People with ADHD often receive pharmacological treatments. Understanding the evidence regarding the effectiveness and the choice of medications and likely situations where caution might need to be exercised are important considerations for clinicians in the comprehensive treatment and support of people with ADHD.
An evidence review (update of NICE 2018) was conducted to explore what principles clinicians should follow when discussing decisions to start, adjust, or discontinue pharmacological treatment for people with ADHD Whilst new evidence was found, it was not integrated because the NICE analysis was already deemed to have reached saturation of thematic data. NICE identified 69 studies and conducted a qualitative review of the views of people with a lived experience of ADHD and their families. Saturation in themes was reached after five themes were identified. Clinicians should be aware of these themes in order to improve outcomes for people with ADHD and adequately support them through the pharmacological treatment process (see below).
Methylphenidate versus placebo
No new evidence was found. NICE identified 2 low-quality studies (Ghuman et al., 2009; Greenhill et al., 2006). A clinically important benefit of methylphenidate for parent-teacher composite rated ADHD total symptoms total and other symptoms was found.
Risperidone versus methylphenidate
Risperidone is not recognised as a treatment for ADHD but is included here due to a clinically significant finding. No new evidence was found. NICE identified one study of very low quality (Arabgol, Panaghi, & Nikzad, 2015). There was no clinical difference between risperidone and methylphenidate on parent-rated ADHD total, inattentive and hyperactivity symptoms. The number of children discontinuing their medication due to adverse events was lower for risperidone compared to methylphenidate, and this was clinically important.
Risperidone and methylphenidate versus methylphenidate
As noted above, risperidone is not generally used to treat ADHD symptoms but has been included here due to clinically significant adverse events. No new evidence was found. NICE identified one study with low- to very low-quality evidence (Safavi, Dehkordi, & Ghasemi, 2016). There was no clinical difference in parent-reported ADHD total, inattention and hyperactivity symptoms and other symptoms. There was a clinically important benefit of methylphenidate and risperidone combined on the Clinical Global Impression scale. There was clinically important harm of risperidone and methylphenidate combined on the outcome measure of discontinuation due to adverse events.
Immediate-release methylphenidate versus placebo
New evidence was identified from one RCT of moderate certainty (Solleveld et al., 2020). This study found statistically significant benefits of immediate-release methylphenidate over placebo for Clinical Global Impression Scale change score and the Disruptive Behaviour Disorder Rating Scale (for attention scores but not hyperactivity scores).
NICE evidence previously identified eight studies of low to moderate quality. There was a clinically important benefit of methylphenidate over placebo for total ADHD symptoms (parent-rated; 2 studies low quality) (teacher rated; 2 studies low quality; 3 studies moderate quality) (teacher rated; 1 study moderate quality), inattention symptoms (parent-rated; 1 study moderate quality) (teacher rated; 1 study moderate quality), hyperactivity symptoms (teacher-rated, 2 studies low to moderate quality; parent-rated 2 studies), a clinical global impression (3 studies moderate quality), and other symptoms (2 studies low quality).
There was no clinical difference for ADHD hyperactivity symptoms (parent-rated at 16 weeks; 1 study low quality), discontinuation due to adverse events (4 studies low quality) and serious adverse events (1 study moderate quality).
Osmotic-controlled Release Oral System (OROS) methylphenidate versus placebo
New evidence was identified in one RCT of low certainty (Newcorn et al., 2017). There were statistically significant benefits of flexible-dose or fixed-dose OROS methylphenidate over placebo for ADHD total, inattention, hyperactivity symptoms, and Clinical Global Impression scale. NICE previously identified four studies.
There was a clinically important benefit of methylphenidate for parent-, teacher- and investigator-rated ADHD total, inattention, and hyperactivity symptoms (4 studies moderate quality), Clinical Global Impression scale (2 studies moderate quality), other symptoms (one study of low quality), quality of life (one study of low quality) and academic achievement (one study of low quality). There was no clinical difference in the number of children discontinuing their medication due to adverse events (3 studies of low quality).
