Comprehensive Treatment and Monthly 2 Year Follow-Up in Children With ADHD: Randomized Controlled Prospective Study

Abstract

Objectives

To evaluate the efficacy of regular monthly follow up in maintaining treatment gains versus follow up as needed in the community for children ages 6 to 12 years with ADHD.

Methods

One hundred and eighty six children, age 6 to 13 with DSM-IV diagnosed Attention Deficit Hyperactivity Disorder (ADHD) (combined presentation) received 3 months of comprehensive treatment which included long-acting stimulant medication titrated to optimal dose (via parent and teacher reports) and weekly parent training groups and child social skills and academic training. After the 3 months of treatment, children were randomized to either monthly follow up by the research team (treatment group) or follow up as usual in the community (control group). Comprehensive evaluations occurred at 3 months at the end of the comprehensive treatment, and at 6, 12, 18, and 24 month follow-up.

Results

Children in both groups improved significantly in all domains after the 3 months of intensive comprehensive treatment. Treatment gains were maintained during the 2 year follow up period in both groups. However, the Hyperactive/Impulsive rating in Conners scale by parent report and some teacher reports were worse in the community group during the follow-up period.

Conclusion

The importance of the initial intensive comprehensive 3 month treatment cannot be underestimated as these treatments resulted in significant improvements which were maintained for 2 years in both groups. Factors such as social economic status and available community treatment which may have influenced these results are discussed.

Introduction

Generally, medication combined with multimodal psychosocial treatments are recommended by most practice guidelines for the treatment of children and adolescents with Attention Deficit Hyperactivity Disorder (ADHD; CADDRA–Canadian ADHD Resource Alliance, 2020; National Institute for Health and Care Excellence, 2018; Wolraich et al., 2019). The guidelines have relied on findings from controlled prospective multimodal treatment studies for example, Multimodal treatment of Children with ADHD (MTA) for these recommendations (The MTA Cooperative Group, 1999, 2004a, 2004b).

In the MTA study 576 children ages 7 to 9 years old with DSM-IV diagnosed ADHD (combined type) were randomly assigned to 14 months of treatment to one of four treatment groups; a medication group, a psychosocial treatment group, a combined medication and psychosocial treatment group, and a community treatment as usual group. The combined group proved superior to the behavioral and community group (66% of whom received medication) in oppositional and aggressive symptoms, internalizing symptoms, teacher rated social skills, parent-child relations, and reading achievements while the medication group did not. However, at 24 months (10 months after treatment ended), the combined group had lost 50% of the treatment gains (The MTA Cooperative Group, 2004a), and at 36 months or 3 year follow up no treatment gains could be detected as all groups were equal (The MTA Cooperative Group, 2004b).

This finding was in contrast to a previous controlled 2-site multimodal treatment study where 103 children age 7 to 9.9 years old with DSM-IV diagnosed ADHD (combined type) were all responsive to methylphenidate and randomized to medication alone, medication and multimodal psychosocial treatment (MPT), or medication and equal attention but not the specifics of the multimodal treatment (attention control treatment [ACT]). Treatment was after school twice per week for 1 year. Monthly follow up in each intervention was received in the second year. Significant gains were seen with medication, further slight gains were reported over time, but generally there were no large group differences over the 2 years (Abikoff, Hechtman, Klein, Gallagher, et al., 2004; Abikoff, Hechtman, Klein, Weiss, et al., 2004; Hechtman, Abikoff, Klein, Greenfield, et al., 2004; Hechtman, Abikoff, Klein, Weiss, et al., 2004; Klein et al., 2004). However, in this second study there were no decreases in initial treatment gains in the second year when monthly follow up visits in each intervention occurred. Medication treatment was also maintained during this follow up year.

The contrast in maintenance of treatment gains in the two studies was striking and led us to this current study where 186 children age 6 to 13 years old were provided with 3 months of comprehensive medication and psychosocial treatment then randomized to monthly follow up by study staff or follow up in the community. We hypothesized that 3 months of medication and intensive psychosocial treatment would significantly benefit the children with ADHD and their families and the monthly follow up, and brief interventions by the clinic staff (psychologists) when needed would help better maintain treatment gains compared to follow up in the community.

The primary outcome measures were the Conners parent and teacher 10 item questionnaire which measures inattention and hyperactivity/impulsivity. We hypothesized that the group which received monthly following by study staff is more likely to retain treatment gains compared to the group followed in the community.

The study design and a general outline regarding analysis was reviewed and approved by the Institutional Review Board. The study was preregistered (NCT 02142140).

Methods

Patient Recruitment

Children 6 to 13 years of age with possible ADHD were recruited via information describing the study sent to school guidance counselors and school psychologists, pediatricians, free community psychosocial intervention clinics (CLSC) and parent groups such as People Apt at Negotiating Deficits of Attention (PANDA) and Centre for ADHD Awareness Canada (CADDAC).

All referred children who passed the initial telephone screening process received comprehensive evaluations by a qualified child psychiatrist to determine ADHD diagnosis, comorbidities, and severity of these comorbidities. A qualified psychologists or one of their students under supervision also provided comprehensive psychological assessments to determine Intelligence Quotient’s (I.Q.’s) and possible learning disorders.

