Efficacy and acceptability of pharmacological, psychosocial, and brain stimulation interventions in children and adolescents with mental disorders: an umbrella review

Abstract

Top‐tier evidence on the safety/tolerability of 80 medications in children/adolescents with mental disorders has recently been reviewed in this jour­nal. To guide clinical practice, such data must be combined with evidence on efficacy and acceptability. Besides medications, psychosocial inter­ventions and brain stimulation techniques are treatment options for children/adolescents with mental disorders. For this umbrella review, we systematically searched network meta‐analyses (NMAs) and meta‐analyses (MAs) of randomized controlled trials (RCTs) evaluating 48 medications, 20 psychosocial interventions, and four brain stimulation techniques in children/adolescents with 52 different mental disorders or groups of mental disorders, reporting on 20 different efficacy/acceptability outcomes. Co‐primary outcomes were disease‐specific symptom reduction and all‐cause discontinuation (“acceptability”). We included 14 NMAs and 90 MAs, reporting on 15 mental disorders or groups of mental disorders. Overall, 21 medications outperformed placebo regarding the co‐primary outcomes, and three psychosocial interventions did so (while seven outperformed waiting list/no treatment). Based on the meta‐analytic evidence, the most convincing efficacy profile emerged for amphetamines, methylphenidate and, to a smaller extent, behavioral therapy in attention‐deficit/hyperactivity disorder; aripiprazole, risperidone and several psychosocial interventions in autism; risperidone and behavioral interventions in disruptive behavior disorders; several antipsychotics in schizophrenia spectrum disorders; fluoxetine, the combination of fluoxetine and cognitive behavioral therapy (CBT), and interpersonal therapy in depression; aripiprazole in mania; fluoxetine and group CBT in anxiety disorders; fluoxetine/selective serotonin reuptake inhibitors, CBT, and behavioral therapy with exposure and response prevention in obsessive‐compulsive disorder; CBT in post‐traumatic stress disorder; imipramine and alarm behavioral intervention in enuresis; behavioral therapy in encopresis; and family therapy in anorexia nervosa. Results from this umbrella review of interventions for mental disorders in children/adolescents provide evidence‐based information for clinical decision making.

Many mental disorders have an onset with clinically relevant manifestations in childhood or adolescence, followed frequently by a chronic illness course into adulthood1, 2. Many disorders with an earlier onset are first diagnosed in adulthood, with a delay ranging for example from 6 to 8 years for mood disorders and from 9 to 23 years for anxiety disorders ³ . Due to their interference with attainment of biopsychosocial milestones, mental and neurodevelopmental disorders in children and adolescents are among the leading causes of global burden of disease and years lived with disability ⁴ . This situation makes the appropriate delivery of evidence‐based and effective treatments for youth with mental disorders a key priority in the public health field.

Pharmacological, psychosocial and brain stimulation options are available for the management of many mental disorders in children and adolescents. However, for several of them, what should be considered the first line treatment strategy – based on efficacy, effectiveness, acceptability and tolerability/safety – remains uncertain.

A number of randomized controlled trials (RCTs) have been conducted to assess the efficacy, acceptability and tolerability of medications across different disorders in children and adolescents. The results from many of these RCTs have been pooled in pairwise meta‐analyses (MAs) or network meta‐analyses (NMAs)5, 6, 7, 8. While most antidepressants outperform placebo to treat depression in adults ⁹ , most antidepressants have not been shown to be superior to placebo in children and adolescents with major depressive disorder7, 10. Similarly, yet to a lower extent, antidepressants may not be as effective in children and adolescents with anxiety disorders as in adults ¹¹ .

On the other hand, RCTs comparing psychosocial interventions with waiting list or no intervention control groups generally show a large effect size in youth with depression ¹⁰ or anxiety ¹² disorders. Yet, when compared with placebo/sham interventions, most significant findings favoring psychosocial interventions vs. placebo disappear10, 12. Effect sizes also vary according to design, blinding, patient selection (baseline severity) and choice of the control group ¹³ in trials assessing combination treatments, whose superiority to monotherapies has not been consistently confirmed within and across disorders in children/adolescents.

Differences in inclusion criteria, outcomes, and a variety of features defining quality across MAs and NMAs limit the clinical value and impact of such a rich, yet complex body of evidence. Umbrella reviews may overcome these problems to some degree by taking the totality of the evidence from existing MAs and NMAs into account, and filtering top‐tier meta‐analytic estimates according to pre‐established criteria. It is paramount to provide clinicians with structured and standardized summaries, translating the massive data into actionable clinical information.

To our knowledge, no umbrella review is available of the evidence from MAs and NMAs of RCTs on the efficacy and acceptability of pharmacological, psychosocial, and brain stimulation treatment options for the core symptoms and associated problems of the full range of mental disorders in children and adolescents. The present study aims to fill this gap, as previously done in this journal concerning the safety and tolerability of 80 pharmacological agents used for the management of child and adolescent mental disorders ¹⁴ .

We focused on disease‐specific symptom reduction and treatment response as efficacy measures, and on measures of acceptability that could be compared across the three different treatment modalities, namely all‐cause discontinuation and intolerability‐related discontinuation. Following this approach, this umbrella review intends to provide practitioners with an evidence‐based atlas of therapeutic tools to inform clinical decision making, where a balance needs to be struck between efficacy, acceptability/tolerability, and safety.

METHODS

Search, inclusion and exclusion criteria

This umbrella review followed an a priori protocol (available upon request). We conducted a systematic search in PubMed, PsycINFO, and Cochrane database up to January 9, 2021, using an exhaustive combination of key words (full search string available upon request). We also manually searched bibliographies of included meta‐analyses. Two independent authors conducted title/abstract screening, full‐text assessment, and data extraction into a pre‐defined excel spreadsheet. A third author triple‐checked extracted data, and resolved any conflict.

Included were: a) NMAs or MAs of RCTs, b) of a priori defined 48 psychotropic medications, 20 psychosocial interventions, and four brain stimulation interventions, c) in children and/or adolescents, d) with any of 52 a priori defined mental disorders, e) reporting on 20 a priori defined outcomes within a specific disorder. Exclusion criteria were: a) systematic reviews without meta‐analysis, b) pooling of studies other than RCTs, c) interventions for other than pre‐defined disorders/outcomes.

Whenever two NMAs or MAs reported on the same combination of disorder, intervention, comparison and outcome, we considered the comparison with more RCTs, the minimum being at least one direct comparison for NMAs.

Included disorders, interventions, and comparisons

Mental disorders of interest, as grouped in the ICD‐11 ¹⁵ , were: a) neurodevelopmental disorders (autism spectrum disorder, attention‐deficit/hyperactivity disorder (ADHD), disorders of intellectual development, developmental speech or sound disorders, developmental learning disorders, developmental motor coordination disorders), b) schizophrenia and other primary psychotic disorders (schizophrenia, schizoaffective disorder, schizotypal disorder, acute and transient psychotic disorder), c) catatonia, d) mood disorders (bipolar and related disorders, depressive disorders), e) anxiety or fear‐related disorders (generalized anxiety disorder, panic disorder, agoraphobia, specific phobia, social anxiety disorder, separation anxiety disorder, selective mutism), f) obsessive‐compulsive and related disorders (obsessive‐compulsive disorder, body dysmorphic disorder, body‐focused repetitive disorders), g) movement disorders (Tourette's disorder, other tic disorder), h) disorders specifically associated with stress (post‐traumatic stress disorder (PTSD), complex PTSD, prolonged grief disorder, reactive attachment disorder, disinhibited social engagement disorder), i) dissociative disorders (dissociative neurological symptom disorder, dissociative amnesia, trance disorder, dissociative identity disorder), j) feeding and eating disorders (anorexia nervosa, bulimia nervosa, binge eating disorder, avoidant‐restrictive food intake disorder, pica, rumination‐regurgitation disorder), k) elimination disorders (enuresis, encopresis), l) disorders of bodily distress or bodily experience (bodily distress disorder, body integrity dysphoria), m) disorders due to substance use or addictive behaviors, n) impulse control disorders (pyromania, kleptomania, compulsive sexual behavior disorder, intermittent explosive disorder), o) disruptive behavior or dissocial disorders (oppositional defiant disorder, conduct disorder).

Interventions included medications, psychosocial interventions, and brain stimulation techniques.

Medications comprised antidepressants (bupropion, mirtazapine, nefazodone, vilazodone, desvenlafaxine, duloxetine, venlafaxine, citalopram, escitalopram, fluoxetine, fluvoxamine, paroxetine, sertraline, clomipramine, desipramine, imipramine, nortriptyline, amitriptyline); antipsychotics (fluphenazine, haloperidol, molindone, trifluoperazine, amisulpride, aripiprazole, asenapine, clozapine, loxapine, lurasidone, olanzapine, paliperidone, quetiapine, risperidone, thioridazine, ziprasidone); anti‐ADHD medications (amphetamines, atomoxetine, clonidine, guanfacine, methylphenidate, modafinil); mood stabilizers (carbamazepine, lamotrigine, lithium, oxcarbazepine, topiramate, valproate); and others (oxybutynin, desmopressin).

Psychosocial interventions included behavioral therapy, cognitive behavioral therapy (CBT), problem solving, dialectical behavioral therapy, family‐based therapy, interpersonal psychotherapy, mentalization based therapy, psychodynamic psychotherapy, supportive therapy, social skills training, acceptance and commitment therapy, mindfulness, eye movement desensitization and reprocessing, narrative exposure therapy, cognitive remediation therapy, cognitive training, parent‐child interaction therapy, play therapy, art therapy, and occupational therapy.

Brain stimulation interventions included transcranial magnetic stimulation, transcranial direct current stimulation, electroconvulsive therapy, and neurofeedback.

Comparators were labeled as active drug, active psychosocial intervention, treatment as usual (TAU)/low intensity psychosocial intervention, waiting list/no treatment, or placebo/sham.

Outcomes

Co‐primary outcomes were disease‐specific primary symptom reduction and all‐cause discontinuation (“acceptability”).

Secondary continuous outcomes were measures of aggressive behavior, anxiety (other than anxiety disorders), cognition (other than ADHD), depressive symptoms (other than depressive episode/disorder), irritability, suicidal ideation, global illness severity, functioning (as defined by authors), and quality of life.

Secondary categorical outcomes were study‐defined treatment response, remission, relapse, hospitalization, discontinuation due to inefficacy, discontinuation due to intolerability, suicide attempt, completed suicide, and death. When available, treatment estimates from clinicians, teachers, parents, and children/adolescents were considered separately.

Quality of evidence

The quality of MAs and NMAs was measured using A Mea­surement Tool for the Assessment of Multiple Systematic Reviews (AMSTAR‐PLUS)16, 17 to quantify both the methodological quality of MAs and NMAs with the first 11 items (AMSTAR) and of included RCTs with six additional items (AMSTAR‐Content).

Methodological quality was categorized into low (<4), medium (4‐7), and high (>7). Content quality was categorized into low (<4), medium (4‐6), and high (>6). The lowest score between methodological and content quality determined the overall MA or NMA quality.

Statistical analysis

We converted continuous non‐standardized outcomes, such as weighted mean differences, to standardized mean differences (SMDs), and binary outcomes to odds ratio (ORs) with Comprehensive Meta‐Analysis (CMA), Version 3 ¹⁸ . We then calculated the mean SMD for the primary efficacy outcome across pharmacological, psychosocial, and brain stimulation interventions for each disorder against placebo/sham and waiting list/no intervention, as well as for active controlled monotherapy and combination treatment studies, prioritizing clinician rating, followed by teacher, parent, and then subject‐rated estimates. For treatment response, in case no data were available for the continuous primary efficacy outcome, we converted ORs to SMDs, using CMA.

Whenever data conversion was not possible, we kept the original effect sizes as reported. Whenever we included data from meta‐analyses that used fixed‐effects models, we recalculated the meta‐analysis using random‐effects models ¹⁹ . For consistent and easy comparison, we harmonized effect sizes as follows: SMD<0 favors intervention, OR/risk ratio (RR) <1 favors intervention for discontinuation, suicide or relapse, while OR/RR>1 favors intervention for response or remission.

RESULTS

Search results and literature coverage

The search process is described in Figure 1. Out of 5,231 initial hits, we assessed 910 MAs and NMAs at full text level. Of these, we excluded 806, with specific reasons (list available upon request). The list of all included MAs and NMAs is available in Table 1, also indicating the number of included RCTs and participants, as well as the methodological quality (AMSTAR score) together with the quality of included RCTs (AMSTAR‐Content median score).

Figure 1.

PRISMA flow chart, MAs – meta‐analyses, NMAs – network meta‐analyses

Table 1.

Network and pairwise meta‐analyses of randomized controlled trials (RCTs) of pharmacological, psychosocial and brain stimulation interventions in children and adolescents with mental disorders included in the umbrella review

