Caregiver–Child Discordance on the DSM-5 Cross-Cutting Symptom Measure Among Youth in Outpatient Psychiatry

Abstract

Psychiatric illness during childhood and adolescence is a growing concern, placing increased pressure on psychiatric services. Reliance on an evidence-based assessment tool may facilitate the identification and management of symptoms and may facilitate accountability. The purposes for this study were to characterize the psychiatric symptoms of a sample of pediatric patients accessing outpatient psychiatry using the DSM Level 1 Cross-Cutting Measure (CCSM), compare patient and caregiver ratings (CCSM), and explore patients’ acceptability of using the CCSM. The sample consisted of 51 patients (mean age 14 years) and 46 caregivers (mean age 43 years). Patient and caregiver ratings suggested problems with sleep, inattention, depression, irritability/anger, and anxiety. The most common discordance concerned suicide ideation. Most patients (34 of 38) reported that the assessment tool was easy to complete. The CCSM may be a useful, evidence-based, standardized, transdiagnostic assessment tool aligned with the DSM-5 that can be used in a variety of mental health settings to identify symptoms, inform treatment planning, and track progress.

1. Introduction

Mental illness in children and adolescents (herein, youth) is a growing medical, public health, and educational concern. The peak age of onset for mental disorders has been estimated as 14.5 years [1] For many youth, symptoms of mental illness are associated with poor physical health [2,3] and have been shown to interfere with educational performance [4,5], interpersonal relationships [6], and occupational functioning [7]. Therefore, identifying and addressing mental illness during childhood and adolescence is important for their physical and psychological health, educational outcomes, and future well-being.

Estimates for the prevalence of psychiatric illness in youth range from 15% [8] to 22% [9]; however, a substantial proportion of these youth (75–80%) do not access specialized treatments for these disorders [10,11]. However, in recent years, there has been a documented rise in the need for specialized outpatient psychiatric services [12,13] for a variety of disorders such as anxiety and personality disorder, and self-harm and suicide in females, and for eating disorders in both females and males [14]. This growing need for specialized care has placed increased demand and strain on an already strained system.

Once accessing mental health care, including outpatient psychiatry, youth are assessed and diagnosed or provisionally diagnosed based on guidance from the Diagnostic and Statistical Manual (DSM-5-TR) of the American Psychiatric Association or the International Classification of Diseases [15,16]. Comorbidity is common and has long-lasting impacts [17,18,19]; therefore, the use of a standardized measure to assess symptoms that cut across multiple mental illnesses, such as the DSM Level 1 Cross-Cutting Symptom Measures (CCSM), may prove useful. The CCSM was developed by the American Psychiatric Association for clinical evaluation and research. The suite of measures includes caregiver-reported and child-reported Level 1 cross-cutting symptom severity ratings, among other tools. In addition, the use of both the caregiver-rated and child-rated forms of the CCSM may provide insights into differing perspectives or congruence on the same symptom domains.

A mental health assessment may be an essential component of care to guide clinical decisions and ultimately improve socio-emotional, behavioural, physical, and educational outcomes. Assessment tools may be useful for identifying areas of concern, tracking progress, and demonstrating accountability. The CCSM is freely available [https://www.psychiatry.org/psychiatrists/practice/dsm/educational-resources/assessment-measures Accessed 30 April 2025] and was developed to provide clear and clinically meaningful ratings to allow for the characterization of patients in terms of presence and severity, offering insights into degrees of dysfunction and not necessarily a diagnosis. The tool was designed to assist clinicians in identifying symptoms that are important across diagnoses and identifying areas for in-depth assessment.

A mental health assessment is a critical component of prevention and intervention services in a wide range of contexts, offering supports and services for psychiatric illness, including youth hubs, schools, and community and medical services. Evidence-based assessments [20] with valid and reliable tools have been shown to be feasible in routine mental health services for youth [21,22]. One of the challenges with assessment concerns the cross-cutting symptoms [23]. However, Corbel and Kinchen (2024) [22] have especially reported the usefulness of using the DSM Level 1 CCSM in the community system, which was expressly designed to assess cross-cutting symptoms.