Exclusion criteria included Autistic Spectrum Disorder (ASD), intellectual disability, overall IQ less than 80, and history of significant brain injury. Additionally, any comorbid disorder severe enough to require treatment (e.g., severe anxiety, depression, Obsessive Compulsive Disorder [OCD], or psychosis) was an exclusion criterion. Patient characteristics in the treatment group and control group are outlined in Table 1. There were no significant differences in gender, SES, ethnicity, I.Q., ADHD severity, or level of comorbidity in the two groups. Academic achievement on the Wechsler Individual Achievement Test (WIAT) was not significantly different between the groups.

Table 1.

Patient Characteristics.

Patient Characteristics	Parent Ratings of Child’s ADHD	Case management group	Community group	Statistic used	p-Value
N		90	96
Age		8.71	8.85	Student’s T-test	.565
Gender				x² test of independence	.828
	Male	66	72
	Female	24	24
SES (median income levels 1–6)		5	5	Mann-Whitney U-test	.258
IQ mean		95.8	97.6	Student’s T-test	.327
Ethnicity				x² test	.446
	Caucasian	75	82
	African	3	4
	East/Southeast Asian	0	2
	Hispanic	1	0
	Other	11	8
ADHD severity (Baseline)
	Parent Global mean	76.8	77.2	Student’s T-test	.766
	Parent hyperactivity + impulsivity mean	76.9	78.8	Student’s T-test	.384
	Parent inattention	83.5	83.2	Welch’s T-test	.810
Generalized anxiety		7	3	x² test of independence	.165
Oppositional defiant disorder		50	54	x² test of independence	.860
Wechsler Individual Achievement Test (WIAT)		89.3–90.5 (For different measures)	91.9–94.5 (For different measures)		.149–.209

Treatment Interventions

All children with ADHD received carefully titrated stimulant medication (usually long-acting methylphenidate e.g., MLR methylphenidate [Biphentin^®] or OROS methylphenidate [Concerta^®]), titrated to optimal dose (using the Conners parent and teacher rating scales) and minimal side effects (using a side effects checklist completed by the child and parents). Choice of medication for example, Biphentin or Concerta was determined by the length of time the child needed coverage and whether or not they could swallow pills. Children needing shorter periods of coverage and those who could not swallow pills were given Biphentin. Those requiring longer periods of coverage and able to swallow pills were given Concerta.

In addition to stimulant medication, all parents received parent training and children received social skills training. Children who scored low on academic functioning scales also received academic skills training.

Parent Training

The parent training program occurred 2 hr/week for a 3 month period with a small group of parents on a Saturday morning (while their children received social skills intervention). Both were given in French and English. The program focused heavily on collaborative and proactive solution method developed by Ross Greene (Greene & Winkler, 2019). Parents were given handouts of the slides presented. The themes covered included education about ADHD, active listening, building a relationship, implementing structure and routine, introduction to collaborative and proactive solutions (Greene & Winkler, 2019) (Detailed description in Appendix 3).

Social Skills Program

The social skills program occurred weekly for 2 hr for 3 months in small groups on a Saturday morning. The intervention followed a manual developed by Dr. Natalie Grizenko who headed the Douglas Hospital site and Dr. William Harvey who ran the program. Dr. Grizenko’s social skills training program emphasized training on use of self-other perspectives (Grizenko et al., 2000) and Dr. Harvey’s program focused on self-control in physical activity (Harvey et al., 2003). These two approaches were integrated into the social skills training.

Academic Training

Academic training occurred after school once per week for 2 hr for 3 months. Training occurred for the core subjects (e.g., English, French, Math) the child had difficulties with (e.g., consistently failing grades below 60%) following consultation with the child’s teacher to clarify the problem areas. Parents attended the academic training for their child by observing the interventions behind a one-way mirror. This ensured that the information given to children was used at home by parents. Brief daily reviews of this material was encouraged at home between academic training sessions. Small rewards were provided for effort, attentiveness, and mastery of material. The language of academic instruction during the academic training was the same as the language of instruction for that subject at school.

These three interventions occurred weekly in a small group format for 3 months. The parent training and the social skills training occurred on a Saturday morning and the academic skills training occurred after school during the week.

The recruitment and service provisions were performed in two sites: the child division of the Douglas Mental Health Institute and the Montreal Children’s Hospital.

After 3 months of intensive treatment, children were randomly assigned to monthly follow-up and brief interventions by study staff versus follow-up as usual in the community. This randomization was conducted equally and similarly at each site.

Follow-Up Interventions

The follow-up provided by the study staff (child psychologist and trainees in graduate psychology programs under supervision, and trained research assistants) consisted of monthly calls to families and an exploration of how their child was doing at school and at home, academically, socially, and emotionally. If problems were reported, follow up services were tailored to the perceived difficulty of the child. Options of services included calls to the teacher and/or meetings with the child and family and/or referral for medication adjustment with the child psychiatrist depending on what the underlying difficulty was judged to be. Interventions were usually brief and pragmatic.

Community Follow-Up Group

The children with ADHD who were randomly assigned to follow-up in the community could obtain whatever follow-up treatment the parents felt they needed or could procure in the community.

As seen in the table of services received by the community group (Table 2), most (96.23%) received long acting stimulant medication. In addition, most (63.73%) received an Individualized Education Plan (IEP) at school. Tutoring was also relatively high with 36.82% receiving tutoring. Only 14.81% received individual therapy, while only 6.17% received family therapy. Parent training and support group was also relatively low at 4.11%. Generally, the community group was fairly affluent with 48% of families earning more than $100,000 per year per household. Follow up evaluations for the community group occurred at the regular follow up intervals, for example, after 3 months of intensive treatment and at 6, 12, 18, and 24-month follow up.