	Source	Number of RCTs/ patients	Intervention	Controls	Outcomes	A	C
Anxiety disorders
Wang et al 79	MA	115/7,719	AD	PBO	PE, REM	10	4
Dobson et al 11	NMA	22/2,623	AD	PBO	RES, ACD, AED, S	7	5
Zhang et al 80	MA	7/358	CB	WL/NT	PE	9	2
James et al 12	MA	87/5,964	CB	PBO, WL/NT, TAU, PS	PE, REM, DEP, F, ACD	11	3
Zhou et al 78	NMA	101/6,625	CB	PBO, WL/NT, TAU, PS	PE, QoL, ACD	11	2
Sigurvinsdóttir et al 81	MA	81/5,913	CB	WL/NT, TAU, PS	REM	10	1
James et al 82	MA	41/1,955	CB	TAU, PS	PE, REM	11	1.5
Anorexia nervosa
Fisher et al 99	MA	21/1,407	FB	TAU, PS	PE, ACD, REM	10	1
van den Berg et al 100	MA	15/1,279	PS	TAU	PE	9	2
Zeeck et al 97	NMA	18/1,247	FB, PSD‐O	PS	PE	7	1
Social anxiety disorder
Yang et al 83	MA	17/1,134	CB	PBO, WL/NT	PE, REM, DEP, QoL, ACD	10	2
Kreuze et al 84	MA	42/3,239	CB	PBO, TAU, LIP	AG, F	10	2.5
Attention‐deficit/hyperactivity disorder (ADHD)
Cortese et al 5	NMA	133/18,199	AD, STIM, α2	PBO, AD, STIM	PE, AED, GLO	11	9
Otasowie et al 22	MA	6/216	AD	PBO	PE, GLO	10	3
Punja et al 23	MA	23/2,675	STIM	PBO	PE, COG, GLO	10	4
Stuhec et al 34	MA	28/4,699	AD	PBO	PE	8	2
Luan et al 21	NMA	73/15,025	AD, STIM, α2	PBO, PHARMA	PE, AED, ID	7	4
Catalá‐López et al 20	NMA	190/26,114	AP, AD, STIM, α2, CB, CT, NF, COMB	PBO	RES, ACD, GLO	10	4
Schachter et al 36	MA	62/2,897	STIM	PBO	AG	9	1
Schwartz et al 37	MA	25/3,928	AD, STIM	PBO	AG, F, QoL, S	7	5
Coghill et al 38	MA	60/1,993	STIM	PBO	COG	8	2
Storebø et al 39	MA	185/12,245	STIM	PBO	QoL	8	5
Bangs et al 40	MA	32/7,248	AD, STIM	PBO	S	3	4
Hirota et al 41	MA	12/2,276	α2+	PBO	PE, ACD, AED, ID	6	3.5
Storebø et al 42	MA	25/2,690	SKILL, COMB	WL/NT	PE, COG, F	11	2
Sun et al 24	MA	8/423	STIM	PBO	PE, ACD, AED	11	2
Battagliese et al 25	MA	24/1,690	BT	MIX	PE, AG, COG, F	7	1
Faraone et al 26	MA	4/216	STIM	STIM	AG	2	3
Van Doren et al 27	MA	10/506	NF	PHARMA, PS	PE, RES, ACD	8	2
Cortese et al 28	MA	16/759	CT	MIX	PE, COG	11	1
Daley et al 29	MA	32/2,077	BT	MIX	PE, COG	9	2
Bikic et al 30	MA	12/1,054	SKILL	MIX	PE, COG	8	2
Mulqueen et al 31	MA	8/399	BT	MIX	PE	6	1
Cortese et al 32	MA	13/520	NF	MIX	PE, COG	9	1.5
Bussalb et al 33	MA	16/706	NF	MIX	PE	4	2
Faraone et al 35	MA	7/384	STIM	PBO	AG	2	2
Autism spectrum disorder
Maneeton et al 44	MA	3/408	AP	PBO	PE, RES, GLO	7	4
Maneeton et al 52	MA	7/372	AP	PBO	REL, RES	7	3.5
Zhou et al 53	MA	64/3,499	STIM	PBO	PP	9	3
Murza et al 54	MA	16/837	SKILL	WL/NT	F	8	0.5
Fletcher‐Watson et al 56	MA	22/695	SKILL	WL/NT, TAU	F	10	1
Sturman et al 55	MA	4/113	STIM	PBO	PE	10	1
Cohen et al 57	MA	15/995	AP	PBO	RES	5	1
Hirota et al 58	MA	7/171	MS	PBO	RES, AG, ACD, AED, ID	6	4
Fallah et al 43	NMA	8/878	AP	PBO, AP	AG	7	1
D'Alò et al 59	MA	15/1,124	AP	PBO	ACD, AED	9	5
Ospina et al 60	MA	69/2,585	BT	WL/NT, PS	PE	9	1
Reichow et al 61	MA	5/196	SKILL	WL/NT	PE	10	1
James et al 12	MA	87/5,964	CB	WL/NT, TAU	ANX	11	0.5
Tachibana et al 62	MA	32/594	PS	TAU	PE	11	1
Nevill et al 63	MA	19/1,205	PCI	TAU/LIP, MIX	PE, COG	5	1
Yu et al 45	MA	14/555	BT	TAU	PE, F	9	0
Oono et al 46	MA	17/919	PCI	MIX	PE, F, GLO	10	1
Parsons et al 47	MA	21/925	SKILL	MIX	PE	9	1
Kreslins et al 48	MA	10/470	CB	MIX	ANX	9	0
Tarver et al 49	MA	9/521	PCI	MIX	AG	8	2
Soares et al 50	MA	18/1,266	SKILL	MIX	F	8	2
Postorino et al 51	MA	8/653	PCI	MIX	IR	8	1
Bipolar disorder, depressive episode
Maneeton et al 106	MA	3/251	AP	PBO	PE, RES, REM, GLO, ACD, AED	9	3
Bipolar disorder, manic episode
Meduri et al 107	MA	22/5,437	AP	PBO	PE, RES, ACD, AED, ID	10	5
Liu et al 108	MA	46/2,666	MS	PBO	RES	7	6
Jochim et al 109	MA	25/3,252	MS, AP	PBO, MS	ACD	10	4
Bulimia nervosa
Linardon et al 101	MA	79/NR	CB	PS	PE	6	0
Depressive disorders
Zhou et al 10	NMA	71/9,510	AD, PSD‐O, FB, CB, COMB	PBO, WL/NT, TAU/LIP, PHARMA, PS	PE, ACD, S	11	5
Cipriani et al 7	NMA	34/5,260	AD	PBO, PHARMA	RES, AED	11	5
Spielmans & Gerwig 64	MA	8/1,756	AD	PBO	QoL	5	5
Kato et al 65	MA	40/8,890	AD	PBO	REL	9	3
Whittington et al 66	MA	2/376	AD	PBO	REM	9	2.5
Watanabe et al 67	MA	27/1,744	PSD‐O	WL/PBO	RES	7	2
Cox et al 68	MA	9/882	AD, CB, COMB	PHARMA, PS	REM, S	10	3
Dubicka et al 69	MA	5/1,206	COMB	PHARMA, PS	RES, F, S	7	3
Klein et al 70	MA	11/809	CB	MIX	PE	8	4
Disruptive behavior/dissocial/conduct disorders
Seida et al 92	MA	62/NR	AP	PBO	PE, AG, GLO	9	3.5
Loy et al 93	MA	10/896	AP	PBO	PE, AG	10	4
Pringsheim et al 94	MA	18/1,195	MS	PBO	AG	10	2
Ipser & Stein 95	MA	14/823	PHARMA	PBO	AG, ACD, GLO, RES	6	1.5
Battagliese et al 25	MA	24/1,690	CB	WL/NT, MIX	PE	7	1.5
McQuire et al 96	MA	14/912	AP, MS	PBO	AG	8	2
Developmental coordination disorder
Miyahara et al 116	MA	15/649	SKILL	WL/NT	PE	10	1
Eating disorders
Couturier et al 98	MA	6/369	FB	PS	REM	8	3
Encopresis
Freeman et al 114	MA	10/562	COMB	TAU	PE, RES	7	1
Brazzelli et al 115	MA	21/1,371	COMB	TAU	RES	10	1
Enuresis
Caldwell et al 86	MA	74/5,983	BT, COMB	PHARMA, PS, WL/NT	PE, RES	11	1
Caldwell et al 87	MA	64/4,071	AD, COMB	PBO, PHARMA, PS	PE, RES	11	1
Caldwell et al 88	MA	16/1,643	BT	PS, WL/NT	RES	10	1
Buckley et al 89	MA	27/1,803	SKILL, COMB	TAU, PHARMA	REM	10	1
Deshpande et al 90	MA	40/2,440	AD, COMB	PHARMA	RES, REL	10	1
Peng et al 91	MA	15/1,502	PHARMA	PS	ACD	9	4
Song et al 85	NMA	18/1,649	PHARMA, COMB	PHARMA, PS	RES, REL	9	4
Obsessive‐compulsive disorder
Skapinakis et al 71	NMA	86/15,585	AD, CB, COMB	PBO, WL/NT, PHARMA, PS	PE, ACD	10	3
Maneeton et al 72	MA	3/188	AD	PBO	RES, GLO	9	2
McGuire et al 73	MA	20/1,296	AD, CB	PBO, TAU/LIP, WL/NT	RES, REM	8	1
Locher et al 74	MA	36/6,778	AD	PBO	AED	10	4
Geller 75	MA	12/1,044	AD	PBO	GLO	8	3
Uhre et al 76	MA	12/791	CB, AD	PBO, WL/NT, PS	REM, F, QoL	9	1
Johnco et al 77	MA	21/1,423	CB, AD	PBO, WL/NT, TAU/LIP, PS	ACD	6	1
Post‐traumatic stress disorder
Gillies et al 117	MA	14/758	CB	WL/NT, TAU/LIP	PE, RES, ANX, DEP, ACD	10	1
Schizophrenia spectrum disorders
Krause et al 102	NMA	28/3,003	AP	PBO, PHARMA	PE, RES, ACD, ID	11	3
Arango et al 103	NMA	13/2,210	AP	PBO, PHARMA	GLO, AED	9	7
Pagsberg et al 8	NMA	12/2,158	AP	PBO, PHARMA	GLO	8	3
Sarkar & Grover 104	MA	15/995	AP	PHARMA	PE	5	1
Kumar et al 105	MA	13/1,112	AP	PHARMA	AED	8	1
Tic disorder
Bloch et al 110	MA	9/477	STIM, AD	PBO	PE	4	1
Yu et al 111	MA	15/1,070	MS	PHARMA	RES	7	3
Tourette's disorder
Hollis et al 112	MA	40/2,422	AP, α2, STIM, BT	PBO, MIX	PE	8	1
Zheng et al 113	MA	6/528	AP	PHARMA	PE	10	2

MA – meta‐analysis, NMA – network meta‐analysis, A – AMSTAR, C – AMSTAR‐Content (median), AD – antidepressants, CB – cognitive‐based, FB – family‐based, PS – active psychosocial, PSD‐O – psychodynamic‐oriented, STIM – stimulants, α2 – α2‐agonists (+=augmentation with), AP – antipsychotics, CT – ­cognition‐targeted, NF – neurofeedback, COMB – combination of more than one treatment, SKILL – skills training, BT – behavioral treatment, MS – mood stabilizers, PCI – parent‐child interaction, PHARMA – mixed medications, PBO – placebo, WL – waiting list, NT – no treatment, TAU – treatment as usual, LIP – low‐intensity psychosocial intervention, MIX – mixed active/inactive control group, PE – primary efficacy outcome, REM – remission, REL – relapse, RES – response, S – suicidality, ACD – all‐cause discontinuation, AED – discontinuation due to adverse events, ID – discontinuation due to inefficacy, DEP – depressive symptoms, ANX – anxiety symptoms, AG – aggressivity, QoL – quality of life, GLO – global illness severity, COG – cognition, F – functioning, NR – not reported

We ultimately included 14 NMAs and 90 MAs, reporting on 15 disorders or groups of disorders. For ADHD, we included three NMAs5, 20, 21 and 21 MAs22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42; for autism, one NMA ⁴³ and 21 MAs12, 44, 45, 46, 47, 48, 49, 50, 51, 52, 53, 54, 55, 56, 57, 58, 59, 60, 61, 62, 63 (including one focusing on comorbid anxiety disorders and autism) ¹² ; for depressive disorders, two NMA7, 10 and seven MAs64, 65, 66, 67, 68, 69, 70; for obsessive‐compulsive disorder, one NMA ⁷¹ and six MAs72, 73, 74, 75, 76, 77; for anxiety disorders, two NMAs11, 78 and five MAs12, 79, 80, 81, 82 (plus two MAs specific on social anxiety disorder83, 84); for enuresis, one NMA ⁸⁵ and six MAs86, 87, 88, 89, 90, 91, for disruptive behavior/dissocial/conduct disorders, five MAs92, 93, 94, 95, 96 (plus one focusing on youth with comorbid ADHD) ²⁵ ; for eating disorders, one NMA ⁹⁷ and four MAs98, 99, 100, 101; for schizophrenia spectrum disorders, three NMAs8, 102, 103 and two MAs104, 105; for bipolar disorder, four MAs106, 107, 108, 109; for tic disorder, two MAs110, 111; for Tourette's disorder, two MAs112, 113; for encopresis, two MAs114, 115; for developmental coordination disorder, one MA ¹¹⁶ ; and for PTSD, one MA ¹¹⁷ .

Overall, 85.4% of a priori selected medications were covered for at least one of the two co‐primary outcomes, which was the case for 55% of the psychosocial interventions, and 25% of the brain stimulation interventions. Moreover, 70% of a priori selected outcomes were covered across monotherapy medication treatments (anti‐ADHD medications: 65%; antidepressants: 55%; antipsychotics: 40%; mood stabilizers: 25%), 80% across psychosocial interventions, and 20% across brain stimulation interventions.

Among monotherapy medication treatments with data on co‐primary outcomes, those most covered by the literature were atomoxetine (11 outcomes), methylphenidate (9 outcomes), amphetamines and risperidone (8 outcomes), aripiprazole, fluoxetine, guanfacine, lurasidone and quetiapine (7 outcomes), and asenapine, clonidine, olanzapine, paliperidone and sertraline (6 outcomes). Monotherapy psychosocial interventions most covered by the literature were CBT (12 outcomes), behavioral therapy (9 outcomes), parent‐child interaction therapy (7 outcomes), and CBT‐oriented, psychodynamic‐oriented and family‐based therapies (6 outcomes). Among brain stimulation interventions, neurofeedback was the only modality with data that could be included in this umbrella review (4 outcomes).

Quality of included evidence

Among 14 NMAs of RCTs, the median AMSTAR score was 9.5 (interquartile range, IQR: 7‐11), and the median AMSTAR‐Content score was 4 (IQR: 2.75‐5). The median overall quality score across all effect sizes was low in six NMAs (42.9%), moderate in six (42.9%), high in the remaining two (14.2%).

Among 90 MAs of RCTs, the median AMSTAR score was 9 (IQR: 7‐10) and the median AMSTAR‐Content score was 2 (IQR: 1‐3). The median overall quality score across all effect sizes was low in 71 MAs (78.9%), moderate in 19 (21.1%), and high in none.

Across NMAs and MAs of RCTs of medications, the median AMSTAR quality score was 10 (IQR: 7‐11), being low in 0.8%, moderate in 24.7%, and high in 74.4% of the NMAs/MAs, while the AMSTAR‐Content median quality score was 4 (IQR: 3‐5), being low in 30.1%, moderate in 58.6%, and high in 11.3%.

Across NMAs and MAs of RCTs of psychosocial interventions, the median AMSTAR quality score was 11 (IQR: 10‐12), being low in none of the NMAs/MAs, moderate in 8.2%, and high in 91.8%, while the median AMSTAR‐Content quality score was 2 (IQR: 1‐3), being low in 87.4%, moderate in 12.6%, and high in none.

Across brain stimulation interventions, the median AMSTAR quality score was 9 (IQR: 8‐10), being low in none of the NMAs/MAs, medium in 16.7%, and high in 83.3%, while the median AMSTAR‐Content quality score was 2 (IQR: 1‐4), being low in 66.7%, moderate in 33.3%, and high in none.

Efficacy, acceptability and tolerability of pharmacological, psychosocial, and brain stimulation interventions (Tables 2‐7)

ADHD

Results for ADHD are shown in Tables 2, 6 and 7. Amphetamines, methylphenidate, desipramine and modafinil had the largest effect size for the primary efficacy outcome.

Table 2.