In order to use parent and child reported symptom severity ratings, it is important to notice and understand differences in ratings between these two sources. Some investigators have reported low to moderate agreement between youth and caregiver reports (e.g., [24]); however, child–caregiver agreement has been reported as strong for observable symptoms and non-school-based symptoms [25] and good to excellent for a range of disorders including conduct disorder, core symptoms of anorexia and bulimia, ADHD and ODD but not anxiety or depression [26]. Studies on youth and caregiver agreement include samples from the general and clinical populations, though only three studies could be located with pediatric patients accessing outpatient psychiatric clinics. In two of these studies in the outpatient clinic [27,28], the Child Behaviour Checklist and Youth Self Report were used, and in the third study [29], the Mini International Neuropsychiatric Interview for Children and Adolescents (MINI-KID) was used. Low to moderate agreements were reported by Mbekou and Duncan, while Jónsdóttir reported fair to excellent agreement. No reports on child and parent discordance with the CCSM in a specialized, child and adolescent outpatient clinic could be located, which makes this study one of the first studies in which the characterization of youth and quantification of child–caregiver discordance with the CCSM are reported for this population and setting. Integrating child and caregiver perspectives on the child’s symptoms may foster optimal care [30] and provide opportunities for improved understanding of child symptoms.

The CCSM was expressly designed for clinical and research purposes based on the current DSM-5TR. The implementation of a transdiagnostic symptom rating scale directly linked to diagnostic criteria may facilitate screening and assessment in clinical and community settings. The aims for the present study were to characterize patients’ symptoms based on the CCSM, explore the discordance between caregiver and child-reported versions of the same domains, and explore the patients’ acceptability of the CCSM. The term, caregiver, is expressly used since some patients may be supported by non-parental guardians. To address these aims, a survey was conducted in a child and adolescent outpatient psychiatry unit; this report is focused on symptom ratings captured shortly after referral and is part of a larger study on youth accessing outpatient psychiatry. Due to the exploratory nature of the study, no specific hypotheses were created.

2. Materials and Methods

2.1. Setting

The survey was conducted in a regional hospital’s child and adolescent outpatient psychiatry unit. This regional health network serves a population of about 800,000, including local cities and rural areas. Patients are referred by professionals in other hospital programs or family physicians, or urgent care clinics. A multi-disciplinary mental health team, including psychiatry, psychology, nursing, social work, and occupational therapy, provides assessments, treatments, and recommendations to family physicians. Referrals to this child and adolescent outpatient psychiatry clinic for suicidal concerns come from two sources: the Emergency department or directly from community practitioners. For youth referred from the Emergency department, typically a Social Worker or Occupational Therapist will first assess the youth to understand the reason for the suicidality, then make referrals to appropriate supports to assist with remedying it (e.g., Psychiatry, counselling, community mental health nurse, school social workers, other community partners). For referrals from community practitioners, a Social Worker or Occupational Therapist will first review the referral to determine the level of acuity of the youth (Urgent, Semi-Urgent, Non-Urgent). Afterwards, that practitioner will complete a telephone triage assessment with the youth and their parent before connecting them with a psychiatrist. For both Emergency department or community practitioner referrals, it is sometimes the case that the acuity of the youth is so severe that during the assessment, they may be directed to immediately visit the Emergency department or in some cases admission to the child and adolescent inpatient unit is arranged by the psychiatrist if the assessing clinician believes that the youth would be at too much of a risk to wait for outpatient support. Research ethics approval was obtained from the Waterloo-Wellington Research Ethics Board (WWREB #: 2022-0750) and the University of Guelph (REB# 22-08-033) in accordance with the Declaration of Helsinki.

2.2. Participants and Procedures

All patients and their caregivers referred to the outpatient clinic with the ability to communicate in English were invited to participate in the study. The term, caregiver, is expressly used because some pediatric patients may not be residing with or supported by their biological parent while at the hospital. To maintain privacy and confidentiality, caregivers were informed by hospital staff about this study opportunity, who asked caregivers if they would like to learn about the study from a research assistant (RA). For caregivers who agreed, the RA contacted the caregiver by telephone, engaged in the consent process, and administered the survey to caregivers who provided informed consent. Caregivers were also asked for consent for their child to complete the survey on their first appointment. At the patients’ first appointment, patients were fully informed of the study opportunity, and fully informed consent was requested. Those who agreed completed the survey questionnaire at the outpatient unit.