Table 2.

Percentage of Participants Who Received Different Services in Each Group Throughout the Follow-Up Period.

Treatment received	Case management group	Community group	T-test results
Medication
Stimulant	91.95% (2.17%)	96.23% (0.66%)	t (6) = −3.76 (p = .009)
Non-stimulant	20.98% (6.74%)	13.40% (2.58%)	t (6) = 2.10 (p = .081)
Individual services	95.75% (3.23%)	19.28% (3.61%)	t (6) = 31.59 (p < .001)
Individual therapy	20.13% (5.18%)	14.81% (3.72%)	t (6) = 1.67 (p = .146)
Other	93.75% (2.18%)	4.45% (1.21%)	t (6) = 71.71 (p < .001)
Family therapy	11.05% (2.16%)	6.17% (1.70%)	t (6) = 3.54 (p = .012)
Academic services	85.68% (2.43%)	80.08% (3.78%)	t (6) = 2.49 (p = .047)
Individualized Education Plan (IEP)	63.35% (1.96%)	63.73% (3.11%)	t (6) = −0.21 (p = .844)
Tutoring	47.95% (2.09%)	36.82% (6.15%)	t(3.686) = −3.43 (p = .030)
Testing	7.83% (3.88%)	7.23% (3.72%)	t (6) = 0.22 (p = .83)
Parent support	9.23% (1.62%)	13.48% (4.80%)	t(3.671) = −1.68 (p = .17)
Parent training & support group	2.20% (1.19%)	4.11% (1.54%)	t (6) = −1.97 (p = .097)
Counseling	6.63% (2.22%)	6.23% (2.48%)	t (6) = 0.24 (p = .82)

Note. Mean percentages (across the five time points) are reported with standard deviations in parentheses. We also report the results of independent-samples t-tests comparing the percentage of participants who received each treatment in the Case Management Group versus the Community Group. Except for tutoring and overall parent support, equal variances were assumed. Note that “other services” mainly included referral for additional support, in-home case management (according to the study protocol), and occupational/physical therapy. In-home case management, which most children in the case management group received, reflects the prevalence of the intervention in the case management group.

Follow-Up Assessments

All children with ADHD received comprehensive social, academic, and emotional evaluation at 6, 12, 18, and 24-month follow-up. At these follow-up points, the Services for Children and Adolescents Parent Interview (SCAPI) was also administered to the parents (Jensen et al., 2004) This provided a view of what services were obtained in the community.

Establishing Similarities at Both Sites

Evaluation and treatment were carried out at two sites after staff were trained to administer and score all measures in a similar fashion. Some interventions occurred at one site for all subjects (e.g., parent training and social skills training), while academic skills training occurred at both sites, one site providing the academic skills in English and the other in French, depending on the language of instruction at school for the subject for which the child was receiving academic skill training. Medication titration to optimal dose occurred at both sites after it was established that the child was responsive to methylphenidate. Parent and teacher Conners global index 10 item questionnaires were used to establish severity and optimal dose and systematic side effects scale was used to monitor side effects. Optimal dose was determined by the best teacher and parent ratings on the Conners scale with side effects being tolerable and not too severe. Common guidelines with regards to titration of different medication and approach to side effects were established between the two sites.

Vignettes and parent and teacher Conners ratings (average from two teachers) were exchanged by the two sites to ensure similarities in interpretation of these Conners ratings. Any discrepancies were discussed until a mutual agreement on the interpretation of these ratings was reached.

Systematic Comprehensive Evaluations

All children referred to the study received a comprehensive psychiatric evaluation by a licensed trained child psychiatrist to establish a Diagnostic Statistical Manual IV (DSM-IV) diagnosis of ADHD and any comorbidities. Children were excluded from the study if they failed to meet diagnosis criteria for ADHD or had comorbidities severe enough to require treatment in their own right. Children also received an IQ assessment via Wechsler Intelligence Scale for Children (WISC-III and IV) administered by the child psychologist or a student under the supervision of a child psychologist to determine if they had IQs over 80 (Wechsler, 1991, 2003).

The Conners Parent/Teacher (CGI-P and CGI-T) Global Rating Scales were used as the primary measures to determine ADHD severity and degree of improvement over the course of treatment. They assess children’s attention, impulsivity, emotional outburst, and motor hyperactivity.

Children were also tested on Academic Skills via the Wechsler Individual Achievement Test (WIAT) word reading, reading comprehension, numerical operations, math reasoning, and listening comprehension by the same staff member (Breaux, 2020).

The Social Skills Rating System (SSRS) was completed by the parents and teacher after 3 months of treatment and at each time point (Gresham & Elliott, 1990). It included ratings on cooperation, assertion, responsibility, and self-control.

Parents completed the Achenbach or Child Behavior Checklist (a measure of behavioral and emotional functioning) and Weiss Functional Impairment Rating Scale (WFIRS; a measure of global functioning) at baseline, after 3 months of treatment, and at 6, 12, 18, and 24 month follow ups (Achenbach, 1991; Weiss et al., 2018).

Statistical Analyses

Treatment effects were analyzed using two-way, mixed design ANOVAs with the repeated measure time points (baseline vs. end of treatment) and independent measure of group.

Maintenance effects were analyzed using two-way, mixed design ANOVAs with factors time points (6, 12, 18, 24 months relative to end of treatments) and group membership. Baseline scores were entered as covariates in these analyses.