Efficacy and effectiveness of pharmacological, psychosocial and brain stimulation interventions vs. inactive control in children/adolescents with neurodevelopmental and disruptive behavior/dissocial/conduct disorders

Outcome	Intervention	Effect size (95% CI)	Control	Number of RCTs/patients	Q
Attention‐deficit/hyperactivity disorder (ADHD)
Pharmacological interventions
Efficacy (clinician‐rated)	Amphetamines	SMD=–1.02 (–1.19 to –0.85)	PBO/Sham	46/9,926	H
	Methylphenidate	SMD=–0.78 (–0.93 to –0.62)	PBO/Sham	46/9,926	H
	Clonidine	SMD=–0.71 (–1.17 to –0.24)	PBO/Sham	46/9,926	H
	Guanfacine	SMD=–0.67 (–0.85 to –0.50)	PBO/Sham	46/9,926	H
	Modafinil	SMD=–0.62 (–0.84 to –0.41)	PBO/Sham	46/9,926	H
	Atomoxetine	SMD=–0.56 (–0.66 to –0.45)	PBO/Sham	46/9,926	H
Efficacy (teacher‐rated)	Desipramine	SMD=–0.97 (–1.66 to –0.28)	PBO/Sham	2/89	L
	Methylphenidate	SMD=–0.82 (–1.16 to –0.48)	PBO/Sham	16/1,843	H
	Modafinil	SMD=–0.76 (–1.15 to –0.37)	PBO/Sham	16/1,843	H
	Amphetamines	SMD=–0.55 (–0.83 to –0.27)	PBO/Sham	5/745	M
	Guanfacine	SMD=–0.63 (–1.62 to 0.35)	PBO/Sham	16/1,843	H
	Atomoxetine	SMD=–0.32 (–0.82 to 0.18)	PBO/Sham	16/1,843	H
Efficacy (parent‐rated)	Desipramine	SMD=–1.42 (–1.99 to –0.85)	PBO/Sham	2/99	L
	Amphetamines	SMD=–1.07 (–1.36 to –0.79)	PBO/Sham	23/3,796	H
	Methylphenidate	SMD=–0.84 (–0.95 to –0.72)	PBO/Sham	23/3,796	H
	Atomoxetine	SMD=–0.60 (–0.71 to –0.50)	PBO/Sham	23/3,796	H
	Modafinil	SMD=–0.46 (–0.61 to –0.31)	PBO/Sham	23/3,796	H
	Bupropion	SMD=–0.32 (–0.69 to 0.05)	PBO/Sham	2/124	L
	Guanfacine	SMD=–0.23 (–0.90 to 0.45)	PBO/Sham	23/3,796	H
Efficacy (mixed‐rated)	Atomoxetine	SMD=–0.17 (–0.23 to –0.11)	PBO/Sham	36/7,579	M
	Amphetamines	SMD=–0.18 (–0.28 to –0.09)	PBO/Sham	36/7,579	M
	Methylphenidate	SMD=–0.14 (–0.21 to –0.08)	PBO/Sham	36/7,579	M
	Guanfacine	SMD=–0.16 (–0.26 to –0.05)	PBO/Sham	36/7,579	M
	Clonidine	SMD=–0.10 (–0.23 to 0.03)	PBO/Sham	36/7,579	M
Response	Desipramine	OR=36.76 (9.17‐214)	PBO/Sham	113/19,398	M
	Amphetamines	OR=7.45 (5.1‐11.09)	PBO/Sham	113/19,398	M
	Modafinil	OR=5.51 (3.04‐10.32)	PBO/Sham	113/19,398	M
	Methylphenidate	OR=5.26 (4.09‐6.82)	PBO/Sham	113/19,398	M
	Clonidine	OR=3.96 (1.89‐8.41)	PBO/Sham	113/19,398	M
	Atomoxetine	OR=3.63 (2.81‐4.73)	PBO/Sham	113/19,398	M
	Guanfacine	OR=3.29 (2.27‐4.82)	PBO/Sham	113/19,398	M
Aggressive behavior	Amphetamines	SMD=–1.15 (–1.38 to –0.93)	PBO/Sham	3/84	L
	Methylphenidate	SMD=–0.26 (–1.10 to 0.68)	PBO/Sham	2/181	L
	Atomoxetine	RR=1.34 (0.91 to 1.97)	PBO/Sham	15/2,067	M
Cognition: executive memory	Methylphenidate	SMD=–0.26 (–0.39 to –0.13)	PBO/Sham	7/468	L
Cognition: non‐executive memory	Methylphenidate	SMD=–0.60 (–0.79 to –0.41)	PBO/Sham	8/635	L
Cognition: reaction time	Methylphenidate	SMD=–0.21 (–0.30 to –0.12)	PBO/Sham	21/1,095	L
Cognition: response inhibition	Methylphenidate	SMD=–0.41 (–0.55 to –0.27)	PBO/Sham	16/846	L
Acceptability	Clonidine	OR=0.40 (0.20‐0.78)	PBO/Sham	171/22,961	M
	Methylphenidate	OR=0.59 (0.46‐0.75)	PBO/Sham	171/22,961	M
	Aripiprazole	OR=0.61 (0.02‐25.34)	PBO/Sham	171/22,961	M
	Modafinil	OR=0.67 (0.37‐1.24)	PBO/Sham	171/22,961	M
	Desipramine	OR=0.70 (0.17‐2.89)	PBO/Sham	171/22,961	M
	Amphetamines	OR=0.78 (0.52‐1.18)	PBO/Sham	171/22,961	M
	Guanfacine	OR=0.79 (0.54‐1.14)	PBO/Sham	171/22,961	M
	Atomoxetine	OR=0.85 (0.68‐1.07)	PBO/Sham	171/22,961	M
	Bupropion	OR=1.54 (0.39‐6.76)	PBO/Sham	171/22,961	M
Tolerability	Methylphenidate	OR=1.31 (0.79‐2.25)	PBO/Sham	60/12,188	M
	Modafinil	OR=1.34 (0.57‐3.18)	PBO/Sham	60/12,188	M
	Amphetamines	OR=1.38 (0.64‐3.00)	PBO/Sham	60/12,188	M
	Clonidine	OR=2.32 (0.63‐8.94)	PBO/Sham	58/NR	H
	Bupropion	OR=3.60 (0.34‐130)	PBO/Sham	60/12,188	M
	Atomoxetine	OR=1.48 (1.01‐2.18)	PBO/Sham	60/12,188	M
	Guanfacine	OR=3.39 (1.93‐6.3)	PBO/Sham	60/12,188	M
Discontinuation due to inefficacy	Amphetamine	OR=0.11 (0.05‐0.20)	PBO/Sham	45/9,087	M
	Clonidine	OR=0.29 (0.13‐0.56)	PBO/Sham	45/9,087	M
	Methylphenidate	OR=0.31 (0.18‐0.53)	PBO/Sham	45/9,087	M
	Guanfacine	OR=0.37 (0.26‐0.54)	PBO/Sham	45/9,087	M
	Atomoxetine	OR=0.47 (0.33‐0.67)	PBO/Sham	45/9,087	M
	Bupropion	OR=1.97 (0.19‐57.4)	PBO/Sham	45/9,087	M
Functioning	Atomoxetine	SMD=–0.48 (–0.62 to –0.33)	PBO/Sham	8/1,308	M
Functioning: academic	Amphetamines	SMD=–0.56 (–0.73 to –0.39)	PBO/Sham	8/826	M
Global illness improvement	Amphetamines	OR=7.71 (5.52‐10.77)	PBO/Sham	40/NR	H
	Atomoxetine	OR=2.28 (1.38‐3.76)	PBO/Sham	40/NR	H
	Guanfacine	OR=3.63 (2.36‐5.57)	PBO/Sham	40/NR	H
	Methylphenidate	OR=5.57 (3.99‐7.79)	PBO/Sham	40/NR	H
	Modafinil	OR=3.22 (1.91‐5.43)	PBO/Sham	40/NR	H
	Clonidine	OR=2.78 (0.91‐8.53)	PBO/Sham	40/NR	H
Global illness severity	Amphetamines	SMD=–0.86 (–1.72 to –0.01)	PBO/Sham	2/86	M
	Desipramine	OR=26.41 (7.41‐94.18)	PBO/Sham	2/103	L
Quality of life	Methylphenidate	SMD=–0.61 (–0.80 to –0.42)	PBO/Sham	3/514	M
	Atomoxetine	SMD=–0.39 (–0.50 to –0.28)	PBO/Sham	16/2,361	M
Suicide attempt	Atomoxetine	RR=0.84 (0.03‐20.00)	PBO/Sham	23/3,883	L
Suicidal ideation	Atomoxetine	RR=1.67 (0.83‐3.36)	PBO/Sham	15/2,517	M
Pharmacological augmentation
Efficacy	α2‐agonists + stimulants	SMD=–0.36 (–0.51 to –0.21)	PBO/Sham	3/719	M
Acceptability	α2‐agonists + stimulants	RR=0.74 (0.37‐1.48)	PBO/Sham	3/726	L
Tolerability	α2‐agonists + stimulants	RR=0.77 (0.05‐12.50)	PBO/Sham	3/726	L
Discontinuation due to inefficacy	α2‐agonists + stimulants	RR=0.49 (0.21‐1.13)	PBO/Sham	3/726	M
Psychosocial interventions
Efficacy (mixed‐rated)	Social skills training	SMD=–0.39 (–0.63 to –0.15)	WL/NT	15/2,857	L
Efficacy (teacher‐rated)	Social skills training	SMD=–0.26 (–0.47 to –0.05)	WL/NT	14/1,379	M
Efficacy (parent‐rated)	Social skills training	SMD=–0.54 (–0.81 to –0.26)	WL/NT	11/1,206	L
Efficacy (clinician‐rated)	Social skills training	SMD=–3.15 (–9.88 to 3.57)	WL/NT	2/107	L
Response	Behavioral therapy	OR=2.97 (1.53‐5.88)	PBO/Sham	113/19,398	M
	Cognitive training	OR=0.70 (0.12‐3.87)	PBO/Sham	113/19,398	M
Acceptability	Behavioral therapy	OR=0.58 (0.33‐0.99)	PBO/Sham	171/22,961	M
	Cognitive training	OR=1.32 (0.71‐2.52)	PBO/Sham	171/22,961	M
Functioning: academic	Social skills training	SMD=–0.15 (–0.31 to 0.01)	WL/NT	5/642	M
Global illness severity	Behavioral therapy	OR=2.99 (1.21‐7.31)	PBO/Sham	113/19,398	M
	Cognitive training	OR=0.39 (0.01‐5.80)	PBO/Sham	113/19,398	M
Functioning: social skills (mixed‐rated)	Social skills training	SMD=–0.29 (–0.47 to –0.11)	WL/NT	19/2,649	L
Functioning: social skills (parent‐rated)	Social skills training + parental involvement	SMD=–0.43 (–0.70 to –0.15)	WL/NT	4/337	L
	Social skills training	SMD=–0.19 (–0.32 to –0.06)	WL/NT	15/1,609	M
Functioning: social skills (teacher‐rated)	Social skills training + parental involvement	SMD=–0.15 (–0.41 to 0.12)	WL/NT	4/632	M
	Social skills training	SMD=–0.11 (–0.22 to 0.00)	WL/NT	11/1,271	M
Functioning: emotional (mixed‐rated)	Social skills training	SMD=0.20 (–0.01 to 0.41)	WL/NT	5/353	L
Functioning: emotional (parent‐rated)	Social skills training	SMD=0.27 (–0.05 to 0.59)	WL/NT	3/173	L
Functioning: emotional (teacher‐rated)	Social skills training	SMD=0.02 (–0.68 to 0.72)	WL/NT	2/129	L
Brain stimulation interventions
Response	Neurofeedback	OR=1.96 (0.52‐8.26)	PBO/Sham	113/19,398	M
Acceptability	Neurofeedback	OR=0.59 (0.31‐1.14)	PBO/Sham	171/22,961	M
Combined interventions
Response	Methylphenidate + parent training	OR=55.63 (3.18‐29.52x10 2 )	PBO/Sham	113/19,398	M
	Methylphenidate + clonidine	OR=21.91 (5.52‐105.40)	PBO/Sham	113/19,398	M
	Atomoxetine + parent training	OR=2.48 (0.51‐11.79)	PBO/Sham	113/19,398	M
Acceptability	Methylphenidate + clonidine	OR=0.32 (0.13‐0.77)	PBO/Sham	171/22,961	M
ADHD and disorders of intellectual development
Efficacy	Methylphenidate	SMD=–0.88 (–1.14 to –0.61)	PBO/Sham	8/424	L
Acceptability	Methylphenidate	OR=1.68 (0.68‐4.14)	PBO/Sham	4/215	L
Tolerability	Methylphenidate	OR=4.82 (0.98‐23.63)	PBO/Sham	4/215	L
Autism spectrum disorder
Pharmacological interventions
Efficacy: inappropriate speech (mixed‐rated)	Aripiprazole	SMD=–0.30 (–0.50 to –0.09)	PBO/Sham	3/400	L
Efficacy: stereotypic (mixed‐rated)	Aripiprazole	SMD=–0.32 (–0.53 to–0.12)	PBO/Sham	3/400	M
	Methylphenidate	SMD=–0.18 (–0.46 to 0.11)	PBO/Sham	5/127	M
	Atomoxetine	SMD=–0.16 (–0.50 to 0.18)	PBO/Sham	4/281	L
Efficacy: overall (teacher‐rated)	Methylphenidate	SMD=–0.53 (–1.26 to 0.19)	PBO/Sham	2/37	L
Efficacy: social interaction (parent‐rated)	Methylphenidate	SMD=–0.21 (–0.6 to 0.18)	PBO/Sham	2/90	L
Efficacy: social interaction (teacher‐rated)	Methylphenidate	SMD=–0.51 (–1.07 to 0.05)	PBO/Sham	3/103	L
Efficacy: stereotypic (parent‐rated)	Methylphenidate	SMD=–0.34 (–0.84 to 0.17)	PBO/Sham	3/NR	L
Efficacy: social withdrawal (mixed‐rated)	Aripiprazole	SMD=–0.13 (–0.33 to 0.08)	PBO/Sham	3/400	M
Response	Risperidone	OR=2.57 (1.35‐4.86)	PBO/Sham	3/241	L
	Aripiprazole	RR=2.08 (1.24‐3.46)	PBO/Sham	3/400	L
Aggressive behavior	Risperidone	SMD=–0.29 (–0.48 to –0.11)	PBO/Sham	8/878	L
	Aripiprazole	SMD=–0.24 (–0.40 to –0.08)	PBO/Sham	8/878	L
	Valproate	SMD=–0.18 (–0.71 to 0.35)	PBO/Sham	2/57	M
	Lurasidone	SMD=–0.05 (–0.27 to 0.18)	PBO/Sham	8/878	L
Acceptability	Risperidone	RR=0.52 (0.32‐0.86)	PBO/sham	6/379	M
	Antipsychotics	RR=0.61 (0.48‐0.78)	PBO/Sham	15/1,124	M
	Aripiprazole	RR=0.67 (0.49‐0.90)	PBO/Sham	5/526	M
	Haloperidol	RR=0.80 (0.24‐2.62)	PBO/Sham	2/60	M
	Mood stabilizers	RR=1.27 (0.53‐3.06)	PBO/Sham	5/125	M
Tolerability	Risperidone	RR=0.71 (0.17‐2.92)	PBO/Sham	5/339	M
	Antipsychotics	RR=0.99 (0.55‐1.79)	PBO/Sham	12/1,010	M
	Mood stabilizers	RR=1.13 (0.36‐3.53)	PBO/Sham	4/112	M
	Aripiprazole	RR=1.24 (0.57‐2.71)	PBO/Sham	4/493	M
Discontinuation due to inefficacy	Mood stabilizers	RR=2.11 (0.36‐12.42)	PBO/Sham	3/60	M
Global illness severity	Aripiprazole	SMD=–0.54 (–0.77 to –0.32)	PBO/Sham	3/400	M
	Risperidone	OR=10.5 (4.80‐22.60)	PBO/Sham	6/446	L
	Mood stabilizers	RR=1.55 (0.39‐6.21)	PBO/Sham	3/77	L
Relapse	Risperidone	RR=0.30 (0.13‐0.68)	PBO/Sham	2/56	M
Psychosocial interventions
Efficacy: emotion recognition (mixed‐rated)	Computer‐assisted interaction	SMD=–0.53 (–1.12 to 0.05)	WL/NT	2/48	L
	Social skills training	SMD=–0.34 (–0.88 to 0.20)	WL/NT	2/54	L
Efficacy: social competence (mixed‐rated)	Social skills training	SMD=–0.47 (–0.78 to –0.16)	WL/NT	4/178	L
Anxiety (subject‐rated)	Cognitive behavioral therapy	SMD=–0.61 (–1.54 to 0.33)	WL/NT	5/181	L
Anxiety (parent‐rated)	Cognitive behavioral therapy	SMD=–1.12 (–1.91 to –0.34)	WL/NT	7/244	L
Functioning: joint attention	Skills training‐joint attention	SMD=–0.66 (–0.93 to –0.40)	WL/NT	9/417	L
Disruptive behavior/dissocial/conduct disorders (with or without ADHD)
Pharmacological interventions
Efficacy (clinician‐rated)	Risperidone	SMD=–0.48 (–0.71 to –0.24)	PBO/Sham	4/293	L
Efficacy (parent‐rated)	Risperidone	SMD=–0.79 (–1.06 to –0.52)	PBO/Sham	2/225	M
Efficacy (mixed‐rated)	Risperidone	SMD=–0.32 (–0.49 to –0.16)	PBO/Sham	4/590	M
Response: aggressive behavior	Valproate	OR=15.6 (1.91‐128.1)	PBO/Sham	2/47	L
	Lithium	RR=4.56 (1.97‐10.56)	PBO/Sham	3/116	L
Aggressive behavior (clinician‐rated)	Mixed (risperidone, quetiapine)	SMD=–0.24 (–0.76 to 0.29)	PBO/Sham	2/57	L
Aggressive behavior (parent‐rated)	Risperidone	SMD=–0.72 (–0.99 to –0.46)	PBO/Sham	3/238	M
Aggressive behavior (mixed‐rated)	Risperidone	SMD=–0.60 (–0.89 to –0.31)	PBO/Sham	2/188	L
	Mixed (risperidone, lithium, methylphenidate)	SMD=–1.93 (–3.88 to 0.02)	PBO/Sham	4/172	L
Acceptability	Mixed (risperidone, lithium, methylphenidate)	RR= 0.97 (0.60‐1.55)	PBO/Sham	8/631	L
Global illness severity	Risperidone	SMD=–1.31 ( –1.88 to –0.74)	PBO/Sham	2/58	L
	Mixed (risperidone, quetiapine)	SMD=–0.30 (–0.49 to –0.12)	PBO/Sham	5/435	M
	Mixed (carbamazepine, lithium, amphetamines)	RR= 2.39 (1.10‐5.21)	PBO/Sham	4/136	L
Psychosocial interventions
Efficacy (parent‐rated)	Parental + child behavioral interventions	SMD=–1.00 (–1.68 to –0.32)	WL/NT	3/207	L
Intellectual disabilities and disruptive behavior/dissocial disorders (with or without ADHD)
Aggressive behavior (clinician‐rated)	Risperidone	SMD=–1.09 (–1.39 to –0.79)	PBO/Sham	4/257	L
	Aripiprazole	SMD=–0.64 (–0.91 to –0.36)	PBO/Sham	2/308	L
	Valproate	SMD=–0.06 (–0.75 to 0.63)	PBO/Sham	2/57	L
Aggressive behavior (mixed‐rated)	Risperidone	SMD=–0.70 (–1.01 to –0.39)	PBO/Sham	3/266	L
Developmental coordination disorders
Efficacy	Skills training	SMD=–0.27 (–0.85 to 0.31)	WL/NT	2/51	L
Tic disorder
Efficacy: tics (clinician‐rated)	Desipramine	SMD=–0.44 (–0.91 to 0.02)	PBO/Sham	2/75	L
	Methylphenidate	SMD=–0.28 (–0.58 to 0.03)	PBO/Sham	4/191	L
Tourette's disorder
Efficacy (clinician‐rated)	Antipsychotics (haloperidol, pimozide, risperidone, ­ziprasidone)	SMD=–0.74 (–1.08 to –0.41)	PBO/Sham	4/75	L
	Guanfacine	SMD=–0.73 (–1.26 to –0.20)	PBO/Sham	2/58	L
	Methylphenidate	SMD=–0.17 (–0.46 to 0.11)	PBO/Sham	4/161	L

