 28.6% of adolescents with a mood disorder reported a treatment focus on a mood disorder whereas 93.9% of adolescents without a diagnosis of mood disorder did not report a treatment focus on a mood disorder, χ² (1) = 4.47, p = .01. Similarly, 27.3% of adolescents with a DBD reported a treatment focus on a DBD whereas 94.4% of adolescents without a DBD did not report a treatment focus on a DBD, χ² (1) = 4.18, p = .04. These effects were of medium size for mood disorder, Φ = .31, and DBD, Φ = .30, analyses. There was no significant association found between diagnoses of anxiety disorders, χ² (1) = 2.15, p = .14, substance abuse disorders, χ² (1) = 0.01, p = .94, or ADHD, χ² (1) = 1.22, p = .27, and corresponding reasons for treatment (see Table 5). These effects were small for anxiety disorder, Φ = .21, substance abuse disorder, Φ = .01, and ADHD, Φ = .16, analyses.

Table 5.

Association Between Adolescent Diagnoses and Reason for Community-Based Treatment (Adolescent Report Only) (N=47)

	Anxiety	DBD	Mood	Substance	ADHD	Suicide +
Agreement
Diagnosis (yes) / Focus (yes)	4.3%	6.4%	85.0%	2.1%	2.1%	17.0%
Diagnosis (no) / Focus (no)	74.5%	72.3%	66.0%	70.2%	83.0%	55.3%
Disagreement
Diagnosis (yes) / Focus (no)	17.0%	17.0%	21.3%	21.3%	4.3%	6.4%
Diagnosis (no) / Focus (yes)	4.3%	4.3%	4.3%	6.4%	10.6%	21.3%
χ2	2.15	4.18	4.47	0.01	1.22
p	.14	.04	.01	.94	.27	.01
Phi	.21	.30	.31	.01	.16
Effect Size	Small	Medium	Medium	Small	Small

Note.

⁺Fisher’s exact test

DBD = Disruptive Behavior Disorders, ADHD =Attention-Deficit Hyperactivity Disorder

Expected cell counts below 5 in analyses related to suicidality precluded the interpretation of chi square tests. Instead, Fisher’s exact test is reported. There was a significant association between suicidality and associated reasons for treatment. Specifically, 44.4% of adolescents with suicidality reported a treatment focus on suicidality whereas 89.7% of adolescents without suicidality did not report a treatment focus on it, p = .01.

3.6. Exploratory Analysis: Diagnosis and Reasons for Treatment - Parent Report Only

There was no significant association found between diagnoses of anxiety disorders, χ²(1) = 0.73, p = .39, DBD, χ² (1) = 0.91, p = .34, mood disorders, χ² (1) = 1.33, p = .25, substance abuse disorders, χ² (1) = 1.74, p = .19, or ADHD, χ² (1) = 0.21, p = .65, and corresponding reasons for treatment (see Table 6). The effects were small for anxiety disorder, Φ = .12, DBD, Φ = .14, mood disorder, Φ = .17, substance abuse disorder, Φ = .19, and ADHD, Φ = −.07, analyses. Parents did not report on the presence of adolescent suicidality and thus associated analyses were not conducted.

Table 6.

Association Between Adolescent Diagnoses and Reason for Community-Based Treatment (Parent Report Only) (N=47)

	Anxiety	DBD	Mood	Substance	ADHD
Agreement
Diagnosis (yes) / Focus (yes)	6.4%	14.9%	21.3%	2.1%	4.3%
Diagnosis (no) / Focus (no)	63.8%	40.4%	38.3%	85.1%	57.4%
Disagreement
Diagnosis (yes) / Focus (no)	19.1%	36.2%	19.1%	4.3%	21.3%
Diagnosis (no) / Focus (yes)	10.6%	8.5%	21.3%	8.5%	17.0%
χ2	0.73	0.91	1.33	1.74	0.21
P	.39	.34	.25	.19	.65
Phi	.12	.14	.17	.19	−.07
Effect Size	Small	Small	Small	Small	Small

DBD = Disruptive Behavior Disorders, ADHD =Attention-Deficit Hyperactivity Disorder

3.7. Sociodemographic Factors and Treatment Match

3.7.1. Insurance

There was no significant association found between insurance status (public vs. private) and “diagnosis-focus of treatment” match (per combined parent and adolescent report), χ² (1) = 1.93, p = .16, Φ = .15.

3.7.2. SES

There was a trend for adolescents with a “diagnosis-focus of treatment” match to have a higher family income than the group without this match (per combined parent and adolescent report) (r = .20, p= .09), though the association was not statistically significant.