Missing data were handled using listwise deletion (complete-case analysis). Control analyses comparing dropouts and retained participants did not indicate a systematic pattern rendering differential attrition in the two groups an unlikely explanation for the observed effects, including null findings (Appendix 2).

Results

Participant Numbers and Attrition

The consort diagram (Figure 1) outlines participant numbers and attrition. Two hundred participants, children age 6 to 12 years with normal (over 80) Intelligence Quotient (I.Q.) and DSM-IV diagnosis of ADHD were recruited for the study. Fourteen were excluded before randomization: 7 because they did not participate in 75% of the initial 3 month treatment and 7 left because they did not have time for the study. One hundred and eighty six were randomized, 90 to the case management group and 96 to the community group. Six of the 90 of the case management group were lost to follow up, 2 at 6 month follow up, 2 at 12 month follow up, and 2 at 24 month follow up, leaving a total of 84 who completed the 2 year follow up. Nineteen of the 96 of the community group were lost to follow up, 7 at 6 month follow up, 3 at 12 month follow up, 6 at 18 month follow up, and 3 at 24 month follow up, leaving a total of 77.

Figure 1.

Participant numbers and attrition.

Attrition was clearly greater in the community follow up group with 19 participants lost to follow up compare to 6 in the case management group.

Patient Characteristics

Table 1 provides a detailed description of the participants’ characteristics. The two groups (case management and the community group) were matched on age (mean age 8.71 and 8.85 years old respectively). Most participants in both groups were male (66 and 72, respectively).

The socioeconomic status (SES) in both groups was similar with a median income of level 5 in a range of 1 to 6 ($75,000–$99,999 per year).

The average IQ measured was 95.8 for the case management group and 97.0 for the community group. This difference was not significant (t(177) = −0.983, p = .327, d = 0.147). Most of the participants in both groups were Caucasian (75 in the case management, 82 in the community group).

ADHD severity as assessed by Conners parent global ratings (M= 76.8, SD= 9.6 in the case management group and M= 77.2, SD= 9.7 in the community group) as well as ratings of hyperactivity-impulsivity (M= 76.9, SD= 15.1 in the case management group and M= 78.8, SD= 4.2 in the community group) and inattention (M= 83.5, SD= 7.06 in the case management group and M= 83.2, SD= 5.51 in the community group) were also similar in both groups. Comorbid anxiety in both groups was low (7 and 3 cases in the case management group and community group, respectively), while Oppositional Defiant Disorder (ODD) was high, at 50 cases and 54 cases, respectively, again similar in both groups. The baseline Wechsler Individual Achievement Test (WIAT) was also not significantly different for both groups on all measures.

Table 2 provides a brief outline and comparison of services received by the intervention and the community control group. Services received by both groups were similar with regard to medication, family therapy, parent training, counseling, and academic services with regard to an Individual Educational Plan, tutoring, and academic testing. Individual services for the child were much more frequent in the intervention groups versus the community control group 95.75% of children in the intervention group received these services versus 19.28% in the community control group. This most probably reflects individual meeting with the child and family to clarify information obtained in the regular monthly calls to the child and family as part of the intervention protocol. A more detailed outline of services received by the community group at the end of treatment and during follow up at 6, 12, 18, 24 months appears in the table in the Appendix 1: Community Followed Participants.

The pattern of results on most measures appears to be similar. This pattern consists of a marked improvement on most measures from baseline to the end of the 3 month treatment phase. This is followed by generally maintenance of the treatment effect at 6, 12, 18, and 24 months. This pattern is seen in both the monthly follow up and the community groups.

Conners Parent/Teachers

Specifically for parent ratings of inattention and hyperactivity/impulsivity on the parent Connors rating scale, there is a significant improvement from baseline to the end of the 3 month phase (Inattention scale: F(1, 180) = 253.18, p < .001, ${\eta }_p^2=0.584$ ; Hyperactivity/Impulsivity scale: F(1, 180) = 204.72, p < .001, ${\eta }_G^2=0.532$ ). This improvement was generally maintained at 6, 12, 18, and 24 months. The main effect of time point was not significant (Inattention scale: F(3, 396) = 2.19, p = .088, ${\eta }_p^2=0.016$ ; Hyperactivity/Impulsivity scale: F(3, 396) = 0.93, p = .429, ${\eta }_p^2=0.007$ ), and no significant difference between the monthly follow up and the community group for the Inattention scale (F(1, 132) = 0.30, p = .588, ${\eta }_p^2=0.002$ ). For the Hyperactivity/Impulsivity scale, however, the ANOVA yielded a significant main effect of group with a large relapse in the community group (F(1, 132) = 6.05, p = .015, ${\eta }_p^2=0.044$ ; Figure 2).

Figure 2.

Inattention and hyperactivity measures.

Note. Evolution of t-scores on parent- and teacher-rated Conners scales (Inattention and Hyperactivity/Impulsivity) across six time points, ranging from baseline (BL) to 24 months, shown separately for Case Management (solid, blue line) and Community groups (dashed, red line). Error bars (95% confidence intervals) are plotted around mean t-scores per group. A slight between-group offset has been added along the x-axis for display purposes only.