RCTs – randomized controlled trials, SMD – standardized mean difference, OR – odds ratio, RR – risk ratio, PBO – placebo, WL – waiting list, NT – no treatment, NR – not reported, Q – quality (H – high, M – medium, L – low). Bold prints indicate significant values. SMDs<0 indicate that intervention is more effective than control. For discontinuation outcomes (acceptability, tolerability, inefficacy) and relapse, OR/RR<1 favors the intervention. For response and remission, OR/RR>1 favors the intervention.

Table 6.

Efficacy and effectiveness of pharmacological, psychosocial and brain stimulation interventions vs. active psychological intervention or drug condition in children/adolescents (only significant differences are reported)

Outcome	Intervention	Effect size (95% CI)	Control	Number of RCTs/patients	Q
Anorexia nervosa
Efficacy: weight gain	FT	SMD=–0.44 (–0.74 to –0.14)	Other than FT	4/178	L
Anxiety disorders
Efficacy (mixed‐rated)	CBT‐Group	SMD=–0.44 (–0.82 to –0.06)	CBT‐Individual	101/6,625	L
Attention‐deficit/hyperactivity disorder (ADHD)
Efficacy (clinician‐rated)	Amphetamines	SMD=–0.24 (–0.44 to –0.05)	Methylphenidate	46/NR	H
	Methylphenidate	SMD=–0.22 (–0.39 to –0.05)	Atomoxetine	46/NR	H
Efficacy (parent‐rated)	Methylphenidate	SMD=–1.07 (–1.74 to –0.40)	Bupropion	23/NR	H
	Methylphenidate	SMD=–0.23 (–0.37 to –0.10)	Atomoxetine	23/NR	H
Response	Methylphenidate	OR=1.44 (1.08‐1.92)	Atomoxetine	113/19,398	M
Aggressive behavior	Amphetamines	SMD=–0.35 (–0.56 to –0.13)	Methylphenidate	2/132	L
Acceptability	Methylphenidate	OR=0.68 (0.52‐0.91)	Atomoxetine	171/22,961	M
Tolerability	Methylphenidate	OR=0.39 (0.18‐0.83)	Guanfacine	60/12,188	M
Discontinuation due to inefficacy	Amphetamines	OR=0.23 (0.10‐0.44)	Atomoxetine	45/9,087	M
Global illness severity	Amphetamines	OR=3.39 (1.95‐5.88)	Atomoxetine	40/NR	H
Efficacy: inattention (mixed‐rated)	Neurofeedback	SMD=0.44 (0.02 to 0.86)	Stimulants	4/161	L
Acceptability	Neurofeedback	OR=0.45 (0.21‐0.95)	COG TR	171/22,961	M
Response	BT+stimulants	OR=4.76 (2.50‐9.09)	BT	113/19,398	M
	BT+stimulants	OR=4.58 (2.49‐8.75)	Stimulants	113/19,398	M
Autism spectrum disorder
Efficacy: stereotypic (clinician‐rated)	BT‐IT	SMD=–0.78 (–1.42 to –0.13)	BT‐CI	2/40	L
Efficacy: distal social behavior (clinician‐rated)	BT‐IT	SMD=–0.98 (–1.64 to –0.32)	BT‐CI	2/40	L
Bipolar disorder, manic episode
Efficacy (clinician‐rated)	Risperidone	SMD=–1.01 (–1.29 to –0.74)	Valproate	2/228	M
Enuresis
Acceptability	Desmopressin	OR=0.45 (0.29‐0.71)	BT‐Alarm	15/1,502	M
Efficacy	BT‐Alarm	SMD= –0.43 (–0.77 to –0.08)	Desmopressin	4/285	L
Relapse	BT‐Alarm	OR=0.15 (0.03‐0.53)	Desmopressin	12/1,381	M
Efficacy	Desmopressin+ BT‐Alarm	SMD= –0.58 (–0.89 to –0.26)	Desmopressin	2/156	L
Response	Desmopressin+anticholinergics	OR=2.80 (1.50‐5.40)	Desmopressin	15/1,350	M
	Imipramine+oxybutynin	RR=1.47 (1.09‐2.00)	Imipramine	2/101	L
	Imipramine+oxybutynin	RR=1.46 (1.06‐2.01)	Oxybutynin	2/100	L
	Desmopressin+BT‐Alarm	RR=1.32 (1.08‐1.62)	Desmopressin	5/359	L
Relapse	Oxybutynin+ imipramine	RR=0.50 (0.30‐0.81)	Oxybutynin	2/81	L
	Oxybutynin+ imipramine	RR=0.48 (0.31‐0.74)	Imipramine	2/85	L
Depressive disorders
Efficacy (clinician‐rated)	Fluoxetine	SMD=–1.65 (–2.34 to –0.95)	Nortriptyline	70/8,906	M
Response	Fluoxetine	OR=3.02 (1.04‐7.22)	Nortriptyline	34/5,260	M
Tolerability	Paroxetine	OR=0.22 (0.08‐0.87)	Imipramine	34/5,260	M
	Fluoxetine	OR=0.31 (0.13‐0.95)	Duloxetine	34/5,260	M
Suicidal ideation	CBT	SMD=–0.27 (–0.51 to –0.03)	SSRIs	2/268	L
Remission	CBT+SSRI	OR=2.15 (1.15‐4.02)	CBT+PBO	2/173	M
Functioning	CBT+SSRI	SMD=–0.20 (–0.33 to –0.08)	Standalone AD	4/850	L
Schizophrenia spectrum disorders
Efficacy (clinician‐rated)	Haloperidol	SMD=–1.35 (–2.16 to –0.55)	Fluphenazine	28/3,003	L
	Clozapine	SMD=–0.86 (–1.54 to –0.17)	Olanzapine	28/3,003	L
	SGAs	SMD=–0.36 (–0.56 to –0.16)	FGAs	4/243	L
Response	Risperidone	OR=5.53 (2.01‐15.18)	Haloperidol	28/3,003	L
Tic disorder
Response	Topiramate	RR=1.10 (1.02‐1.18)	Haloperidol/tiapride	14/1,017	M
	Topiramate	RR=1.09 (1.01‐1.19)	Haloperidol	10/727	L

RCTs – randomized controlled trials, SMD – standardized mean difference, OR – odds ratio, RR – risk ratio, PBO – placebo, Q – quality (H – high, M – medium, L – low), BT – behavioral therapy, BT‐IT– behavioral therapy imitative interaction, BT‐CI – behavioral therapy contingency interaction, CBT – cognitive behavioral therapy, FT – family therapy, COG TR ‐ cognitive training, AD – antidepressant, SSRI – selective serotonin reuptake inhibitor, SGAs – second‐generation antipsychotics, FGAs – first‐­generation antipsychotics, NR – not reported. SMDs<0 indicate that intervention is more effective than control. For discontinuation outcomes (acceptability, tolerability, inefficacy) and relapse, OR/RR<1 favors the intervention. For response and remission, OR/RR>1 favors the intervention.

Table 7.

Efficacy and effectiveness of pharmacological, psychosocial and brain stimulation interventions vs. mixed control conditions in ­children/adolescents (only significant differences are reported)

Outcome	Intervention	Effect size (95% CI)	Control	Number of RCTs/patients	Q
Attention‐deficit/hyperactivity disorder (ADHD)
Efficacy (mixed‐rated)	BI	SMD=–0.55 (–0.77 to –0.32)	WL/AC/LIP	6/333	L
Efficacy (probably blinded rater)	COG TR	SMD=–0.20 (–0.40 to –0.01)	Mixed	11/566	L
Efficacy (most proximal rater)	COG TR	SMD=–0.37 (–0.66 to –0.09)	Mixed	14/727	L
	BT	SMD=–0.35 (–0.50 to –0.19)	Mixed	19/1,430	L
Efficacy (teacher‐rated)	ST	SMD=–0.26 (–0.52 to –0.01)	Mixed	6/615	L
Efficacy (parent‐rated)	BT‐Parental	SMD=–0.65 (–1.05 to –0.25)	TAU/WL/LIP	8/399	L
	ST	SMD=–0.56 (–0.74 to –0.38)	Mixed	10/934	L
Aggressive behavior	BI	SMD=–0.40 (–0.71 to –0.10)	Mixed	5/350	L
Functioning: academic	ST	SMD=–0.33 (–0.51 to –0.14)	Mixed	7/695	L
	BT	SMD=–0.28 (–0.59 to –0.06)	Mixed	9/817	L
Efficacy (most proximal rater)	Neurofeedback	SMD=–0.35 (–0.59 to –0.11)	Mixed	13/540	M
Efficacy (parent‐rated)	Neurofeedback	SMD=–0.32 (p=0.013)	Mixed	16/706	L
Autism spectrum disorder
Efficacy: socialization (mixed‐rated)	PCIT	SMD=–0.22 (–0.36 to –0.09)	Mixed	13/846	L
Efficacy: language (mixed‐rated)	PCIT	SMD=–0.16 (–0.31 to –0.02)	Mixed	13/785	L
Efficacy: language comprehension (parent‐rated)	PCIT	SMD=–0.29 (–0.56 to –0.01)	Mixed	3/204	L
Anxiety (clinician‐rated)	CBT	SMD=–1.05 (–1.65 to –0.45)	TAU/WL	6/208	L
Anxiety (parent‐rated)	CBT	SMD=–1.00 (–1.80 to –0.21)	TAU/WL	7/283	L
Aggressive behavior	PCIT	SMD = –0.67 (–0.85 to –0.49)	Mixed	9/521	L
Functioning: shared/joint attention	ST‐ToM	SMD=–0.55 (–0.99 to –0.11)	TAU/WL	2/88	L
	PCIT	SMD=–0.41 (–0.68 to –0.14)	Mixed	3/215	L
Functioning: social skills	SST‐Computer	SMD=–0.93 (–1.29 to –0.57)	TAU/WL	5/138	L
	SST	SMD=–0.83 (–1.07 to –0.60)	TAU/WL	18/1,266	L
	SST‐Face to face	SMD=–0.81 (–1.08 to –0.53)	TAU/WL	14/1,128	L
Functioning: parent synchrony	PCIT	SMD=–0.90 (–1.23 to –0.56)	Mixed	3/244	L
Global illness severity	PCIT	SMD=–0.30 (–0.52 to –0.08)	Mixed	6/316	L
Irritability	PCIT	SMD=–0.59 (–0.88 to –0.30)	Mixed	8/653	L
Depressive disorders
Efficacy (mixed‐ rated)	CBT	SMD=–0.53 (–0.82 to –0.24)	Mixed	11/809	M
Oppositional defiant disorder (ODD)
Efficacy (mixed‐rated)	BI	SMD=–0.79 (–0.93 to –0.64)	WL/AC	17/NR	L
Tourette's disorder
Efficacy (clinician‐rated)	BT	SMD=–0.64 (–0.99 to –0.29)	WL/LIP	2/133	L
Disruptive behavior/dissocial/conduct disorders (with or without ADHD)
Efficacy: ADHD symptoms (mixed‐ rated)	BI	SMD=–0.34 (–0.64 to –0.05)	WL/AC	11/518	L
Efficacy: ADHD symptoms (parent‐rated)	BI	SMD=–0.68 (–0.91 to –0.44)	WL/AC	5/322	L
Efficacy: externalizing (mixed‐rated)	BI	SMD=–0.52 (–0.68 to –0.36)	WL/AC	10/881	L
Efficacy: ODD symptoms (mixed‐ rated)	BI	SMD=–0.88 (–1.24 to –0.51)	WL/AC	10/335	L
Efficacy: ODD symptoms (parent‐rated)	BI	SMD=–0.81 (–1.20 to –0.42)	WL/AC	4/199	L
Aggressive behavior	BI	SMD =–0.28 (–0.46 to –0.10)	WL/AC	18/794	L
Cognition: attention	BI	SMD=–0.38 (–0.52 to –0.23)	WL/AC	15/588	L
Functioning	BI	SMD=–0.39 (–0.52 to –0.26)	WL/AC	22/1,027	L

RCTs – randomized controlled trials, SMD – standardized mean difference, WL – waiting list, AC – active control, TAU – treatment as usual, LIP – low intensity psychosocial intervention, Q – ­quality (H – high, M – medium, L – low), BT – behavioral therapy, CBT – cognitive behavioral therapy, COG TR – cognitive training, BI – combination of parental and child behavioral interventions, ST – skills training, PCIT – parent‐child interaction therapy, SST – social skills training, ST‐ToM – skills training: precursors of Theory of Mind, NR – not reported. SMDs<0 indicate that intervention is more effective than control.

Focusing on the two best interventions, amphetamines had the highest effect size based on the clinician‐rated primary efficacy outcome vs. placebo (large effect size), and were superior to placebo also regarding response (large effect size), aggressive behavior (large effect size), academic functioning (medium effect size), global illness severity (large effect size), and less discontinuation due to inefficacy (large effect size), without significant differences regarding all‐cause discontinuation (“acceptability”) or discontinuation due to intolerability (see Table 2).