2.3. Measures

Survey questionnaires were developed for the caregiver and the patient that included demographic items and standardized measures. To capture child symptoms, caregivers completed the DSM-5-TR Parent/Guardian-Rated Level 1 Cross-Cutting Symptom Measure (CCSM) for children aged six to 17 years [31]. This 25-item measure was designed by the American Psychiatric Association for the assessment of patients’ symptoms across common domains of mental health problems over the past 2 weeks. Nineteen items were designed to assess 10 psychiatric domains, including somatic problems, sleep disturbance, inattention, depression, irritability and anger, mania, anxiety, psychosis, and repetitive thoughts and behaviours, rated on a 5-point scale from 0 (not at all) to 4 (Severe, nearly every day). A score of 2 (mild or experienced most days) or greater would indicate that the clinician should further investigate that domain, potentially using one of the Level 2 measures designed for each domain (though no Level 2 measure exists for psychosis or suicide). Six items of the DSM CCSM were designed to assess two domains, suicide attempt/ideation and substance use, and are rated on a “Yes, No, or Don’t Know” scale. This tool has been shown to be valid and reliable [31,32,33].

Patients completed the DSM-5-TR Self-Rated Level 1 Cross-Cutting Symptom Measure (DSM-SR-CCSM). It consists of the same items, domains, and response options as the Caregiver/Guardian version, though the wording of the items has been adapted for the child patient. After completing the CCSM, patients were also asked their opinions on using the measure with open-ended questions on ease of use.

2.4. Data Analysis

Demographic information is presented with descriptive statistics including means, standard deviations, and proportions. Comparisons of youth and caregiver reporting of symptom severity were performed with paired t tests for continuous ratings (i.e., items 1–19 on the CCSM reflecting 10 domains). The significance level was set at p < 0.05. Cohen’s Kappa coefficient was used for categorical ratings [34] for the CCSM items 20 to 25 on substance use and suicide (Yes, No responses). Cohen’s Kappa [35] can range from −1 to 1, where 1 represents perfect agreement, 0 represents the amount of agreement to be expected due to chance, and kappa values below 0 are rare. Kappa values can be interpreted as poor agreement (0 to 0.40), fair (0.41–0.59), good (0.60–0.74), and excellent (>0.75). Discordance between patient and caregiver reporting on substance use and suicidal thoughts and attempts was presented with number and percent disagreement with A (patient yes, Caregiver no), B (patient no, caregiver yes), C (patient no, caregiver left the item blank), D (patient yes, caregiver left the item blank). The sample size was too small to permit subgroup analyses. Statistical analyses were performed using the Statistical Package for the Social Sciences (SPSS) 26.0 (IBM Corp., Armonk, NY, USA) [36]. Patients’ highly brief responses to the question on the acceptability of completing the CCSM were recorded verbatim and categorized with content analysis [37].

3. Results

The sample comprised 51 pediatric patients, of whom there were 46 dyads of patients and their caregivers who completed surveys between February 2023 and December 2024. The mean age of the patients was 14.41 years, most (65%) identified as female, and most (73%) identified as white (see

Table 1. Patient characteristics, n = 51.

Characteristic	Patient
Age, m (SD)	14.41 (2.3)
Gender, n (%)
Female	33 (65)
Male	14 (28)
Non-binary	4 (8)
Ethnicity, n (%)
White/European	37 (73)
Indigenous	4 (8)
Black/Caribbean	2 (4)
Southeast Asian	2 (4)
Middle eastern	2 (4)
Latin American	2 (4)
Not sure	2 (4)
Prior outpatient visit, n (%)
Yes	24 (47%)
Number of prior visits, n (%)
One	12 (50)
Two to four	7 (29)
Ten or more	5 (21)
Reason for referral, n
Suicide ideation/self-harm	27
Depression	21
Behaviour problems	11
Anxiety	9
Gender dysphoria	5
Psychotic symptoms	4
Eating disorder	2
Substance use disorder	1

for patient characteristics). The reasons for referral were available for 46 youth; the main reasons for referral were suicide ideation/self-harm/safety (n = 27/46; 58.7%) and depression (n = 21/46; 45.7%). Most patients had more than one reason for referral.