Immediate-release methylphenidate versus OROS methylphenidate
No new evidence was found. NICE previously identified one study. There was no clinically important difference for ADHD inattention and hyperactivity symptoms (teacher-rated; one study of moderate quality) (parent-rated; one study of moderate quality), Clinical Global Impressions Scale (one study of low quality) and discontinuation due to adverse events (one study of low quality).
OROS methylphenidate versus lisdexamfetamine
New evidence was found in two RCTs in one study (Newcorn et al., 2017). There were statistically significant benefits of fixed-dose lisdexamfetamine over fixed-dose OROS methylphenidate for ADHD total, inattention and hyperactivity symptoms, and on the Clinical Global Impressions Scale (one RCT with low-certainty evidence).
There were no statistically significant differences between flexible-dose lisdexamfetamine and OROS methylphenidate for ADHD total, inattention, or hyperactivity symptoms and on the Clinical Global Impressions Scale (one RCT with low-certainty evidence). NICE previously identified one study. There was a clinically important benefit of lisdexamfetamine for investigator-rated ADHD total symptoms (one study of moderate quality) and clinical global impressions (one study of low quality). There was no clinical difference for discontinuation due to adverse events, academic achievement and other symptoms (one study of low quality).
In addition to the comparisons above, there were also comparisons between methylphenidate and dextromethorphan and piracetam which were not considered to be clinically relevant for the treatment of ADHD (see Technical Report).
Lisdexamfetamine versus placebo
New evidence was identified in 2 RCTs, reported in one article (Newcorn et al., 2017). There were statistically significant benefits of lisdexamfetamine 30 mg, 50 mg and 70 mg for ADHD total, inattention and hyperactivity symptoms (investigator-rated; one RCT, very low-certainty evidence), and Clinical Global Impression Improvement scale (investigator rated; one RCT, very low certainty) compared with placebo.
There were statistically significant benefits of flexible-dose and fixed-dose lisdexamfetamine over placebo for ADHD total, inattention, hyperactivity symptoms and on the Clinical Global Impressions scale (one RCT with low-certainty evidence). NICE previously identified one RCT. There was a clinically important benefit of lisdexamfetamine for ADHD total symptoms (investigator rated; one study of moderate quality), Clinical Global Impression scale, academic achievement and other symptoms (one study of moderate quality). There was no clinical difference for discontinuation due to adverse events (2 studies of very low quality).
Atomoxetine versus placebo
No new evidence was found. NICE previously identified 26 studies. There was a clinically important benefit of atomoxetine for: quality of life (2 studies of moderate quality, one study of low quality), treatment response (2 studies of low quality), ADHD total symptoms (investigator-rated; 3 studies of low quality and 6 studies of moderate quality) (teacher-rated; 5 studies of moderate quality, one study of low quality) (parent-rated; 9 studies of high quality, 2 studies of low quality, 3 studies of moderate quality), ADHD inattention symptoms (investigator rated; 5 studies of low quality) (teacher-rated; 5 studies of low quality) (parent-rated; 9 studies of low quality at four to12 weeks; 2 studies low quality at four weeks; 3 studies moderate quality), ADHD hyperactivity symptoms (investigator-rated; five studies of moderate quality) (teacher-rated; 4 studies of moderate quality, one study of low quality) (parent rated; 12 studies of moderate quality, 2 studies of very low quality), Clinical Global Impression (5 studies of moderate quality) and other outcomes (2 studies low quality).
There was no clinical difference for other symptoms (3 studies of moderate quality), academic achievement (one study of low quality), discontinuation due to adverse events (16 studies of moderate quality and 2 studies of low quality), or serious adverse events (3 studies of low quality).
Atomoxetine versus methylphenidate
New evidence was found from one study (Zhu, Sun, Zhang, Liu, & Zhao, 2017). NICE previously identified 3 RCTs. Integrated evidence showed there were statistically significant benefits of methylphenidate over atomoxetine for ADHD total and inattention symptoms (3 RCTs with moderate-certainty evidence), hyperactivity symptoms (one RCT with low-certainty evidence) and Clinical Global Impression scale (one RCT with low-certainty evidence).