Methylphenidate had medium to large effect sizes regarding the primary efficacy outcome vs. placebo across different raters, and was superior to placebo regarding other‐than‐attention cognition broadly (small to medium effect size), global illness improvement (large effect size), quality of life (medium effect size), acceptability (small effect size), and less discontinuation due to inefficacy (medium effect size), without significant differences concerning discontinuation due to intolerability. The efficacy of methylphenidate was also confirmed in youth with comorbid intellectual disability (see Table 2).

Clonidine, guanfacine and atomoxetine were also effective regarding the primary efficacy outcome, but with less consistent results across raters. Among psychosocial interventions, social skills training improved the primary efficacy outcome and functioning (small to medium effect size); however, the control group was waiting list/no treatment. Only behavioral therapy outperformed placebo for response (small effect size), impact on global illness severity (small effect size), and acceptability (small effect size). Neurofeedback did not show any significant efficacy outcome, nor any difference emerged on acceptability (see Table 2).

Alpha‐2 agonists were an effective augmentation strategy when added to stimulants vs. placebo (small effect size). Importantly, combined interventions, and specifically methylphenidate with parent training or with clonidine, and atomoxetine with parent training, showed large effect sizes regarding response vs. placebo (see Table 2). Additionally, behavioral therapy plus stimulants was superior both to behavioral therapy alone and to stimulants alone regarding response (large effect size), without any differences in acceptability (see Table 6).

In head‐to‐head comparisons, amphetamines outperformed methylphenidate, which outperformed bupropion (large effect sizes) and atomoxetine (small effect size) on the primary efficacy outcome. Amphetamines were superior to atomoxetine in reducing discontinuation due to inefficacy, and better than methylphenidate for aggressive behavior (small effect size), while methylphenidate was superior to atomoxetine regarding acceptability (medium effect size), and to guanfacine regarding less discontinuation due to intolerability (medium effect size). Stimulants were superior to neurofeedback regarding cognition, and neurofeedback outperformed cognitive training on acceptability (see Table 6).

Autism spectrum disorder

Results for autism spectrum disorder are shown in Tables 2, 5, 6 and 7.

Table 5.

Efficacy and effectiveness of pharmacological, psychosocial and brain stimulation interventions vs. treatment as usual (TAU) or low intensity psychosocial intervention (LIP) in children/adolescents (only significant differences are reported)

Outcome	Intervention	Effect size (95% CI)	Control	Number of RCTs/patients	Q
Anxiety disorders
Efficacy (mixed‐ rated)	CBT‐Group	SMD=–0.84 (–1.47 to –0.21)	TAU	101/6,625	L
Functioning	CBT	SMD=–1.06 (–1.57 to –0.55)	TAU/LIP/PBO/Sham	5/467	L
Remission	CBT‐Individual+P	OR=8.56 (3.10‐23.66)	TAU	5/172	L
Autism spectrum disorder
Efficacy: overall (mixed‐rated)	PCIT	SMD=–0.22 (–0.41 to –0.03)	TAU/LIP	6/420	L
Efficacy: reciprocity (clinician‐rated)	Mixed psychosocial interventions	SMD=–0.53 (–0.78 to –0.29)	TAU	8/380	L
Cognition: developmental quotient	Mixed psychosocial interventions	SMD=–0.36 (–0.66 to –0.05)	TAU	5/232	L
Cognition	PCIT	SMD=–0.24 (–0.46 to –0.03)	TAU/LIP	6/334	L
Anxiety disorder remission	CBT	OR=11.25 (3.11‐40.79)	TAU	4/142	L
Depressive disorders
Efficacy (clinician ‐rated)	IPT	SMD=–0.66 (–1.22 to –0.09)	TAU	70/8,906	L
Encopresis
Efficacy: soiling	BT+TAU	SMD=–0.35 (–0.63 to –0.07)	TAU	4/209	L
Response	BT+TAU	RR=1.78 (1.25‐2.55)	TAU	4/216	L
Obsessive‐compulsive disorder
Response	BT‐ERP	RR=1.71 (1.29‐2.25)	TAU/LIP	4/271	L
Acceptability	BT‐ERP	RR=0.60 (0.39‐0.93)	TAU/LIP	4/251	L

RCTs – randomized controlled trials, SMD – standardized mean difference, OR – odds ratio, RR – risk ratio, PBO – placebo, Q – quality (H – high, M – medium, L – low), BT – behavioral therapy, BT‐ERP – behavioral therapy with exposure and response prevention, CBT – cognitive behavioral therapy, IPT – interpersonal therapy, PCIT – parent‐child ­interaction therapy, P – parental involvement. SMDs<0 indicate that intervention is more effective than control. For discontinuation outcomes (acceptability, ­tolerability, inefficacy) and relapse, OR/RR<1 favors the intervention. For response and remission, OR/RR>1 favors the intervention.

Aripiprazole was superior to placebo regarding the primary efficacy outcome, as well as response, aggressive behavior, global illness severity, and acceptability (all small effect sizes). Risperidone showed the same profile, yet with a large effect size regarding response. Both aripiprazole and risperidone were not different from placebo concerning discontinuation due to intolerability (see Table 2).

Among psychosocial interventions, social skills training had a small to large effect size regarding the primary efficacy outcome and functioning, and CBT had a large effect concerning anxiety across different control groups (see Table 2). Parent‐child interaction therapy and other mixed psychosocial interventions had a small to medium effect size for the primary efficacy outcome vs. TAU, as well as a small effect regarding cognition. Parent‐child interaction therapy also improved aggression (medium effect size), irritability (medium effect size), and functioning (large effect size). Finally, behavioral therapy with an imitative component had a large effect size for the primary efficacy outcome against other active psychosocial interventions without the imitative component (see Tables 5, 6 and 7).

Depressive disorders

Results for depressive disorders are shown in Tables 3, 5, 6 and 7.

Table 3.

Efficacy and effectiveness of pharmacological, psychosocial and brain stimulation interventions vs. inactive control in children/adoles­cents with schizophrenia spectrum, depressive, and bipolar disorders

Outcome	Intervention	Effect size (95% CI)	Control	Number of RCTs/patients	Q
Schizophrenia spectrum disorders
Efficacy (clinician‐rated)	Olanzapine	SMD=–0.74 (–1.05 to –0.44)	PBO/Sham	28/3,003	L
	Risperidone	SMD=–0.62 (–0.89 to –0.34)	PBO/Sham	28/3,003	L
	Lurasidone	SMD=–0.48 (–0.71 to –0.25)	PBO/Sham	28/3,003	M
	Aripiprazole	SMD=–0.43 (–0.63 to –0.24)	PBO/Sham	28/3,003	M
	Quetiapine	SMD=–0.42 (–0.65 to –0.19)	PBO/Sham	28/3,003	M
	Paliperidone	SMD=–0.42 (–0.66 to –0.18)	PBO/Sham	28/3,003	L
	Asenapine	SMD=–0.38 (–0.66 to –0.11)	PBO/Sham	28/3,003	M
	Ziprasidone	SMD=–0.14 (–0.40 to 0.11)	PBO/Sham	28/3,003	L
Response	Risperidone	OR=3.46 (1.92‐6.23)	PBO/Sham	28/3,003	L
	Olanzapine	OR=2.64 (1.07‐4.18)	PBO/Sham	28/3,003	L
	Lurasidone	OR=2.56 (1.45‐4.48)	PBO/Sham	28/3,003	M
	Paliperidone	OR=2.12 (1.07‐4.18)	PBO/Sham	28/3,003	L
	Quetiapine	OR=1.86 (1.03‐3.32)	PBO/Sham	28/3,003	M
	Asenapine	OR=1.73 (0.96‐3.10)	PBO/Sham	28/3,003	M
Global illness severity	Olanzapine	SMD=–0.6 (–1.18 to –0.02)	PBO/Sham	13/2,210	M
	Risperidone	SMD=–0.50 (–0.73 to –0.27)	PBO/Sham	12/2,158	L
	Paliperidone	SMD=–0.44 (–0.67 to –0.22)	PBO/Sham	12/2,158	L
	Lurasidone	SMD=–0.41 (–0.77 to –0.05)	PBO/Sham	13/2,210	M
	Quetiapine	SMD=–0.41 (–0.77 to –0.05)	PBO/Sham	13/2,210	M
	Ziprasidone	SMD=–0.40 (–0.68 to –0.12)	PBO/Sham	13/2,210	M
	Aripiprazole	SMD=–0.35 (–0.59 to –0.11)	PBO/Sham	13/2,210	M
	Asenapine	SMD=–0.29 (–0.53 to –0.06)	PBO/Sham	13/2,210	M
Acceptability	Paliperidone	OR=0.26 (0.08‐0.80)	PBO/Sham	28/3,003	L
	Risperidone	OR=0.31 (0.14‐0.72)	PBO/Sham	28/3,003	L
	Olanzapine	OR=0.36 (0.15‐0.85)	PBO/Sham	28/3,003	L
	Lurasidone	OR=0.53 (0.18‐1.55)	PBO/Sham	28/3,003	M
	Ziprasidone	OR=0.59 (0.22‐1.58)	PBO/Sham	28/3,003	L
	Quetiapine	OR=0.63 (0.27‐1.43)	PBO/Sham	28/3,003	M
	Asenapine	OR=0.91 (0.33‐2.56)	PBO/Sham	28/3,003	M
	Aripiprazole	OR=1.48 (0.60‐3.67)	PBO/Sham	28/3,003	M
Tolerability	Lurasidone	OR=0.45 (0.16‐1.22)	PBO/Sham	13/2,210	M
	Ziprasidone	OR=0.99 (0.45‐2.30)	PBO/Sham	13/2,210	M
	Risperidone	OR=2.38 (0.57‐13.56)	PBO/Sham	13/2,210	M
	Aripiprazole	OR=2.54 (0.70‐14.48)	PBO/Sham	13/2,210	M
	Asenapine	OR=2.67 (0.82‐12.47)	PBO/Sham	13/2,210	M
	Quetiapine	OR=3.29 (0.92‐16.75)	PBO/Sham	13/2,210	M
	Olanzapine	OR=7.76 (1.23‐87.44)	PBO/Sham	13/2,210	M
	Paliperidone	OR=23.12 (2.38‐778.70)	PBO/Sham	13/2,210	M
Discontinuation due to inefficacy	Paliperidone	OR=0.10 (0.04‐0.28)	PBO/Sham	28/3,003	L
	Olanzapine	OR=0.14 (0.06‐0.31)	PBO/Sham	28/3,003	L
	Risperidone	OR=0.17 (0.07‐0.42)	PBO/Sham	28/3,003	L
	Ziprasidone	OR=0.41 (0.20‐0.84)	PBO/Sham	28/3,003	L
	Lurasidone	OR=0.39 (0.09‐1.77)	PBO/Sham	28/3,003	M
	Asenapine	OR=0.63 (0.23‐1.73)	PBO/Sham	28/3,003	M
Depressive disorders
Pharmacological interventions
Efficacy (clinician‐rated)	Fluoxetine	SMD=–0.51 (–0.84 to –0.18)	PBO/Sham	70/8,906	M
	Desipramine	SMD=–0.43 (–1.26 to 0.39)	PBO/Sham	70/8,906	M
	Duloxetine	SMD = –0.22 (–0.85 to 0.42)	PBO/Sham	70/8,906	M
	Venlafaxine	SMD = –0.25 (–0.87 to 0.36)	PBO/Sham	70/8,906	M
	Mirtazapine	SMD = –0.23 (–0.97 to 0.51)	PBO/Sham	70/8,906	M
	Citalopram	SMD=–0.18 (–0.89 to 0.55)	PBO/Sham	70/8,906	M
	Escitalopram	SMD=–0.17 (–0.88 to 0.54)	PBO/Sham	70/8,906	M
	Paroxetine	SMD=–0.16 (–0.67 to 0.35)	PBO/Sham	70/8,906	M
	Nefazodone	SMD=–0.14 (–0.85 to 0.57)	PBO/Sham	70/8,906	M
	Desvenlafaxine	SMD=–0.12 (–0.79 to 0.54)	PBO/Sham	70/8,906	M
	Sertraline	SMD=–0.11 (–0.71 to 0.49)	PBO/Sham	70/8,906	M
	Imipramine	SMD=–0.03 (–0.75 to 0.68)	PBO/Sham	70/8,906	M
	Vilazodone	SMD=–0.09 (–1.09 to 0.90)	PBO/Sham	70/8,906	M
	Amitriptyline	SMD=0.08 (–1.11 to 1.27)	PBO/Sham	70/8,906	M
	Nortriptyline	SMD= 1.14 (0.46‐1.81)	PBO/Sham	70/8,906	M
Response	Nefazodone	OR=2.1 (1.06‐4.89)	PBO/Sham	34/5,260	M
	Duloxetine	OR=1.74 (1.12‐2.84)	PBO/Sham	34/5,260	M
	Fluoxetine	OR=1.70 (1.25‐2.39)	PBO/Sham	34/5,260	M
	Desipramine	OR=1.59 (0.67‐4.84)	PBO/Sham	34/5,260	M
	Escitalopram	OR=1.53 (0.96‐2.58)	PBO/Sham	34/5,260	M
	Sertraline	OR=1.44 (0.79‐2.97)	PBO/Sham	34/5,260	M
	Paroxetine	OR=1.3 (0.89‐1.99)	PBO/Sham	34/5,260	M
	Venlafaxine	OR=1.16 (0.72‐2.03)	PBO/Sham	34/5,260	M
	Citalopram	OR=1.02 (0.62‐1.82)	PBO/Sham	34/5,260	M
	Imipramine	OR=0.83 (0.48‐1.54)	PBO/Sham	34/5,260	M
	Nortriptyline	OR=0.57 (0.24‐1.64)	PBO/Sham	34/5,260	M
	Amitriptyline	OR=0.22 (0.05‐2.78)	PBO/Sham	34/5,260	M
Acceptability	Nefazodone	OR=0.49 (0.21‐1.39)	PBO/Sham	66/9,075	M
	Vilazodone	OR=0.59 (0.27‐1.54)	PBO/Sham	66/9,075	M
	Nortriptyline	OR=0.76 (0.28‐3.41)	PBO/Sham	66/9,075	M
	Fluoxetine	OR=0.78 (0.56‐1.15)	PBO/Sham	66/9,075	M
	Mirtazapine	OR=0.83 (0.40‐2.08)	PBO/Sham	66/9,075	M
	Desvenlafaxine	OR=0.85 (0.47‐1.74)	PBO/Sham	66/9,075	M
	Citalopram	OR=0.96 (0.52‐1.97)	PBO/Sham	66/9,075	M
	Duloxetine	OR=1.04 (0.62‐1.96)	PBO/Sham	66/9,075	M
	Venlafaxine	OR=1.12 (0.53‐2.70)	PBO/Sham	66/9,075	M
	Amitriptyline	OR=1.16 (0.29‐12.13)	PBO/Sham	66/9,075	M
	Paroxetine	OR=1.3 (0.81‐2.27)	PBO/Sham	66/9,075	M
	Escitalopram	OR=1.4 (0.77‐2.86)	PBO/Sham	66/9,075	M
	Sertraline	OR=162 (0.83‐3.22)	PBO/Sham	66/9,075	M
	Desipramine	OR=2.21 (0.88‐7.67)	PBO/Sham	66/9,075	M
	Imipramine	OR=2.51 (1.26‐6.25)	PBO/Sham	66/9,075	M
Tolerability	Amitriptyline	OR=0.10 (0.02‐32.16)	PBO/Sham	34/5,260	M
	Fluoxetine	OR=1.03 (0.5‐2.7)	PBO/Sham	34/5,260	M
	Citalopram	OR=1.13 (0.45‐3.66)	PBO/Sham	34/5,260	M
	Nefazodone	OR=1.29 (0.3‐21.89)	PBO/Sham	34/5,260	M
	Mirtazapine	OR=1.36 (0.41‐10.99)	PBO/Sham	34/5,260	M
	Paroxetine	OR=1.59 (0.77‐3.95)	PBO/Sham	34/5,260	M
	Escitalopram	OR=1.64 (0.46‐13.49)	PBO/Sham	34/5,260	M
	Desipramine	OR=2.85 (0.83‐21.8)	PBO/Sham	34/5,260	M
	Sertraline	OR=2.94 (0.94‐17.19)	PBO/Sham	34/5,260	M
	Duloxetine	OR=2.80 (1.20‐9.42)	PBO/Sham	34/5,260	M
	Venlafaxine	OR=3.19 (1.01‐18.7)	PBO/Sham	34/5,260	M
	Imipramine	OR=5.49 (1.96‐20.86)	PBO/Sham	34/5,260	M
Quality of life	Mixed (fluoxetine, paroxetine, sertraline)	SMD=–0.11 (–0.26 to 0.03)	PBO/Sham	3/765	M
Relapse	SSRIs	OR=0.34 (0.18‐0.64)	PBO/Sham	3/164	L
Remission	Fluoxetine	RR=1.82 (1.25‐2.63)	PBO/Sham	2/315	M
	Sertraline	RR=1.09 (0.72‐1.61)	PBO/Sham	2/376	M
Suicide attempt/ideation	Nefazodone	OR=0.29 (0.06‐6.31)	PBO/Sham	34/NR	M
	Mirtazapine	OR=0.53 (0.10‐40.83)	PBO/Sham	34/NR	M
	Imipramine	OR=0.59 (0.19‐3.07)	PBO/Sham	34/NR	M
	Desvenlafaxine	OR=0.74 (0.41‐1.49)	PBO/Sham	34/NR	M
	Escitalopram	OR=0.94 (0.44‐2.55)	PBO/Sham	34/NR	M
	Duloxetine	OR=0.93 (0.55‐1.71)	PBO/Sham	34/NR	M
	Fluoxetine	OR=1.11 (0.74‐1.75)	PBO/Sham	34/NR	M
	Paroxetine	OR=1.71 (0.81‐5.05)	PBO/Sham	34/NR	M
	Citalopram	OR=1.18 (0.46‐4.43)	PBO/Sham	34/NR	M
	Vilazodone	OR=1.96 (0.45‐100.00)	PBO/Sham	34/NR	M
	Sertraline	OR=2.22 (0.75‐12.5)	PBO/Sham	34/NR	M
	Venlafaxine	OR=8.33 (1.92‐NC)	PBO/Sham	34/NR	M
Psychosocial interventions
Efficacy (clinician‐rated)	IPT	SMD=–1.37 (–2.04 to –0.7)	WL/NT	70/8,906	L
	PSOLV	SMD=–1.26 (–2.48 to –0.03)	WL/NT	70/8,906	L
	FT	SMD=–1.03 (–1.66 to –0.4)	WL/NT	70/8,906	L
	CBT	SMD=–0.94 (–1.40 to –0.48)	WL/NT	70/8,906	L
	IPT	SMD=–0.70 (–1.29 to –0.12)	PBO/Sham	70/8,906	L
	FT	SMD=–0.36 (–0.95 to 0.24)	PBO/Sham	70/8,906	L
	CBT	SMD=–0.27 (–0.72 to 0.18)	PBO/Sham	70/8,906	L
	PSD‐O	SMD=0.08 (–0.67 to 0.84)	PBO/Sham	70/8,906	L
Response	PSD‐O	RR=1.68 (1.08‐2.63)	WL/PBO/Sham	2/83	L
Acceptability	IPT	OR=0.53 (0.20‐1.15)	PBO/Sham	66/9,075	M
	IPT	OR=0.65 (0.19‐1.62)	WL/NT	66/9,075	M
	CBT	OR=0.65 (0.32‐1.16)	PBO/Sham	66/9,075	M
	PSOLV	OR=0.77 (0.01‐4.40)	WL/NT	66/9,075	M
	CBT	OR=0.77 (0.34‐1.48)	WL/NT	66/9,075	M
	FT	OR=0.84 (0.35‐1.72)	PBO/Sham	66/9,075	M
	PSD‐O	OR=0.96 (0.37‐1.93)	PBO/Sham	66/9,075	M
	BT	OR=1.27 (0.19‐4.32)	PBO/Sham	66/9,075	M
Suicide attempt/ideation	IPT	OR=0.64 (0.04‐2.59)	PBO/Sham	34/NR	M
	CBT	OR=11.31 (0.01‐46.11)	PBO/Sham	34/NR	M
	PSD‐O	OR=8.64 (0.01‐40.05)	PBO/Sham	34/NR	M
Combination interventions
Efficacy (clinician‐rated)	Fluoxetine+CBT	SMD=–0.73 (–1.39 to –0.07)	PBO/Sham	70/8,906	M
Acceptability	Fluoxetine+CBT	OR=0.75 (0.39‐1.65)	PBO/Sham	66/9,075	M
Suicide attempt/ideation	Fluoxetine+CBT	OR=0.88 (0.41‐2.35)	PBO/Sham	34/NR	M
Bipolar disorder, depressive episode
Efficacy (clinician‐rated)	Quetiapine	SMD=–0.10 (–0.32 to 0.13)	PBO/Sham	2/224	M
Response	Quetiapine	RR=1.1 (0.89‐1.35)	PBO/Sham	3/250	L
Acceptability	Quetiapine	RR=0.73 (0.36‐1.49)	PBO/Sham	2/225	L
Global illness severity	Quetiapine	SMD=–0.20 (–0.46 to –0.06)	PBO/Sham	2/224	M
Remission	Quetiapine	RR=1.23 (0.90‐1.68)	PBO/Sham	3/250	L
Tolerability	Quetiapine	RR=0.31 (0.11‐1.01)	PBO/Sham	2/225	L
Bipolar disorder, manic episode
Efficacy (clinician‐rated)	Aripiprazole	SMD=–1.08 (–1.32 to –0.85)	PBO/Sham	2/339	M
Response	Mixed (mood stabilizers and antipsychotics)	OR=2.24 (z=8.12, p<0.001)	PBO/Sham	9/1,362	M
	Aripiprazole	RR=1.86 (1.43‐2.43)	PBO/Sham	2/332	M
	SGAs	z=10.34, p<0.001	PBO/Sham	6/1,190	H
	Mood stabilizers	z=2.06, p=0.04	PBO/Sham	2/172	M
Acceptability	Aripiprazole	RR=0.80 (0.51‐1.27)	PBO/Sham	2/339	M
	Valproate	OR=1.77 (0.83‐3.78)	PBO/Sham	2/179	M
Tolerability	Aripiprazole	RR=5.19 (0.92‐29.25)	PBO/Sham	2/339	M
Discontinuation due to inefficacy	Aripiprazole	RR=0.27 (0.09‐0.82)	PBO/Sham	2/339	M