The mean age of caregivers was 43 years (SD 7.4), and 41 (89%) caregivers indicated being female. Most (36 of 41 who answered this item, 87.8%) reported being the mother of the patient, 3 indicated being the father, 1 reported being the grandmother, and 1 reported being the sister of the patient. Most caregivers (33/45; 73.3%) indicated being in a spousal relationship. Of the 36 caregivers who provided this information, the mean household income was $106,694CAN (SD 7 7913.44).

Statistically significant differences between patient and caregiver ratings were evident in the domains of depression, mania, anxiety, psychosis and repetitive thoughts and behaviours, and the overall total score (

Table 2. Comparison of symptoms ratings, m (SD).

Characteristic	Patient	Caregiver	Mean Difference	Paired T	p Value	d Index
Somatic	1.64 (1.93)	1.26 (1.22)	0.38	−1.9 (45)	0.060	0.24
Sleep	2.29 (1.46)	2.15 (1.40)	0.14	−0.5 (45)	0.628
Inattention	2.76 (1.29)	2.39 (1.36)	0.37	−1.3 (45)	0.207
Depression	2.43 (1.29)	1.83 (1.09)	0.60	−2.5 (45)	0.017	0.45
Irritability/Anger	2.47 (1.18)	2.21 (1.38)	0.26	−0.9 (45)	0.378
Mania	1.63 (0.90)	0.80 (0.96)	0.83	−4.5 (45)	<0.001	0.89
Anxiety	2.18 (1.14)	1.73 (1.36)	0.45	−2.1 (45)	0.044	0.36
Psychosis	0.74 (1.07)	0.42 (0.73)	0.32	−2.5 (45)	0.018	0.35
Repetitive thoughts and behaviours	1.58 (1.13)	0.64 (0.80)	0.94	−5.4 (45)	<0.001	0.96
DSM Total	35.72 (14.16)	25.32 (13.83)	10.4	−3.76 (45)	<0.001	0.74

). For each item, the patients rated themselves as more severe than the caregiver rated the patient as represented by the negative t-values. The greatest difference between youth and caregiver ratings was for repetitive thoughts and behaviours; youth rated this symptom 0.94 points higher than caregivers (1.58(1.13) vs. 0.64(0.80), t = −5.4(45), <0.001). Based on sample means of 2 or greater, youth scores indicated that further clinical investigation would be warranted for sleep (n = 43 youth rated themselves as 2 or higher, 84%), inattention (n = 41, 80%), depression (n = 38, 75%), irritability/anger (n = 36, 71%) and anxiety (n = 33, 65%), while caregiver ratings suggested sleep (n = 31 caregivers rated the youth as 2 or higher, 67%), inattention (n = 34, 74%) and irritability/anger (n = 28, 61%) required further clinical investigation. Cohen’s d index, representing the magnitude of difference between patient and caregiver ratings for statistically significant differences (or nearly statistically significant differences), can be interpreted as small (0.20, medium (0.50 or large (0.80) differences [38].

Kappa coefficients suggested good agreement on two of the six items on substance use and suicide, namely the use of drugs like marijuana, cocaine, or crack, or club drugs, and whether the patient ever attempted suicide (

Table 3. Kappa coefficients categorial items.

CCSM Item	Patient, n, yes	Caregiver, n, yes	Kappa	Agreement
1. Had an alcoholic beverage	7	4	0.377	poor
2. Smoked cigarette, cigar, pipe, or chewing tobacco	5	4	−0.094	poor
3. Used drugs like marijuana, cocaine or crack, club drugs	8	6	0.613	good
4. Used any medicine without a doctor’s prescription	2	2	−0.050	poor
5. talked about wanting to commit suicide	21	17	0.149	poor
6. ever tried to kill your/himself/herself?	19	14	0.596	good

), and poor agreement on all the other categorical items. The number and types of discordance appear in

Table 4. Discordance between patient and caregiver, n = 46 dyads.