The teacher ratings of inattention and hyperactivity/impulsivity on the teachers Connors rating scale with significant improvement from baseline to 3 months (Inattention scale: F(1, 157) = 81.78, p < .001, ${\eta }_p^2=0.343$ ; Hyperactivity/Impulsivity scale: F(1, 178) = 189.91, p < .001, ${\eta }_p^2=0.516$ ) and the maintenance of this improvement at 6, 12, 18, and 24 months, as indicated by a non-significant effect of time point (Inattention scale: F(3, 369) = 1.19, p = .313, ${\eta }_p^2=0.010$ ; Hyperactivity/Impulsivity scale: F(3, 369) = 0.32, p = .812, ${\eta }_p^2=0.003$ ), with no significant difference between the monthly follow up group and the community group (Inattention scale: F(1, 123) = 0.01, p = .912, ${\eta }_p^2=0.000$ ; Hyperactivity/Impulsivity scale: F(1, 123) = 0.38, p = .537, ${\eta }_p^2=0.003$ ). No significant interaction between time point and group was found for the Inattention scale (F(3, 369) = 1.74, p = .159, ${\eta }_p^2=0.014$ ). However, we found a significant time point by group interaction for the Parent Conners and some Teacher Conners Hyperactivity/Impulsivity scale (F(3, 369) = 3.57, p = .014, ${\eta }_p^2=0.028$ ), likely driven by higher average t-scores in the Community versus the Case Management group at 18 months follow-up (note, however, that these group effects were resolved at 24 months follow-up; see (Figure 2D)).

Social Emotional Functional Rating

Similarly, the Social Skills Rating System (SSRS), reflecting cooperation, assertion, responsibility, and self control as rated by parents and teachers showed significant improvements by both raters between baseline and the end of the 3 months of treatment (parent-rated SSRS: F(1, 75) = 46.94, p < .001, ${\eta }_p^2=0.385$ ; teacher-rated SSRS: F(1, 178) = 104.00, p < .001, ${\eta }_p^2=0.368$ ). This improvement was maintained according to both raters (i.e., there was no significant effect of time point in the maintenance analysis (parent-rated SSRS: F(3, 177) = 1.68, p = .172, ${\eta }_p^2=0.028$ ; teacher-rated SSRS: F(3, 357) = 0.63, p = .599, ${\eta }_p^2=0.005$ ) and no significant differences between the monthly follow up group and the community group at 6, 12, 18, and 24 months (parent-rated SSRS: F(1, 59) = 0.09, p = .767, ${\eta }_p^2=0.001$ ; teacher-rated SSRS: F(1, 119) = 1.25, p = .265, ${\eta }_p^2=0.010$ ). Further, no significant interactions between time point and group were found (parent-rated SSRS: F(3, 177) = 0.39, p = .760, ${\eta }_p^2=0.007$ , teacher-rated SSRS: F(3, 357) = 0.84, p = .475, ${\eta }_p^2=0.007$ ; Figure 3).

Figure 3.

Social skills, behavioral problems, and functional impairment. (A, B) Evolution of Cooperation, Assertion, Responsibility, Self-Control standard scores, as assessed by the parent- and teacher-rated Social Skills Rating Scale, across six time points, ranging from baseline (BL) to 24 months, shown separately for Case Management (solid, blue line) and Community groups (dashed, red line). (C, D) Evolution of total scores on the Child Behavior Checklist and Weiss Functional Impairment Rating Scale across six time points, ranging from baseline (BL) to 24 months, shown separately for Case Management (solid, blue line) and Community groups (dashed, red line).

Note. Error bars (95% confidence intervals) are plotted around mean t-scores per group. A slight between-group offset has been added along the x-axis for display purposes only.

Likewise, the Child Behavioral Checklist (Achenbach) measuring emotional problems and the Weiss Functional Impairment Rating Scale (WFIRS) measuring both emotional and functional problems completed by the parents showed significant decreases in these problems between baseline to the end of treatment (Achenbach: F(1, 180) = 133.26, p < .001, ${\eta }_p^2=0.425$ ; WFIRS: F(1, 180) = 101.54, p < .001, ${\eta }_p^2=0.361$ ). These treatment effects were maintained at 6, 12, 18, and 24 months with further significant improvement in the Achenbach (F(3, 393) = 3.39, p = .018, ${\eta }_p^2=0.025$ ) and the WFIRS (F(3, 393) = 3.37, p = .019, ${\eta }_p^2=0.025$ ) in both groups. There were no significant differences between groups (Achenbach: F(1, 131) = 1.81, p = . 181, ${\eta }_p^2=0.014$ ; WFIRS: F(1, 131) = 0.17, p = .680, ${\eta }_p^2=0.001$ ). No significant interactions between time point and group were found for the Achenbach (F(3, 393) = 0.37, p = .772, ${\eta }_p^2=0.003$ ) or the WFIRS (F(1, 131) = 0.86, p = .463, ${\eta }_p^2=0.007$ ).

The academic achievement tests via Wechsler Individual Achievement Test (WIAT) was administered at baseline, 12 months, and 24 months. Word reading, reading comprehension, numerical operations, and math reasoning were measured. At each time point the case management group appeared better than the community group on these measures though the differences did not reach statistical significance at any point.

In sum, the results showed significant improvements in all measures between baseline and the end of treatment at 3 months. This improvement was generally maintained or further improved at 6, 12, 18, and 24 month follow ups in both the monthly follow up group and the community control group with generally no significant difference between the groups in maintaining these treatment effects (with the exception of parent-rated Hyperactivity/Impulsivity which deteriorated in the community group).

Discussion

The efficacy of well titrated long acting methylphenidate medication to an individual optimal dose combined with parent training, social skills training, and academic training on a weekly basis for 3 months proved very effective with significant improvement in most domains.