RCTs – randomized controlled trials, SMD – standardized mean difference, OR – odds ratio, RR – risk ratio, PBO – placebo, WL – waiting list, NT – no treatment, NR – not reported, NC – not calculable, Q – quality (H – high, M – ­medium, L – low), BT – behavioral therapy, CBT – cognitive behavioral therapy, FT – family therapy, IPT – interpersonal therapy, PSD‐O – psychodynamic‐oriented, PSOLV – problem solving, SSRIs – selective serotonin reuptake inhibitors, SGAs – second‐generation antipsychotics. Bold prints indicate significant values. SMDs<0 indicate that intervention is more effective than control. For discontinuation outcomes (acceptability, tolerability, inefficacy) and relapse, OR/RR<1 favors the intervention. For response and remission, OR/RR>1 favors the intervention.

Fluoxetine was the only pharmacological intervention that was superior to placebo on the primary efficacy outcome (medium effect size), as well as on response and remission (both small effect size). Nortriptyline worsened the primary efficacy outcome (large effect size), imipramine increased all‐cause drop‐out (small effect size), and imipramine, venlafaxine and duloxetine increased discontinuation due to intolerability (small to medium effect size). Venlafaxine increased suicidality (large effect size) (see Table 3).

Among psychosocial interventions, a large effect size on the primary efficacy outcome was apparent for interpersonal therapy, problem‐solving therapy, family therapy, and CBT vs. waiting list/no treatment. However, these results were not confirmed vs. placebo or vs. TAU, except for interpersonal therapy, that remained superior when compared to placebo and TAU (medium effect size) (see Tables 3 and 5).

CBT was also superior to mixed interventions regarding the primary efficacy outcome (medium effect size), and to selective serotonin reuptake inhibitors (SSRIs) regarding suicidality (small effect size) (see Tables 3 and 6). Psychodynamically‐oriented psychotherapy had a small effect size advantage regarding response, but no significant effect on the primary efficacy outcome vs. placebo (see Table 3).

As a combination treatment, CBT plus fluoxetine had a medium effect size advantage regarding the primary efficacy outcome vs. placebo (see Table 3), and CBT plus SSRI was superior concerning remission vs. CBT monotherapy, and functioning vs. antidepressant monotherapy (small effect size) (see Table 6).

Enuresis

Results for enuresis are shown in Tables 4 and 6.

Table 4.

Efficacy and effectiveness of pharmacological, psychosocial and brain stimulation interventions vs. inactive control in children/adolescents with anxiety, obsessive‐compulsive, stress‐related, and mixed disorders