CCSM Item, n	A	B	C	D	Total	Percent
1. Had an alcoholic beverage	3	2	4	1	10	22
2. Smoked cigarette, cigar, pipe, or chewing tobacco	3	3	4	1	11	24
3. Used drugs like marijuana, cocaine or crack, club drugs	2	3	3	0	8	17
4. Used any medicine without a doctor’s prescription	2	1	4	0	7	15
5. talked about wanting to commit suicide	9	9	0	1	19	41
6. ever tried to kill your/himself/herself?	5	3	0	1	9	20
Total	24	21	15	4

A patient yes, caregiver no; B patient no, caregiver yes; C patient no, caregiver blank; D patient yes, caregiver blank.

. The greatest number of disagreements was for the item on suicide ideation (19 disagreements, 41%). For this item, the common type of disagreement was a patient response of yes and a caregiver response of no (n = 24); however, it was also common (n = 21) for patients to report no and caregivers to report yes for this item on suicide ideation. Some caregivers left items on substance use and suicide blank, though they completed the symptom rating part of the CCSM. One caregiver left all six of these items blank; two caregivers left the items on substance use blank, but reported on the suicide items, and one caregiver left some items blank but responded to one substance use item and one suicide item. Caregiver omissions were predominantly on the substance use items and in cases where the patient reported not consuming the substances.

Of the 38 patients who provided comments about the CCSM, most (n = 34) considered the tool easy to complete. Five patients noted that having a range of response options (0–4) was helpful. However, four patients noted that it was difficult to recall precisely if the symptoms occurred within the last two weeks (e.g., did the symptoms occur within the last 14 days, or was it 16 or 17 days?). Three patients clarified that providing a longer time frame than two weeks would make the tool easier and more accurate for them to complete. One patient noted that the symptoms depended on the situation, and one patient noted concern about what people would think when they viewed the responses.

4. Discussion

The findings of this study revealed important characteristics of youth referred to specialized care and discordance between youth and caregiver ratings of the youths’ symptoms and behaviours. In this characterization of youth accessing outpatient psychiatry, the main reasons for referral were suicide ideation or attempt or self-harm and depression; for two of these reasons, there was significant discordance between patient and caregiver ratings. In this analysis of child ratings and caregiver ratings of child psychiatric symptoms across multiple domains, child sample means of 2 or greater suggested difficulties with sleep, inattention, depression, irritability/anger, and anxiety that should be further explored in a clinical assessment. Caregivers noted the same domains except depression and anxiety. This finding is consistent with previous research on agreement between child and caregiver ratings of child psychiatric symptoms; caregiver ratings for non-observable (e.g., depression and anxiety) have been reported as lower than their child’s ratings [25]. Significant differences in ratings of substance use and suicide ideation and attempt were also noted. The greatest number of discrepancies between the patient and the caregiver was on the item regarding suicide ideation.

Clinicians may have the opportunity to include both caregiver and youth standardized assessment tools; however, there may be instances where the caregiver may not be available, or the youth prefers the caregivers’ involvement over their own. Therefore, having the opportunity to use either a youth- or caregiver-reported tool or both for an assessment is advantageous. Understanding how youth and caregivers tend to rate the severity or tend to disagree assists the clinician when interpreting the scores. Clinicians should expect discordance, not only because disagreement has been documented with clinical assessment tools, but also because clinicians would understand family dynamics and adolescent development (e.g., seeking autonomy from parents), and exploring discordance in clinical practice may foster better understanding of and between the youth and caregiver. The discordance may suggest treatment goals that include assisting the youth to develop communication skills and may include supports for the caregiver to learn skills in listening, validating, and supporting their child.