4. Discussion

Numerous adolescents meet criteria for a mental health disorder in the United States and most do not receive evidence-based services (Kessler et al., 2009; Merikangas et al., 2010). This includes receipt of evidence-based diagnostic assessment, with parent and child, to guide treatment planning (De Los Reyes & Kazdin, 2005; Jensen-Doss & Hawley, 2010). In the absence of this assessment approach, adolescents may be less likely to receive treatment that is appropriately matched to their mental health needs (De Los Reyes & Kazdin, 2005; Achenbach, 2017). The present study was the first to examine the association between psychiatric disorders derived via evidence-based assessment as well as suicidality and corresponding focus of treatment in community-based mental healthcare per adolescent and parent report. This was examined in a clinical sample of treatment seeking adolescents. Overall, results suggest that adolescents with current mood disorders and DBDs were most likely to receive mental health care focused on their diagnoses per combined parent and adolescent report. The same pattern was found for suicidality (suicidal thoughts and/or attempts) per adolescent report. This was not true for anxiety disorders, ADHD, or substance abuse disorders. Notably, in exploratory analyses that examined the same questions using only adolescent and only parent report of adolescent diagnosis and reasons for treatment, adolescent report paralleled the combined parent-adolescent report findings, but parent report did not, suggesting that adolescents report drove study findings. We also found that the association between diagnosis and reasons for treatment was not affected by insurance status but was marginally (trend level) related to income.

As hypothesized, there was a significant association between the presence of a DBD and focus of treatment on these problems. This is not surprising given that there was also significant agreement between adolescent and parent report about the presence of a DBD, consistent with prior research conducted with clinical and community-based samples (De Los Reyes & Kazdin, 2005; Lewis et al., 2014), as well as agreement about whether DBD was a reason for treatment. Externalizing disorders are often the most commonly reported reason for treatment seeking, in part, because they tend to cause significant impairment at home (Foley et al., 2005; Grills and Ollendick, 2002; Salbach-Andrae et al., 2009).

In contrast, hypotheses were only partially supported with regard to internalizing disorders. There was a significant association between a diagnosis of mood disorder and focus on treatment on this problem area. These findings were supported by both parent-adolescent agreement about the presence of a mood disorder as well as whether mood disorder was a reason for treatment. However, there was not a significant association found between the presence of an anxiety disorder and associated reason for treatment, nor was there significant parent-child concordance about the presence of an anxiety disorder nor focus of treatment on this disorder. Prior literature suggests that internalizing disorders are often underreported by parents (Salbach-Andrae et al., 2009), but in the present study there was a clear differentiatio n between mood and anxiety disorders. Mood disorders, when accompanied by overt behavioral changes (e.g., drop in grades, withdrawal from activities, etc.), may be harder to keep hidden than anxiety disorders (Rothen et al., 2009). This may be particularly true when adolescents report accompanying suicidality.

Findings with regard to ADHD and substance abuse disorders were interesting. Contrary to hypotheses, the association between ADHD and focus of treatment on this disorder was not significant, nor was there a significant agreement between parent and adolescent report about the presence of ADHD. However, they did agree about whether it was a reason for treatment when it was reportedly addressed in treatment. Adolescents may find it challenging to evaluate their behaviors in relation to others, which is an important component of an ADHD assessment (Lahey et al., 2004). ADHD is also commonly comorbid with DBDs (Cummings et al., 2014; Lewinsohn et al., 1997; Steinberg & Drabick, 2015), and thus ADHD symptoms may be mistaken as acting out behavior.

Perhaps more interesting than ADHD findings are those associated with substance abuse. Contrary to prior research (Fisher et al., 2006), there was significant agreement between parent-adolescent ratings about the presence of substance abuse disorders. This higher rate of parental knowledge may stem from the context of the sample. Families were recruited through a study testing a substance abuse, suicide, and HIV prevention program. Thus, parents with knowledge of their adolescents’ substance abuse may have been seeking help through the study. More importantly, though adolescents and parents were aware of substance abuse problems, they were not reportedly a focus of treatment in community-based treatment when present. However, they did agree about whether it was a focus of treatment when reportedly addressed in mental health care. It is possible that adolescents with comorbid mental health problems were seeing therapists for their mental health concerns who may not have training in treatment of substance abuse. The mental health and substance abuse treatment systems are often fragmented, necessitating adolescents to see two therapists instead of one for treatment of both conditions (Hawkins et al., 2009). Given the burden associated with seeing two therapists (costs, time, etc.), this often does not occur.

These are a few notable issues about the study findings in general that deserve comment. First, exploratory analyses suggested that study findings pertaining to correspondence between adolescent diagnosis and focus of care were largely driven by adolescents. It is possible that this discrepancy in adolescent and parent report reflects that parents were not particularly involved in their adolescent’s care or notified about the focus of treatment. This may be particularly problematic for treatment of disorders that require significant family involvement for optimal treatment success (e.g., DBD, substance abuse) and safety (e.g., suicidality) (Asarnow et al., 2017; Dowell & Ogles, 2010). It is also possible that parents chose not to fully engage in their adolescent’s mental health care, which is not uncommon (Baker-Ericzen et al., 2013; Haine-Schlagel & Walsh, 2015). Also, important to note is that although there was an association between mood, DBDs, as well as suicidality, and focus of treatment on these conditions, there was still significant room for improvement. For example, only 8 out of 18 (44%) adolescents with suicidal ideation or a recent suicide attempt reportedly focused on suicidality in treatment. Overall, these results may suggest that clinical interviews administered in community-based mental health care may miss psychiatric diagnoses and associated symptoms, and that parents might be unaware of youth diagnoses or unengaged in the mental health treatment of their child.