Outcome	Intervention	Effect size (95% CI)	Control	Number of RCTs/patients	Q
Anxiety disorders
Pharmacological interventions
Efficacy (clinician‐rated)	Paroxetine	SMD=–0.43 (–0.75 to –0.10)	PBO/Sham	14/2,502	M
	Fluvoxamine	SMD=–0.36 (–0.61 to –0.10)	PBO/Sham	14/2,502	M
	Imipramine	SMD=–0.27 (–0.92 to 0.39)	PBO/Sham	14/2,502	M
	Guanfacine	SMD=–0.13 (–0.39 to 0.12)	PBO/Sham	14/2,502	M
	Fluoxetine	SMD=–0.11 (–0.33 to 0.12)	PBO/Sham	14/2,502	M
	Atomoxetine	SMD=–0.11 (–0.38 to 0.16)	PBO/Sham	14/2,502	M
	Duloxetine	SMD=–0.09 (–0.27 to 0.09)	PBO/Sham	14/2,502	M
	Sertraline	SMD=–0.08 (–0.25 to 0.09)	PBO/Sham	14/2,502	M
	Venlafaxine	SMD=–0.06 (–0.22 to 0.04)	PBO/Sham	14/2,502	M
Efficacy (subject‐rated)	Fluoxetine	SMD=–0.51 (–0.85 to –0.18)	PBO/Sham	2/154	M
	SNRIs	SMD=–2.14 (–9.75 to 5.48)	PBO/Sham	3/622	M
	Venlafaxine	SMD=–1.71 (–3.93 to 0.51)	PBO/Sham	2/443	M
	SSRIs	SMD=–0.42 (–0.96 to 0.12)	PBO/Sham	4/197	M
	Atomoxetine	SMD=–0.29 (–0.51 to 0.08)	PBO/Sham	2/331	M
	TCAs	SMD= 0.36 (–0.27 to 0.99)	PBO/Sham	2/41	M
Efficacy (parent‐rated)	SSRIs	SMD=–0.82 (–1.38 to –0.27)	PBO/Sham	2/96	L
Response	Fluvoxamine	OR=8.17 (1.35‐49.40)	PBO/Sham	19/2,656	M
	Sertraline	OR=6.05 (2.23‐49.40)	PBO/Sham	19/2,656	M
	Fluoxetine	OR=4.06 (1.49‐18.17)	PBO/Sham	19/2,656	M
	Guanfacine	OR=5.47 (0.74‐49.40)	PBO/Sham	19/2,656	M
	Atomoxetine	OR=4.06 (0.67‐24.53)	PBO/Sham	19/2,656	M
	Paroxetine	OR=3.67 (0.67‐20.09)	PBO/Sham	19/2,656	M
	Imipramine	OR=3.00 (0.61‐14.88)	PBO/Sham	19/2,656	M
	Venlafaxine	OR=2.46 (0.90‐6.69)	PBO/Sham	19/2,656	M
	Duloxetine	OR=2.01 (0.37‐11.02)	PBO/Sham	19/2,656	M
	Clomipramine	OR=1.22 (0.22‐6.69)	PBO/Sham	19/2,656	M
Acceptability	Clomipramine	OR=0.55 (0.02‐7.39)	PBO/Sham	20/2,679	M
	Paroxetine	OR=0.61 (0.12‐3.32)	PBO/Sham	20/2,679	M
	Fluvoxamine	OR=0.67 (0.11‐4.06)	PBO/Sham	20/2,679	M
	Sertraline	OR=0.67 (0.14‐2.72)	PBO/Sham	20/2,679	M
	Guanfacine	OR=0.67 (0.10‐4.95)	PBO/Sham	20/2,679	M
	Atomoxetine	OR=0.82 (0.15‐4.95)	PBO/Sham	20/2,679	M
	Duloxetine	OR=1.00 (0.18‐5.47)	PBO/Sham	20/2,679	M
	Venlafaxine	OR=1.11 (0.33‐3.67)	PBO/Sham	20/2,679	M
	Fluoxetine	OR=1.65 (0.50‐6.69)	PBO/Sham	20/2,679	M
	Imipramine	OR=2.01 (0.37‐9.97)	PBO/Sham	20/2,679	M
Remission	Fluoxetine	RR=2.52 (1.19‐5.32)	PBO/Sham	2/95	L
Suicide attempt/ ideation	Sertraline	LogOR=–19.8 (–61.7 to 0.7)	PBO/Sham	9/1,648	M
	Duloxetine	LogOR=0.2 (–2.5 to 2.8)	PBO/Sham	9/1,648	M
	Venlafaxine	LogOR=1.4 (–1.4 to 5.24)	PBO/Sham	9/1,648	M
	Atomoxetine	LogOR=6.6 (–31.6 to 22.7)	PBO/Sham	9/1,648	M
	Guanfacine	LogOR=16.1 (–1.0 to 58.3)	PBO/Sham	9/1,648	M
	Imipramine	LogOR=17.3 (–0.1 to 54.8)	PBO/Sham	9/1,648	M
	Paroxetine	LogOR=20.0 (1.7 to 60.47)	PBO/Sham	9/1,648	M
Tolerability	Venlafaxine	LogOR=–0.8 (–3.8 to 2.1)	PBO/Sham	15/2,516	M
	Atomoxetine	LogOR=0.0 (–5.3 to 5.3)	PBO/Sham	15/2,516	M
	Duloxetine	LogOR=0.2 (–3.9 to 4.3)	PBO/Sham	15/2,516	M
	Sertraline	LogOR=1.7 (–2.8 to 6.6)	PBO/Sham	15/2,516	M
	Paroxetine	LogOR=1.7 (–2.5 to 6.0)	PBO/Sham	15/2,516	M
	Fluovoxamine	LogOR=2.1 (–2.4 to 7.0)	PBO/Sham	15/2,516	M
	Fluoxetine	LogOR=2.5 (–1.8 to 7.9)	PBO/Sham	15/2,516	M
	Imipramine	LogOR=16.6 (–37.5 to 83.7)	PBO/Sham	15/2,516	M
	Guanfacine	LogOR=29.2 (2.2‐94.3)	PBO/Sham	15/2,516	M
Psychosocial interventions
Efficacy (clinician ‐rated)	CBT/BT	SMD=–0.85 (–1.12 to –0.57)	WL/NT	7/358	L
Efficacy (subject‐rated)	CBT‐Child only	SMD=–1.04 (–1.41 to –0.67)	WL/NT	24/1,239	L
	CBT‐Group	SMD=–0.91 (–1.22 to –0.60)	WL/NT	27/1,268	L
	CBT	SMD=–0.67 (–0.88 to –0.47)	WL/NT	45/2,831	L
	CBT‐Child+P	SMD=–0.45 (–0.67 to –0.23)	WL/NT	20/1,285	L
	CBT‐Individual	SMD=–0.39 (–0.64 to –0.15)	WL/NT	21/1,203	L
	CBT	SMD=–0.31 (–0.51 to –0.11)	PBO/Sham	15/978	L
	CBT‐Parent only	SMD=0.04 (–0.38 to 0.46)	WL/NT	5/307	L
Efficacy (parent‐rated)	CBT‐Group	SMD=–0.92 (–1.21 to –0.62)	WL/NT	21/1,279	L
	CBT‐Child only	SMD=–0.87 (–1.21 to –0.53)	WL/NT	13/734	L
	CBT	SMD=–0.70 (–0.90 to –0.51)	WL/NT	35/2137	L
	CBT‐Child+P	SMD=–0.69 (–0.98 to –0.39)	WL/NT	17/1,031	L
	CBT‐Individual	SMD=–0.43 (–0.65 to –0.21)	WL/NT	17/858	L
	CBT‐Parent only	SMD=–0.37 (–0.77 to 0.04)	WL/NT	5/372	L
	CBT	SMD=–0.25 (–0.61 to 0.11)	PBO/Sham	8/638	L
Efficacy (mixed‐rated)	BT‐Group	SMD=–1.43 (–2.36 to –0.51)	WL/NT	101/6,625	L
	CBT‐Group	SMD=–1.43 (–1.76 to –1.09)	WL/NT	101/6,625	L
	BT‐Individual+P	SMD=–1.09 (–1.93 to –0.25)	WL/NT	101/6,625	L
	CBT‐Group+P	SMD=–0.99 (–1.31 to –0.68)	WL/NT	101/6,625	L
	CBT‐Individual	SMD=–0.99 (–1.30 to –0.68)	WL/NT	101/6,625	L
	CBT‐Individual+P	SMD=–0.84 (–1.16 to –0.53)	WL/NT	101/6,625	L
	CBT‐Group	SMD=–0.76 (–1.16 to –0.36)	PBO/Sham	101/6,625	L
	CBT‐Parent only	SMD=–0.70 (–1.22 to –0.19)	WL/NT	101/6,625	L
	CBT‐Internet	SMD=–0.61 (–1.02 to –0.20)	WL/NT	101/6,625	L
	BT‐Individual+Group	SMD=–0.73 (–1.59 to 0.13)	WL/NT	101/6,625	L
	CBT‐Individual+Group	SMD=–0.64 (–1.69 to 0.41)	WL/NT	101/6,625	L
	BT‐Individual+P	SMD=–0.42 (–1.29 to 0.44)	PBO/Sham	101/6,625	L
	CBT‐Group+P	SMD=–0.33 (–0.78 to 0.13)	PBO/Sham	101/6,625	L
	CBT‐Individual	SMD=–0.32 (–0.72 to 0.07)	PBO/Sham	101/6,625	L
	CBT‐Individual+P	SMD=–0.18 (–0.61 to 0.25)	PBO/Sham	101/6,625	L
	BT‐Individual+Group	SMD=–0.06 (–0.94 to 0.82)	PBO/Sham	101/6,625	L
	CBT‐Internet	SMD=0.06 (–0.48 to 0.60)	PBO/Sham	101/6,625	L
Acceptability	CBT‐Individual+Group	OR=0.26 (0.05‐5.73)	WL/NT	101/6,625	L
	BT‐Individual+P	OR=0.64 (0.22‐2.72)	WL/NT	101/6,625	L
	BT‐Individual+P	OR=0.81 (0.19‐2.27)	PBO/Sham	101/6,625	L
	CBT‐Group+P	OR=0.90 (0.46‐1.60)	PBO/Sham	101/6,625	L
	CBT‐Group	OR=0.85 (0.46‐1.44)	PBO/Sham	101/6,625	L
	BT	OR=0.90 (0.32‐3.95)	WL/NT	101/6,625	M
	CBT‐Individual	OR=0.92 (0.52‐1.52)	PBO/Sham	101/6,625	L
	CBT‐Group	OR=0.93 (0.57‐1.63)	WL/NT	101/6,625	L
	CBT	OR=1.09 (0.85‐1.41)	WL/NT	45/3,158	L
	CBT‐Group+P	OR=0.99 (0.67‐1.55)	WL/NT	101/6,625	M
	CBT	OR=1.00 (0.68‐1.49)	PBO/Sham	12/797	L
	CBT‐Internet	OR=1.02 (0.42‐2.08)	PBO/Sham	101/6,625	L
	CBT‐Individual	OR=1.02 (0.67‐1.67)	WL/NT	101/6,625	L
	CBT‐Internet	OR=1.05 (0.59‐2.05)	WL/NT	101/6,625	L
	CBT‐Individual+P	OR=1.11 (0.60‐1.90)	PBO/Sham	101/6,625	L
	BT‐Individual+Group	OR=1.13 (0.28‐3.19)	PBO/Sham	101/6,625	L
	BT‐Group	OR=1.21 (0.27‐22.51)	WL/NT	101/6,625	L
	CBT‐Individual+P	OR=1.23 (0.80‐2.02)	WL/NT	101/6,625	L
	CBT‐Parent only	OR=1.43 (0.75‐3.15)	WL/NT	101/6,625	L
Depressive symptoms	CBT	SMD=–0.34 (–0.51 to –0.17)	WL/NT	17/1,157	L
	CBT	SMD=–0.18 (–0.45 to 0.09)	PBO/Sham	10/613	L
Functioning	CBT	SMD=–1.03 (–1.38 to –0.68)	WL/NT	11/557	L
Quality of life	CBT‐Parent only	SMD=–1.87 (–3.04 to –0.71)	WL/NT	101/6,625	L
	CBT‐Individual	SMD=–1.13 (–1.82 to –0.45)	PBO/Sham	101/6,625	L
	CBT‐Individual	SMD=–1.01 (–1.55 to –0.48)	WL/NT	101/6,625	L
	CBT‐Internet	SMD=–0.86 (–1.57 to –0.15)	PBO/Sham	101/6,625	L
	CBT‐Group	SMD=–0.85 (–1.45 to –0.26)	PBO/Sham	101/6,625	L
	CBT‐Individual+P	SMD=–0.80 (–1.33 to –0.27)	WL/NT	101/6,625	L
	CBT‐Group+P	SMD=–0.75 (–1.34 to –0.17)	WL/NT	101/6,625	L
	CBT‐Group	SMD=–0.73 (–1.34 to –0.11)	WL/NT	101/6,625	L
	CBT‐Internet	SMD=–0.73 (–1.14 to –0.33)	PBO/Sham	101/6,625	L
	BT‐Individual+Group	SMD=–0.79 (–1.68 to 0.09)	WL/NT	101/6,625	L
	BT‐Individual+Group	SMD=–0.67 (–1.56 to 0.21)	WL/NT	101/6,625	L
	CBT‐Individual+Group	SMD=–0.55 (–1.78 to 0.69)	WL/NT	101/6,625	L
Remission	CBT‐Child only	OR=10.42 (5.84‐7.60)	WL/NT	19/1,184	M
	CBT‐Group	OR=6.25 (4.45‐8.78)	WL/NT	25/1,532	M
	CBT‐Remote	OR=6.14 (2.97‐12.71)	WL/NT	10/591	L
	CBT	OR=5.45 (3.90‐7.60)	WL/NT	39/2,697	L
	CBT‐Individual	OR=4.53 (2.55‐8.03)	WL/NT	17/1,165	L
	CBT‐Individual+P	OR=4.08 (2.72‐6.11)	WL/NT	19/1,142	M
	CBT‐Child only	OR=3.58 (1.92‐6.65)	PBO/Sham	7/509	L
	CBT‐Group	OR=3.10 (1.14‐8.45)	PBO/Sham	5/353	L
	CBT‐Parent only	OR=2.83 (1.12‐7.16)	WL/NT	4/371	L
	CBT	OR=2.28 (1.33‐3.89)	PBO/Sham	10/822	L
	CBT‐Individual	OR=2.04 (1.06‐3.91)	PBO/Sham	5/469	L
	CBT‐Individual+P	OR=1.12 (0.65‐1.92)	PBO/Sham	4/313	L
Social anxiety disorder
Efficacy (subject‐rated)	CBT	SMD=–1.59 (–2.33 to –0.86)	WL/NT	11/603	L
	BT	SMD=–1.22 (–2.06 to –0.38)	WL/NT/PBO/Sham	4/169	L
	CBT	SMD=–1.19 (–1.72 to –0.67)	WL/NT/PBO/Sham	14/872	L
	CBT‐Group	SMD=–1.19 (–1.93 to –0.45)	WL/NT/PBO/Sham	11/670	L
	CBT/BT	SMD=–1.13 (–1.59 to –0.68)	WL/NT/PBO/Sham	17/1,016	L
	CBT+P	SMD=–1.13 (–1.59 to –0.67)	WL/NT/PBO/Sham	17/983	L
	CBT‐Individual	SMD=–1.10 (–1.91 to –0.29)	WL/NT/PBO/Sham	3/127	L
	CBT‐Individual+Group	SMD=–0.80 (–1.19 to –0.41)	WL/NT/PBO/Sham	3/115	L
	CBT‐Child only	SMD=–0.75 (–1.24 to –0.26)	WL/NT/PBO/Sham	2/70	L
	CBT‐Internet	SMD=–0.52 (–1.01 to –0.03)	WL/NT/PBO/Sham	2/143	L
Acceptability	CBT	RR=1.00 (0.72‐1.41)	WL/NT/PBO/Sham	16/1,052	M
Depressive symptoms	CBT/BT	SMD=–0.39 (–0.63 to –0.16)	WL/NT/PBO/Sham	8/299	L
Quality of life	CBT/BT	SMD=–0.79 (–1.17 to –0.41)	WL/NT/PBO/Sham	9/552	L
Remission	CBT/BT	RR=8.99 (5.27‐15.33)	WL/NT/PBO/Sham	13/832	L
Obsessive‐compulsive disorder
Pharmacological interventions
Efficacy (clinician‐rated)	Sertraline	SMD=–0.24 (–0.46 to –0.03)	PBO/Sham	17/991	L
	Fluoxetine	SMD=–0.24 (–0.47 to –0.01)	PBO/Sham	17/991	L
	Clomipramine	SMD=–0.31 (–0.64 to 0.02)	PBO/Sham	17/991	L
	Fluvoxamine	SMD=–0.21 (–0.49 to 0.06)	PBO/Sham	17/991	L
Response	Fluoxetine	RR=1.49 (1.15‐1.96)	PBO/Sham	2/146	L
	SSRI/TCAs	RR=1.80 (1.43‐2.26)	PBO/Sham	7/692	L
Acceptability	Fluoxetine	MOR=0.74 (0.25‐1.68)	PBO/Sham	18/1,143	L
	Fluvoxamine	MOR=0.79 (0.24‐2.07)	PBO/Sham	18/1,143	L
	Sertraline	MOR=0.89 (0.32‐2.07)	PBO/Sham	18/1,143	L
	Paroxetine	MOR=1.12 (0.37‐3.42)	PBO/Sham	18/1,143	L
	Clomipramine	MOR=3.06 (0.54‐21.69)	PBO/Sham	18/1,143	L
Tolerability	SSRIs	RR=3.59 (1.89‐6.84)	PBO/Sham	7/807	L
Global illness severity	Fluoxetine	SMD=–0.52 (–0.86 to –0.18)	PBO/Sham	2/146	L
	SSRIs	SMD=–0.42 (–0.61 to –0.23)	PBO/Sham	5/556	M
Remission	SSRIs	RR=2.06 (1.03‐4.13)	PBO/Sham	3/302	L
Pharmacological augmentation (in SSRI‐refractory cases)
Response	Risperidone	OR=6.35 (1.48‐27.3)	PBO/Sham	3/72	M
	Quetiapine	OR=2.33 (0.88‐6.20)	PBO/Sham	3/102	M
	Olanzapine	OR=2.74 (0.34‐21.9)	PBO/Sham	2/70	L
Psychosocial interventions
Efficacy (clinician‐rated)	CBT	SMD=–0.78 (–1.05 to –0.51)	WL/NT	17/991	L
	BT	SMD=–0.72 (–1.20 to –0.24)	WL/NT	17/991	L
	CBT	SMD=–0.23 (–0.56 to 0.11)	PBO/Sham	17/991	L
Response	CBT/BT‐ERP	RR=3.93 (2.52‐6.14)	WL/NT/PBO/Sham	6/236	L
Acceptability	CBT	MOR=0.49 (0.09‐2.40)	PBO/Sham	18/1,143	L
	BT‐ERP	RR=0.80 (0.35‐1.84)	PBO/WL	6/301	L
	CBT	MOR=0.86 (0.23‐3.24)	PBO/Sham	18/1,143	L
	CBT	MOR=0.94 (0.21‐4.79)	WL/NT	18/1,143	L
	BT	MOR=14.28 (0.87‐785.20)	WL/NT	18/1,143	L
Functioning (subject‐rated)	CBT	SMD=–1.15 (–2.11 to –0.19)	WL/NT	3/194	L
Functioning (parent‐rated)	CBT	SMD=–0.95 (–1.61 to –0.28)	WL/NT	3/194	L
	CBT	SMD=–0.31 (–0.63 to 0.01)	PBO/Sham	2/183	L
Remission	CBT	RR=2.33 (1.33‐4.00)	WL/NT	4/271	L
	CBT	RR=1.59 (1.28‐1.96)	PBO/Sham	3/153	L
Quality of life	CBT	SMD=–0.39 (–0.77 to –0.02)	WL/PBO/Sham	2/223	L
Combined interventions
Efficacy	CBT+sertraline	SMD=–0.58 (–0.91 to –0.25)	PBO/Sham	17/991	L
Acceptability	CBT+sertraline	MOR=0.54 (0.08‐3.15)	PBO/Sham	18/1,143	L
Post‐traumatic stress disorder
Efficacy	CBT	SMD=–1.34 (–1.79 to –0.89)	WL/NT	3/98	L
	EMDR	SMD=–0.61 (–1.96 to 0.74)	WL/NT	2/65	L
	NET	SMD=–0.57 (–1.23 to 0.09)	WL/NT	2/79	L
Response	CBT	OR=8.64 (2.01‐37.14)	WL/NT	2/49	L
	NET	OR=3.82 (0.67‐21.8)	WL/NT	2/78	L
Acceptability	NET	OR=5.13 (0.56‐47.28)	WL/NT	2/83	L
Anxiety symptoms	NET	SMD=–0.66 (–1.33 to 0.01)	WL/NT	2/59	L
Depressive symptoms	CBT	SMD=–0.8 (–1.47 to –0.131)	WL/NT	3/98	L
Enuresis
Pharmacological interventions
Efficacy	Imipramine	SMD=–0.46 (–0.67 to –0.24)	PBO/Sham	4/347	M
Response	Amitriptyline	RR=1.22 (1.02‐1.45)	PBO/Sham	2/98	L
	Imipramine	RR=1.35 (1.11‐1.64)	PBO/Sham	12/831	L
Psychosocial interventions
Efficacy	BT‐Alarm	SMD=–1.30 (–2.16 to –0.44)	WL/NT	4/127	L
Response	BT‐Alarm	RR=7.23 (1.40‐37.77)	WL/NT	18/827	L
	BT‐Alarm	RR=1.59 (1.16‐2.17)	PBO/Sham	2/181	L
	BT‐Reward	RR=1.22 (1.03‐1.45)	WL/NT	2/325	L

RCTs – randomized controlled trials, SMD – standardized mean difference, OR – odds ratio, MOR – median odds ratio, RR – risk ratio, PBO – placebo, WL – waiting list, NT – no treatment, Q – ­quality (H – high, M – medium, L – low), BT – behavioral therapy, BT‐ERP – behavioral therapy with exposure and response prevention, CBT – cognitive ­behavioral therapy, EMDR – eye movement desensitization and reprocessing, NET – narrative exposure therapy, P – parental involvement, SSRIs – selective ­serotonin reuptake inhibitors, SNRIs – serotonin‐norepinephrine reuptake inhibitors, TCAs – tricyclic antidepressants. Bold prints indicate significant values. SMDs<0 indicate that intervention is more effective than control. For discontinuation outcomes (acceptability, tolerability, inefficacy) and relapse, OR/RR<1 favors the intervention. For response and remission, OR/RR>1 favors the intervention.