This is not surprising as other multimodal treatment studies for children with ADHD which combined medication and a number of psychosocial treatments; for example, the MTA and the early multimodal study by Abikoff and Hechtman also showed significant improvement in many domains (Abikoff, Hechtman, Klein, Gallagher, et al., 2004; Abikoff, Hechtman, Klein, Weiss, et al., 2004; Hechtman, Abikoff, Klein, Greenfield, et al., 2004; The MTA Cooperative Group, 1999, 2004a, 2004b).

However, the maintenance of treatment effect in both the monthly follow up and the community follow up group was not in line with our expectations. We expected the intervention group which received regular monthly follow up by our study staff to better maintain treatment gains compared to the community follow up group.

The monthly follow-up interventions were generally brief and focused on the most important problems. Despite their limitations, this follow-up strategy did succeed in maintaining treatment gains for 2 years. The only measure that was significantly worse in the community group was Conners Parent ratings of Hyperactivity/Impulsivity, Some time points, for example, 18-month follow up were also worse on the teacher ratings of hyperactivity/impulsivity for the community group. The monthly follow up may have been useful in this regard. However, the lack of differences between the monthly follow-up and the community follow up group on most measure is still striking.

Generally, all patients in the community group continued on long-acting stimulant medication (96.23%). This may have been influenced by our careful initial titration to optimal dose, recommendations to the child’s pediatrician after the 3 months of intensive intervention and the parent training during the intensive initial intervention. Parents had a better grasp as to how to deal effectively with their child with ADHD and also saw the significant benefit of well-titrated medication for their child. The fact that there is a universal pharmacare plan in this province is another factor, which enables most families to have their children on long-acting stimulants free of charge. The high compliance (91.9%–96.2%) with long acting free stimulant medication treatment at optimal dose occurred in both the case management and community groups and may well have contributed to the positive outcomes in both groups over the 2 years. It is striking that medication compliance in this study after 2 years was much higher than seen in other comparable studies, for example, the MTA. The ongoing availability of long acting free stimulant medication (via the Provincial Drug Plan) and parent training received in the first 3 months as well as regular research follow-ups every 6 months may well have contributed to this improved compliance in this follow up study. Many also received academic support at school via an Individualized Education Plan (IEP; 63.35%–63.73%) and psychoeducational support. Some (47.95%–36.82%) also received individual tutoring in various subjects, which their parents arrange privately.

The Hechtman, Abikoff, and Klein (Abikoff, Hechtman, Klein, Gallagher, et al., 2004; Hechtman, Abikoff, Klein, Greenfield, et al., 2004; Hechtman, Abikoff, Klein, Weiss, et al., 2004) medication and multimodal treatment follow up study also had monthly follow up in the second year which may well have contributed to the maintenance of treatment gains.

In summary, the community group did receive a range of services in the community, which may account for the similarity in maintenance effects between the monthly follow up group and the community group. This may have been made possible by the availability of a provincial drug plan, a mandate for schools to provide an Individualized Education Plan for children with ADHD, and the relatively high SES level (EG level 5 on the Hollingshead SES scale) of both groups. In the absence the above availability of free long-acting stimulant medication, a school mandated individual educational plan for ADHD and higher socio-economic status, which allow parents to provide tutoring and other supports, regular monthly follow up, may be more urgently required to maintain treatment gains.

Limitations

Even though we documented what the community group received in the community, we could not control what they received. Thus, what this group received in the community may well have helped them maintain their treatment gains thus resulting in the similarities in this maintenance between the monthly follow up and the community groups.

The monthly follow up intervention may not have been sufficiently robust thus decreasing the possible differences between the monthly follow up and the community follow up group. However, since the treatment gains were generally maintained for 2 years by both groups this possibility is less likely.

Although complete-case analysis seemed unproblematic in the present study (see the control analysis under “Statistical analyses”), it is worth noting that future studies with the appropriate data structure could be encouraged to employ linear mixed models or multiple imputation methods, provided data are missing at random.

Conclusions

Children age 6 to 13 years old with ADHD, IQs over 80, and comorbidities which are mild enough not to require active treatment do well with 3 months of comprehensive treatment which includes carefully titrated stimulant medication to optimal dose, as well as parent training, social skills training, and academic training on a weekly basis. Treatment gains after 3 months of this intensive treatment are significant in ADHD symptoms, and social, emotional, and academic functioning.

These treatment gains were maintained for 2 years as documented by comprehensive assessments every 6 months both in the monthly follow up group and the community group. The latter received services (mainly stimulant medication and an Individualized Education Plan) in the community which may have helped them maintain treatment gains without the systematic follow up.

Thus, 3 months of intensive comprehensive treatment which includes careful medication titration, weekly small group social skills training, academic training, parent training (including the understanding of the importance of ongoing medication) with some ongoing school and academic supports appears helpful in maintaining treatment gains for 2 years.

Author Biographies

Dr. Lily Hechtman is a Professor of Psychiatry and Pediatrics at McGill University and Director of Research in the Division of Child Psychiatry. She is an internationally recognized researcher in ADHD. Dr. Hechtman’s research has focused on long-term prospective follow-up studies and multimodal treatment studies of children with ADHD.

Dr. Natalie Grizenko is associate professor in the Department of Psychiatry at McGill University. She is the medical director of the Disruptive Behavior Disorders Program and ADHD Clinic at the Douglas Mental Health University Institute. She is also a researcher at the Douglas Research Center.