The kappa coefficient is widely used in health and psychology, though in this study, the results suggest it was less informative than the examination of discordance. Other investigators (e.g., [39]) have reported similar shortcomings with interpreting kappa values. Some discrepancies may be due to the child’s symptoms varying by context (e.g., home versus school; [30]), and caregivers may vary in their attunement or quality of connection with the youth and may benefit from professional guidance in repairing disagreements or ruptures within the parent–child relationship [40].

While youth tend to disclose suicidal thoughts and behaviours to friends [41], caregiver input is desired in clinical consultations. A high rate of discordance between child and caregiver reports of child suicide ideation or attempt has been reported [42,43], with many caregivers being unaware of their child’s suicide ideation. It is also possible that some caregivers overestimate (false positives) or misunderstand their youth behaviours. However, adolescents have also concealed or denied thoughts of suicide or death that their caregivers reported [44]. Notably, many youth often do not disclose suicidal thoughts and behaviours to their clinicians, including mental health professionals [42], for fear that the clinician will inform their caregiver, and the patient does not want to worry the caregiver or suffer their reaction. Discordance between youth and caregiver ratings could also stem from each using a different reference period; recall that the reference period for the symptoms was the prior two weeks, and some youth expressed difficulty with reporting within this time frame. Taken together, these findings suggest that implementing both patient- and caregiver-reported measures may enhance understanding. Caregivers may have an important role in the mental health of the children in their care and may be involved in treatment plans or involved with helping their child enact treatment plans. Involving caregivers in assessment and intervention may have beneficial impacts on child outcomes.

The biopsychosocial environment in which children and youth develop may impact their outcomes and self-rated scores. Additionally, caregiver psychopathology may also influence the developing child and the caregiver ratings of the youth in their care. For example, a pattern of intergenerational continuity of affectivity within families, noting environmental contingencies in addition to neurocognitive and temperamental susceptibility, may lead to differential impacts on a child’s emotional functioning has been reported [44]. Other investigations of parental psychology (e.g., Refs. [45,46] indicate the significant impacts caregiver psychopathology may have on the developing child, which suggests it should be a normal part of clinical exploration in practice.

Limitations

The study should be interpreted with a view to its limitations. This study was based on perceptions, more specifically, perceptions of youth experiencing psychiatric symptoms severe enough to be referred to specialized care, and the perceptions of their caregivers. The psychopathology of the caregiver was not captured for this study, limiting an exploration of its role in the findings. The results of this study may have been affected by perception bias, and almost half the sample reported prior outpatient visits, which may have affected their ratings; however, this bias would also exist in clinical practice. In this sense, the study may reflect real-world experiences of youth, caregivers, and clinicians. Other limitations include the reliance on a single outpatient clinic, a modest sample size of 51 youth but only 46 patient-caregiver dyads, a predominantly female and White sample, and an unknown participation rate. The number of families who declined participation was not recorded, which could introduce selection bias. For this study, group means and frequencies were reported; however, individual and nuanced symptoms or problems are not presented, though they would receive attention in clinical practice. In addition, we did not adjust for multiple comparisons due to the exploratory nature of the study; results should be interpreted with caution, though observed patterns align with prior research. The ratings could also have been affected by response, selection, and social desirability biases. In particular, the risk of stigma may have affected participants’ responses. Further research is needed across clinical settings to confirm the utility of the CCSM, including community clinics, and to explore the use of the CCSM for longitudinal analyses to monitor symptom change and to assess intervention effects.

5. Conclusions

Youth referred to specialized care reported difficulties with sleep, anxiety, depression, inattention, and a considerable proportion reported suicide ideation and attempts. Based on previous research, there were some expected differences between youth and caregiver ratings of youth psychiatric symptoms. The parallel youth and caregiver CCSM forms may provide advantages for use in clinical practice and research since they are based on the current version of the DSM (i.e., DSM-5-TR) and allow for a range of response options consistent with viewing psychiatric disorders in degrees rather than presence or absence of disorder. Apart from the instructions for completing the CCSM, limiting the experience to the previous two weeks, youth reported acceptability of completing the measure, suggesting the CCSM could be useful for identifying, informing treatment, and tracking progress in a variety of settings. Future research should be conducted to formally assess the clinical utility of the CCSM.