Among pharmacological interventions, imipramine outperformed placebo regarding the primary efficacy outcome and response (small effect size), and amitriptyline was superior to placebo with respect to response (small effect size) (see Table 4).

Behavioral therapy with alarm outperformed waiting list on the primary efficacy outcome (small effect size) and response (large effect size), and maintained a small effect size regarding response vs. placebo (see Table 4).

No clear superior treatment emerged in monotherapy head‐to‐head comparisons. Combination of desmopressin plus behavioral therapy with alarm was superior to desmopressin alone regarding the primary efficacy outcome (medium effect size) and response (small effect size), while combination of oxybutynin plus imipramine was superior to either imipramine or oxybutynin monotherapy (small effect size) (see Table 6).

Obsessive‐compulsive disorder

Results for obsessive‐compulsive disorder are shown in Tables 4 and 5.

Fluoxetine was the pharmacological intervention with the broadest efficacy, including primary efficacy outcome, response, and global illness severity vs. placebo (small effect sizes). SSRIs as a class also improved response, remission and global illness severity, yet had a higher discontinuation rate due to intolerability than placebo (see Table 4).

Among monotherapy psychosocial interventions, CBT was superior to waiting list regarding the primary efficacy outcome (medium effect size), response (small effect size), remission (small effect size), quality of life (small effect size) and functioning (large effect size), and also to placebo concerning remission (small effect size) (see Table 4). Behavioral therapy with exposure and response prevention outperformed TAU for both response and acceptability (small effect size) ­(see Table 5).

As a combination treatment, CBT and sertraline outperformed placebo (medium effect size) (see Table 4). No significant differences emerged in head‐to‐head comparisons.

Anxiety disorders

Results for anxiety disorders are shown in Tables 4, 5 and 6.

SSRIs (fluoxetine, fluvoxamine, paroxetine) outperformed pla­cebo regarding the primary efficacy outcome, and response (small to medium effect). Fluoxetine also outperformed placebo with respect to remission (small effect size) (see Table 4). Sertraline reduced suicidality compared with placebo, but paroxetine increased it.

CBT was superior to waiting list in different formats (i.e., individual, Internet, group) regarding the primary efficacy outcome (small to large effect size), depressive symptoms (small effect size), remission (small to large effect size) and quality of life (large effect size). CBT was also superior to placebo with respect to quality of life (large effect size) and to TAU regarding the primary efficacy outcome, remission and functioning (large effect size). Group CBT was superior to individual CBT in head‐to‐head comparisons (small effect size) (see Tables 4, 5 and 6).

No meta‐analysis compared pharmacological vs. psychosocial interventions or combined treatment strategies.

Disruptive behavior/dissocial/conduct disorders

Results for disruptive behavior/dissocial/conduct disorders are shown in Tables 2 and 7.

Among pharmacological interventions, risperidone outperformed placebo across different raters regarding the primary efficacy outcome (medium effect size), aggressive behavior (medium effect size, also in people with intellectual disability), and global illness severity (medium effect size). Aggressive behavior was also improved by lithium and valproate (see Table 2).

Among psychosocial interventions, a combination of parental and child behavioral interventions had a large effect size vs. waiting list concerning the primary efficacy outcome, and a medium effect size vs. a mixed control group (see Tables 2 and 7).

Eating disorders

Results for eating disorders are shown in Table 6.

No meta‐analysis on pharmacological intervention met the inclusion criteria of this umbrella review. Among psychosocial interventions, family therapy outperformed other interventions in anorexia nervosa regarding the primary efficacy outcome (body weight, small effect size).

Schizophrenia spectrum disorders

Results for schizophrenia spectrum disorders are shown in Tables 3 and 6.

For schizophrenia, only pharmacological interventions were covered. All investigated antipsychotics but ziprasidone outperformed placebo, with a small effect size, except for olanzapine and risperidone, which had a large effect size. Small effect sizes emerged regarding response (except for asenapine), and all antipsychotics improved global illness severity. Acceptability was superior vs. placebo for paliperidone, risperidone and olanzapine, without differences for the other antipsychotics. Paliperidone and olanzapine were associated with more discontinuation due to intolerability than placebo, while discontinuation due to inefficacy favored paliperidone, olanzapine, risperidone and ziprasidone (see Table 3).

In head‐to‐head comparisons, risperidone and second‐generation antipsychotics outperformed first‐generation antipsychotics (large effect size), and clozapine outperformed olanzapine on the primary efficacy outcome (large effect size) (see Table 6).

Bipolar disorder

Results for bipolar disorder are shown in Tables 3 and 6.

Regarding bipolar depression, quetiapine was not superior to placebo regarding the primary efficacy outcome, separating only on global illness severity (small effect size). Regarding mania, aripiprazole was more effective than placebo regarding the primary efficacy outcome (large effect size) and response (small effect size), without differences vs. placebo regarding acceptability, while being superior regarding less discontinuations for inefficacy (see Table 3).

Other disorders

Results for tic disorder are shown in Tables 2 and 6. Desipramine and methylphenidate were similar to placebo, but topiramate was superior to haloperidol regarding the primary outcome.

Results for Tourette's disorder are shown in Tables 2 and 7. Antipsychotics (including haloperidol, pimozide, risperidone and ziprasidone) and guanfacine were superior to placebo regarding the primary efficacy outcome (both moderate effect size). No significant difference vs. placebo emerged for methylphenidate (see Table 2). Among psychosocial interventions, behavioral therapy outperformed waiting list or low intensity psychosocial intervention (medium effect size) regarding the primary efficacy outcome (see Table 7).

Results for encopresis are shown in Table 5. No pharmacological intervention was eligible. Behavioral therapy outperformed TAU regarding the primary efficacy outcome and response (small effect size).

Results for developmental coordination disorders are shown in Table 2. In the single meta‐analysis meeting inclusion criteria, skills training had no significant effect vs. waiting list on motor coordination.

Results for PTSD are shown in Table 4. No pharmacological intervention met inclusion criteria. CBT was superior regarding the primary efficacy outcome, response and depressive symptoms vs. waiting list (large effect sizes).

DISCUSSION

Pooling top‐tier evidence from 104 MAs/NMAs of RCTs reporting on the effects of pharmacological, psychosocial and brain stimulation interventions, targeting 20 different outcomes in 15 mental disorders or groups of mental disorders, this umbrella review provides a comprehensive meta‐analytic view of the evidence base regarding the efficacy, acceptability and other relevant outcomes of psychiatric treatments in children and adolescents (see supplementary information for further details).

Considered together with a complementary umbrella review published in this journal ¹⁴ , focusing on the detailed evaluation of tolerability and safety of pharmacological interventions, the current review can inform clinicians, youth and their families, as well as other stakeholders, in making evidence‐based decisions regarding the choice and use of pharmacological, psychosocial and brain stimulation interventions in children/adolescents, in monotherapy and in combination. On the basis of these reviews, some evidence‐based recommendation can be made.

For ADHD, amphetamines and methylphenidate are the most effective interventions on a broad set of outcomes. Whilst amphetamines outperform methylphenidate on the primary efficacy outcome, methylphenidate is the medication least different from placebo concerning safety ¹⁴ . Some evidence is available regarding behavioral therapy, covering a narrow set of efficacy outcomes, and with small effect sizes compared with those for medications. Importantly, whilst social skills training shows promising results against waiting list, no evidence is available comparing this intervention with placebo. Hence, amphetamines or methylphenidate can be considered the first‐line treatment, augmented with alpha‐2 agonists if needed, and ideally in combination with behavioral therapy as an optimal treatment regimen. Behavioral therapy could be considered if medications are contraindicated.

For autism, aripiprazole and risperidone are the pharmacological treatment options of choice. However, various psychosocial interventions have proven efficacy on a broad set of outcomes, ranging from anxiety (CBT), to irritability, aggressive behavior and functioning (parent‐child interaction therapy), to the primary efficacy outcome and functioning (social skills training, and behavioral therapy with imitative component). These benefits are not only observed vs. waiting list, but also against other active interventions. Given the different outcomes that these treatment modalities target, a variety of therapeutic tools can be considered, according to the patient's and family's resources, needs and choice, as well as the disease course and the presence of environmental stressors.

For depressive disorders in youth, fluoxetine is the only evidence‐based pharmacological option. All other medications do not improve depression vs. placebo, but placebo effects are considerable. Imipramine, nortriptyline, and likely also venlafaxine should be avoided, given poor acceptability, tolerability and safety. As an alternative to medications, interpersonal therapy is the only psychosocial intervention outperforming placebo. The combination of CBT with fluoxetine also outperformed placebo on the primary efficacy outcome, and was superior to either monotherapy.

For enuresis, imipramine is the most effective pharmacological intervention. It can be combined with oxybutynin to maximize efficacy. However, due to the potential problems with tolerability of this medication in youth, psychosocial interventions should be tried first, including especially alarm behavioral therapy, that is supported by the largest body of evidence. No difference emerges among different types of alarms, and alarm maintains its efficacy after stopping the ­intervention ⁸⁶ .

For obsessive‐compulsive disorder, fluoxetine and SSRIs as a class should be considered the first‐line pharmacological treatment. Among psychosocial interventions, CBT and behavioral therapy with exposure and response prevention are effective options. If fluoxetine/SSRIs are ineffective, a switch to psychosocial interventions should be performed, and vice versa ⁷¹ .

For anxiety disorders, fluoxetine and fluvoxamine are evidence‐based pharmacological treatment strategies. Among psychosocial interventions, CBT – and in particular group CBT – should be offered as first‐line treatment, likely before medications, given the large effect size and broad beneficial effect even vs. placebo in children and adolescents.

For disruptive behavior/dissocial/conduct disorders, risperidone emerges as the most effective pharmacological agent, but different types of behavioral treatment (including parent training) should be regarded as the first‐line treatment ­options118, 119.

For anorexia nervosa in children and adolescents, family therapy is the intervention supported by the most significant evidence.

For schizophrenia spectrum disorders, antipsychotic treatment is the cornerstone of treatment. All tested antipsychotics, except for ziprasidone, have broadly similar superior efficacy vs. placebo, with olanzapine and risperidone being the most effective, and lurasidone/aripiprazole a more tolerable treatment option ¹⁰² . Ideally, starting with safer medications minimizing the risk of adverse events and maximizing adherence is a recommended strategy ¹⁴ .

For bipolar disorder, little meta‐analytic evidence is available overall. For mania, the only positive data are available for aripiprazole, yet lithium is also an evidence‐based treatment based on RCT evidence ¹²⁰ . For bipolar depression, only quetiapine is superior to placebo, and only on a single outcome, namely global illness severity, but not on the primary symptom outcome. This finding is different from adults ¹²¹ , and at least partially due to the larger placebo effects in youth. Our umbrella review did not include lurasidone and olanzapine/fluoxetine combination, as no meta‐analysis has been conducted on them, but these are evidence‐based options to treat bipolar depression in youth based on single RCTs122, 123, which led to their approval by the US Food and Drug Administration for bipolar depression in children and adolescents.

The available evidence presented in this umbrella review is not equally large across individual disorders, and also across monotherapies with pharmacological or psychosocial interventions. Even less meta‐analytic data are available for head‐to‐head studies, within and across treatment modalities, and regarding combination treatments. Furthermore, little meta‐analytic evidence exists on treatment‐resistant youth with a given mental disorder. This is concerning, as early illness onset and disruption of healthy development may portend poorer response and outcomes, requiring information on non‐responding conditions after first‐ and second‐line treatments have been tried.

Among the 104 included meta‐analyses, virtually none reported data on long‐term treatment or relapse prevention. This is problematic, as most of these disorders are chronic and require long‐term treatment.

This umbrella review clearly shows that large effect sizes emerge for psychosocial interventions when they are compared with waiting list or no treatment, where no placebo or expectation of study effect diminishes the treatment effect size. However, when those treatments are compared against psychological placebo or minimally active controls, significant effects either diminish in magnitude or disappear. This finding is relevant for indirect comparisons with pharmacological trials, in which the use of placebo makes the effect size appear smaller. The much greater difficulty of blinding treatment assignment in psychosocial trials is also to be taken into account. The risk of inflated effect sizes due to weak and methodologically flawed comparators (e.g., waiting list, no intervention) is that such interventions might be preferred to other superior treatments, delaying response and remission ¹²¹ .

The results from this umbrella review should be considered within its limitations. First, we only considered evidence that was evaluated quantitatively via MAs/NMAs. This approach has excluded data from RCTs that have not (yet) been meta‐analyzed. In particular, Internet‐based psychosocial interventions, whose development has been recent and which may be particularly favored by youth125, 126, have not been sufficiently covered.

Second, we focused mainly on efficacy outcomes, while choices need to be made considering both efficacy and tolerability/safety. However, we included all‐cause discontinuation as a global acceptability measure, as well as discontinuation due to intolerability as a core tolerability outcome, because these two events are typically measured and reported across both pharmacological and non‐pharmacological treatment modalities. Detailed tolerability outcomes of pharmacological interventions in youth with mental disorders, that can be used to complement the present work on efficacy, have been recently published in this journal ¹⁴ . Such detailed data are not generally reported for psychosocial interventions, which is currently a major unmet need ¹²⁷ .

Third, as mentioned above, most meta‐analytic evidence concerns the acute and short‐term treatment effects, and much more data are required regarding the efficacy and safety of long‐term and relapse prevention interventions for mental disorders in youth. Fourth, most evidence is available for monotherapy and vs. placebo/no treatment, although combination and augmentation treatments across and within pharmacological and psychosocial treatment modalities are commonly used in clinical practice, in youth as well as in adults ¹²⁸ . Fifth, although 14 of the 104 included meta‐analyses were NMAs that allow for direct and indirect head‐to‐head comparisons, most data were not derived from direct comparisons of active treatments, limiting the confidence with which comparative treatment choices can be made.

Sixth, since design, population and illness characteristics, as well as choice of control groups and blinding methods influence effect sizes, and these characteristics often differ substantially between pharmacological and non‐pharmacological trials, indirect comparisons of effect sizes across these treatment modalities need to be interpreted with caution. To overcome this limitation, more head‐to‐head comparisons and combination trials need to be conducted both within and across treatment modalities. Finally, we focused on those disorders that are most common and studied in youth, maximizing the chance of finding meta‐analytic evidence, but other mental conditions could also be of interest.

Despite these limitations, inherent in the umbrella review methodology and available RCT data, this study provides the most comprehensive account of the available RCT evidence concerning pharmacological, psychosocial and brain stimulation interventions for the main psychiatric disorders in childhood and adolescents. The large body of literature reviewed here can inform future research aimed at addressing identified gaps, as well as current clinical care and guidelines regarding the choice of interventions for mental health conditions in youth, merging state‐of‐the‐art efficacy and acceptability data with information on tolerability and safety.