Dr. Ridha Joober is a psychiatrist, researcher, and Full Professor at McGill University specializing in schizophrenia and ADHD. His research spans molecular mechanisms, clinical research, and mental health services. Author of over 450 publications, he is a Fellow of the Canadian Academy of Health Sciences and Foreign Corresponding Member of France’s National Academy of Medicine.

Dr. Jens Kreitewolf is a Faculty Lecturer at McGill University. He received his Ph.D. in Psychology from Humboldt University of Berlin. His research focuses on auditory cognition, particularly speech and voice perception, including how listeners use voice information and talker familiarity to understand speech under challenging everyday listening conditions.

Mariam Ter-Stepanian, PhD, is a researcher specializing in ADHD, learning disorders, and executive functions. She holds a doctorate degree from McGill University and completed postdoctoral training at the Université de Sherbrooke. Her work includes providing consultation, and direct intervention for children with neurodevelopmental disorders and the professionals who support them.

Marie-Ève Fortier holds a Ph.D. and worked as a research assistant involved in data gathering and analysis on this project at the ADHD clinic at the Douglas Mental Health University Institute.

Kayla Eisenberg is a doctoral student in Clinical Psychology at Concordia University. Prior to her graduate studies at Concordia, she completed a master’s degree in Psychiatry research at McGill University. Broadly, her research interests focus on improving the effectiveness of ADHD interventions for youth.

Joy Etcovitch, M.A., was a research coordinator at McGill University Health Centre. She was involved in ADHD clinical research and major clinical trials examining the diagnosis, treatment, and long-term outcomes of children and adolescents with ADHD, including the landmark Multimodal Treatment Study of ADHD (MTA).

Dr. William Harvey is the Director of the Choices in Health, Action, Motivation, Pedagogy and Skills (CHAMPS) physical activity lab at the Department of Kinesiology and Physical Education, McGill University. His research group studies self-regulation of persons with and without ADHD in physical activity and leisure interventions within community settings.

Appendix 1

Community Followed Participants

Table A1.

Proportion of Participants Who Received Specific Services During the Follow-Up Period.

Time point (months)	End of treatment-6	6–12	12–18	18–24
Medication, N(%)	75/75	74/74	71/71	72/72
Stimulants	72/75 (96)	71/74 (95.9)	68/71 (95.8)	70/72 (97.2)
Non-stimulants	8/75 (10.7)	9/74 (12.2)	10/71 (14.1)	12/72 (16.7)
Individual services, N(%)	12/75 (16.0)	15/74 (20.3)	17/71 (23.9)	12/71 (16.9)
Individual therapy	8/75 (10.7)	13/74 (17.6)	13/71 (18.3)	9/71 (12.7)
Other (specify)	4/75 (5.3)	4/74 (5.4)	2/71 (2.8)	3/71 (4.2)
Family therapy, N(%)	4/74 (5.4)	6/74 (8.1)	3/71 (4.2)	5/72 (6.9)
Academic services, N(%)	64/75 (85.3)	59/74 (79.7)	56/71 (78.9)	55/72 (76.4)
IEP	46/75 (61.3)	48/74 (64.9)	48/71 (67.6)	44/72 (61.1)
Tutoring	22/75 (29.3)	25/73 (34.2)	30/71 (42.3)	29/70 (41.4)
Testing, N(%)	4/75 (5.3)	3/74 (4.1)	5/71 (7.0)	9/72 (12.5)
Parent support, N(%)	6/75 (8.0)	8/73 (11.0)	12/71 (16.9)	13/72 (18.1)
Training classes & support groups	4/75 (5.3)	2/73 (2.7)	2/71 (2.8)	4/72 (5.6)
Counseling	2/75 (2.7)	5/73 (6.8)	6/71 (8.5)	5/72 (6.9)

Note. The total number of participants varies based on the number of people who completed the SCAPI interview and answered the appropriate question for the service. N represents the number of participants that meet each criterion. % represents the valid percentage of participants that meet each criterion (i.e., the percentage of the participants who responded to the question; displayed in brackets). This table includes the main categories of service of interest as well as the specific services within these categories that were the most common.

Appendix 2

Teacher – Hyperactivity.

Drop-out at follow-up (months)	Case management group	Community group
3	n= 2, Δ = 10.61 (6.21)	n= 2, Δ = 11.95 (4.35)
6	n= 4, Δ = −4.58 (11.83)	n= 21, Δ = −4.35 (4.41)
12	n= 6, Δ = −11.21 (7.16)	n= 19, Δ = −5.63 (4.44)
18	n= 9, Δ = −10.45 (6.03)	n= 26, Δ = −2.25 (3.89)
24	n= 10, Δ = −11.39 (6.61)	n= 25, Δ = −3.00 (3.83)

Note. Values represent baseline mean differences between drop-outs and retained participants (Δ = Drop-out − Retained). Standard errors of the mean difference are shown in parentheses. Comparisons based on low n were added for completeness only.

Parent – Hyperactivity.

Drop-out at follow-up (months)	Case management group	Community group
3	n= 2, Δ = 13.40 (1.62)	n= 1, Δ = 11.34 (SE n/a)
6	n= 3, Δ = 9.07 (4.63)	n= 17, Δ = 4.85 (3.11)
12	n= 5, Δ = 2.86 (7.76)	n= 16, Δ = −0.79 (3.76)
18	n= 7, Δ = −1.44 (7.67)	n= 25, Δ = −0.84 (3.48)
24	n= 6, Δ = −2.39 (8.08)	n= 20, Δ = −0.99 (3.67)

Teacher – Inattention.