With regard to the influence of socio-economic factors, private versus public insurance was not associated with the degree of match between diagnoses and corresponding focus of care. However, there was a trend for families with higher incomes to have greater correspondence between adolescent diagnosis and focus of care. Prior research does suggest that those with lower incomes may receive services of lower quality that may be inadequate or inappropriate to their needs (Wilk et al., 2005; Bisgaier & Rhodes, 2011).

The results of the present study hold important clinical implications. First, consistent with extant literature, results highlight the importance of assessing for the full spectrum of psychiatric disorders and high-risk behaviors, preferably with an evidence-based diagnostic interview, to capture all comorbid conditions and areas in need of treatment (Achenbach, 2017). Clinicians may also choose to use evidence-based self-report screening tools to indicate areas in need of diagnostic assessment. It will also be important to administer assessments to both adolescents and parents, preferably separately, for the most valid diagnostic assessment (De Los Reyes & Kazdin, 2005). Full understanding of all diagnoses present is needed to yield optimal treatment plans and appropriate selection of evidence-based interventions (Jensen-Doss & Weisz, 2008; Weisz & Addis, 2006).

Study results also emphasize the importance of offering education and training in evidence-based assessment to students enrolled in practitioner focused programs as well as practicing clinicians. Training should include attention to navigation of barriers to implementation of evidence-based assessment. This will be particularly important given new mandates by accrediting bodies of healthcare institutions, such as The Joint Commission, to engage in measurement-based care, or the use of evidence-based assessment to guide treatment planning and decisions (Black et al., 2018). Insurance companies may soon follow suit. In this vein, web-based, HIPAA compliant, measurement-based care software platforms have been developed that can be implemented in community-based settings. Community-based agencies may also consider partnering with educational institutions for consultation (Black et al., 2019).

Though the present study yields novel data with important clinical implications, it also has a number of limitations that deserve mention. First, the sample size was small, and thus some of the findings warrant caution in interpretation. Relatedly, the small sample size negated our ability to explore other sociodemographic variables (e.g., ethnicity, race, etc.) associated with disparities in quality of care. While small sample sizes are not uncommon in clinical psychology research, and are especially quite common in research conducted with underrepresented populations (Etz & Arroyo, 2015), such as the majority of the present sample, this is an important area for future research. This study represents an preliminary step in an important but neglected area of research. Second, while Kappa does correct for chance agreement, given our research questions, it may have over/under corrected agreement (e.g., kappa could be low if a disorder occurred infrequently, not necessarily because there is low agreement) due to how the responses were distributed in the data. Third, we did not have access to clinical records, and thus could not compare diagnoses derived via the study to those made independently by clinicians in the community. However, if clinicians used their patient diagnoses to derive treatment plans, study results may suggest that not all diagnoses were captured. Relatedly, we did not capture clinician report of reasons for treatment and thus cannot confirm correspondence with adolescent or parent report. It is possible that families and therapists did not agree on treatment goals (Hawley & Weisz, 2003) or that symptom presentation changed over the course of 6 months and families failed to recall the focus of earlier treatment, which could lead to discrepancies. Fourth, the DISC-4.0 tends to display relatively higher reliability estimates for externalizing disorders diagnoses relative to internalizing disorder diagnoses, which should be taken into account when interpreting study results (Silverman & Ollendick, 2005, Klein et al., 2005). Notably, in the present study, there was significant parent-adolescent agreement on mood disorders and DBDs. Fifth, selection bias may be present as families were recruited in the context of a family-based prevention program. Parents in this study may be more involved with their adolescents and thus more knowledgeable about their mental health conditions. This bias may have led to a more optimistic picture than is typical. Last, just under half of the sample was referred from a youth shelter, and thus results may not generalize to youth recruited from other clinical settings. Future research in this area should employ larger samples from naturalistic community-based settings.

5. Conclusions

Overall, the present study suggests a potential need for closer examination of diagnostic assessment and treatment planning methods used in community-based mental health care. In the present study, adolescents met criteria for several diagnoses that were not reported to be a focus of treatment per adolescent and/or parent report. Notably, evidence-based assessments are not often used in routine care (Ng & Weisz, 2016) and thus efforts are needed improve dissemination of EBPs in this area to improve quality of care (Harvey & Gumport, 2016).

Highlights.

• Diagnostic interviews yield more accurate diagnoses than clinical interviews

• Youth who are not properly diagnosed may not receive needed mental healthcare

• Focus of care was consistent with mood and disruptive behavior disorder diagnoses

• The same pattern was found for focus of care and youth suicidality

• This was not true for anxiety disorders, ADHD, or substance abuse disorders