Study designs of evaluations included in the review
Randomised controlled trials (RCTs). Non-RCT studies were included if they met all of the other inclusion criteria and if they evaluated adverse effects associated with treatment for more than 12 weeks and included more than 10 patients.
Specific interventions included in the review
Drug-to-Drug Comparisons: studies that included one of the following head-to-head comparisons:
1. Stimulants (methylphenidate, dextroamphetamine or pemoline) vs. stimulants (trials comparing the same drug were included if different formulations of different enantiomers).
2. Stimulants vs. tricyclic antidepressants.
3. Stimulants vs. clonidine, buproprion, or selective serotonin-reuptake inhibitors were eligible for inclusion.
Tricyclic Antidepressants vs. Placebo: Studies that compared placebo with amitriptyline, imipramine, or desipramine.
Drug vs. Nondrug Studies: studies in which:
1. One of the study arms included only a stimulant drug.
2. One or more of the control arms included other modes of intervention such as behaviour modification, dietary interventions, or other psychosocial interventions.
Combination Therapy: studies in which:
1. One of the study arms included two or more interventions given in combination.
2. One of the study arms included a stimulant.
Long-Term Therapy: studies that evaluated treatment for 12 weeks or more in all study arms.
Specific interventions included: amphetamine, attention control/child training, behaviour modification, bibliotherapy protocol, buproprion, caffeine, child training, chlorpromazine, clonidine, cognitive behavioural therapy, desipramine, dexamphetamine, EEG biofeedback, efamol, haloperidol, hydroxyzine, imipramine, L-amphetamine, lithium carbonate, lethylphenidate, nicotine, nothing (e.g. waiting lists), parent training, pemoline, phenylalanine, pinodol, placebo, racemic amphetamine, secobarbitol, selegiline, supportive therapy and thiordazine.
Participants included in the review
Both adults and children suffering from ADHD. For treatment of ADHD in adults, a study was included if treatments were evaluated in patients older than 18 years of age. Studies that included conditions other than ADHD were included only if they provided separate analyses for patients with ADHD.
Outcomes assessed in the review
Outcomes of interest extracted were: abdominal pain; achievement tests; aggressiveness; anorexia; anxiety; arithmetic; behaviour disturbances; CD; core symptoms; crying; decreased appetite; depression; disturbed sleep; emotional; fatigue/tiredness; global mood; global side effects; global symptoms; grades; growth; headache; height; hyperactivity; impulsivity; inattention; increased appetite; interaction with family; interaction with friends; irritability; listening; motor tics; nausea; nightmares; ODD; reading; performance; sadness; self-control; self-esteem; sleep disorder - insomnia; sleep disorder - sleep; social competence; somatic effects; spelling; verbal; weight; weight gain; and weight loss.
The following tests were used: ACTeRS; ADHD rating scale; Beck Depression Inventory; Behaviour Problem Checklist; Brief Psychiatric Rating Scale; CGI - Global Improvement; CGI - Severity of Illness; Child Psychological Rating; Children's Checking Task; Children's Depression Inventory; Clinical Global Improvement; Conners Child/Self; Conners Parent Rating Scale; Conners Teacher/Spouse; CPT; DETROIT13; DETROIT16; DETROIT6; DTS; Durrell Anal. Reading; EFT; Gates McGinty Test; Global Assessment Scale; HAM-A; Hamilton Psychiatric Rating - Depression; How I Feel; Humphries Self-control - Teacher; KBIT; MFFT; Other Clinician Report; Other Parent report; Other Teacher Report; Peterson Quay Checklist; PGS; Physician Global Assessment of Change; Piers Harris Scale; POMS; PTSR; Report Only; SNAP; STESS; STM Task; Unspecified reaction performance; Werry Weiss Activity Scale; Weschler; and WRAT.
How were decisions on the relevance of primary studies made?
Hard copies of studies were obtained and independently assessed by two members of the research team who then decided, by consensus, whether to include the study.