Drop-out at follow-up (months)	Case management group	Community group
3	n= 2, Δ = 0.83 (10.04)	n= 2, Δ = −1.50 (4.60)
6	n= 4, Δ = −2.81 (5.31)	n= 21, Δ = −1.00 (2.56)
12	n= 6, Δ = −2.97 (4.07)	n= 19, Δ = −2.78 (2.72)
18	n= 9, Δ = −3.67 (3.06)	n= 26, Δ = 2.65 (2.25)
24	n= 10, Δ = −3.70 (2.96)	n= 25, Δ = 0.64 (2.15)

Table A2.

Baseline Differences Between Drop-Outs and Retained Participants Across Outcome Measures.

Parent – Inattention.

Drop-out at follow-up (months)	Case management group	Community group
3	n= 2, Δ = 2.08 (1.68)	n= 1, Δ = 6.88 (SE n/a)
6	n= 3, Δ = 0.90 (5.72)	n= 17, Δ = 3.06 (2.12)
12	n= 5, Δ = 3.95 (1.56)	n= 16, Δ = −0.90 (2.79)
18	n= 7, Δ = 3.99 (1.58)	n= 25, Δ = −2.75 (2.50)
24	n= 6, Δ = −0.68 (4.76)	n= 20, Δ = −3.03 (2.71)

Parent – SSRS CAS.

Drop-out at follow-up (months)	Case management group	Community group
3	n= 0	n= 0
6	n= 1, Δ = −6.57 (SE n/a)	n= 4, Δ = −0.98 (5.42)
12	n= 3, Δ = −9.48 (2.63)	n= 5, Δ = 2.84 (3.85)
18	n= 3, Δ = −8.03 (2.28)	n= 9, Δ = −2.19 (4.46)
24	n= 2, Δ = −7.81 (2.38)	n= 9, Δ = 3.16 (4.25)

Teacher – SSRS CAS.

Drop-out at follow-up (months)	Case management group	Community group
3	n= 3, Δ = −5.42 (4.78)	n= 1, Δ = 8.32 (SE n/a)
6	n= 5, Δ = 1.94 (4.63)	n= 21, Δ = 0.33 (3.93)
12	n= 6, Δ = −1.33 (3.92)	n= 19, Δ = 1.44 (4.26)
18	n= 10, Δ = 7.96 (2.57)	n= 26, Δ = −1.54 (3.48)
24	n= 13, Δ = −2.20 (2.42)	n= 25, Δ = 0.39 (3.48)

Note. Values represent baseline mean differences between drop-outs and retained participants (Δ = Drop-out − Retained). Standard errors of the mean difference are shown in parentheses. Comparisons based on low n were added for completeness only.

CBCL total scores.

Drop-out at follow-up (months)	Case management group	Community group
3	n= 2, Δ = 6.23 (0.92)	n= 1, Δ = −10.60 (SE n/a)
6	n= 4, Δ = −2.52 (4.45)	n= 18, Δ = 2.90 (2.06)
12	n= 5, Δ = 1.05 (3.71)	n= 15, Δ = 1.72 (2.42)
18	n= 7, Δ = 0.79 (2.75)	n= 25, Δ = 0.43 (2.01)
24	n= 6, Δ = 0.99 (3.26)	n= 20, Δ = −1.25 (2.10)

Note. Values represent baseline mean differences between drop-outs and retained participants (Δ = Drop-out − Retained). Standard errors of the mean difference are shown in parentheses. Comparisons based on low n were added for completeness only.

WFIRS total scores.

Drop-out at follow-up (months)	Case management group	Community group
3	n= 2, Δ = 0.36 (0.17)	n= 1, Δ = 0.29 (SE n/a)
6	n= 3, Δ = 0.02 (0.27)	n= 17, Δ = 0.15 (0.11)
12	n= 5, Δ = 0.17 (0.24)	n= 18, Δ = 0.01 (0.11)
18	n= 8, Δ = 0.08 (0.17)	n= 26, Δ = 0.16 (0.10)
24	n= 6, Δ = 0.11 (0.18)	n= 20, Δ = 0.06 (0.11)

Note. Values represent baseline mean differences between drop-outs and retained participants (Δ = Drop-out − Retained). Standard errors of the mean difference are shown in parentheses. Comparisons based on low n were added for completeness only.

Appendix 3

Parent Training

ADHD Clinic Parent Training Program

Overview: This parent training program has been developed at the ADHD clinic with an intent to provide parents with a comprehensive overview of ADHD and integrated essential components of parenting approaches that had been at the clinic. The overarching philosophy of the approach is based on Collaborative & Proactive Solutions (CPS) approach (Greene & Winkler, 2019).

Materials: The material was presented as a presentation slides and parents were given handouts. The themes were: Psychoeducation about ADHD, active listening, building a relationship, implementing structure and routine, introduction to CPS approach https://livesinthebalance.org/.

Program leaders: The program was developed by the clinical team at the ADHD clinic. The training for the research project was provided by three leaders (in French and English) who were leading the groups prior to commencement of the research. All co-leaders were trained by these professionals. The following professionals lead the clinical team at the ADHD clinic: a clinical psychologist (leader), a social worker (leader), a nurse clinician (leader), and masters level students in psychology acted as co-leaders.

Setting and delivery: The program consisted of 8-week, 2-hr, group workshops. All workshops were held on Saturday morning at the Douglas Institute ADHD clinic. The material was presented to the families and was followed by discussions. After the training the parents were asked to complete a satisfaction questionnaire.