There were no clinical differences for quality of life (one study of moderate quality), hyperactivity symptoms (parent-rated; 3 RCTs of moderate quality), other symptoms (one study of moderate quality) or the Conners’ scale measures of learning problems, confrontation, and ADHD index (one RCT with low-certainty evidence). More children in atomoxetine treatment groups discontinued due to adverse events, compared with methylphenidate treatment groups (2 RCTs of moderate quality). There were no statistically significant differences between methylphenidate over atomoxetine for ADHD total, inattention, or hyperactivity symptoms, or on the Clinical Global Impression scale (one RCT with low-certainty evidence).
Atomoxetine versus guanfacine extended release
No new evidence was found. NICE previously identified one low-quality study. There was a clinically important benefit of guanfacine for investigator-rated ADHD total symptoms and Clinical Global Impression scale. There was no clinically important difference in the number of children discontinuing due to adverse events.
Guanfacine versus placebo
No new evidence was found. NICE previously identified one RCT. There was a clinically important benefit of guanfacine for ADHD total and hyperactivity symptoms (investigator-rated; one study of moderate quality) and the Clinical Global Impression scale (one study of high quality). There was no clinically important difference for ADHD inattention symptoms (investigator-rated; one study of moderate quality).
Extended-release guanfacine versus placebo
No new evidence was found. NICE previously identified 8 RCTs. There was a clinically important benefit of extended-release guanfacine for ADHD total symptoms (investigator-rated; 7 studies of low quality), ADHD inattention symptoms (investigator-rated; 4 studies of low quality), ADHD hyperactivity symptoms (investigator-rated; 5 studies of high to moderate quality) and Clinical Global Impression scale (5 studies of moderate quality).
There was clinically important harm of extended-release guanfacine for serious adverse events (one study of very low quality): one participant in the guanfacine arm had a serious adverse event, compared with zero in the placebo arm. There was no clinically important difference for academic outcomes (one study of high quality) and discontinuation due to adverse events (8 studies of high quality).
Clonidine versus placebo
No new evidence was found. NICE previously identified 4 RCTs. There was a clinically important benefit of clonidine for the following outcome measures: ADHD total symptoms – parent-rated (2 studies of low quality), teacher-rated (2 studies of low quality), and investigator-rated (one study of low quality) ADHD inattention symptoms – investigator-rated (one study of low quality); hyperactivity symptoms – investigator-rated (one study of low quality) and parent-/teacher-rated (one study high quality)]; other symptoms (2 studies of very low quality).
There was no clinical difference for discontinuation due to adverse events (2 studies of moderate quality) or serious adverse events (one study of high quality).
Clonidine versus methylphenidate
No new evidence was found. NICE previously identified one RCT. The only evidence identified was for ADHD total symptoms, discontinuation due to adverse events and other symptoms, as measured by the Children’s Global Assessment Scale. There was a clinically important benefit of methylphenidate for ADHD total symptoms: teacher-rated (one study of very low quality and parent-rated (one study of very low quality). There was no clinical difference for other symptoms (one study low quality) or in discontinuation rates due to adverse events (one study of very low quality).
Clonidine versus desipramine
No new evidence was found. NICE previously identified one RCT. The only evidence identified was for ADHD hyperactivity symptoms. There was a clinically important benefit of desipramine for ADHD hyperactivity symptoms (parent-/teacher-rated; one study of high quality).
Clonidine versus carbamazepine
No new evidence was found. NICE previously identified one RCT. The only evidence identified was for ADHD symptoms. There was a clinically important benefit of clonidine for ADHD inattention symptoms (investigator-rated; one study of very low quality), ADHD hyperactivity symptoms (investigator-rated; one study of low quality) and ADHD impulsivity symptoms (investigator-rated; one study of low quality).
Immediate-release methylphenidate versus placebo
No new evidence was found. NICE previously identified 8 RCTs. There was no evidence identified for quality of life or serious adverse events. There was a clinically important benefit of methylphenidate for ADHD total symptoms (investigator-rated; 3 studies of very low to moderate quality), treatment response (2 studies of low quality) and Clinical Global Impression scale (2 studies of moderate quality). There were clinically important harms of methylphenidate for discontinuation due to adverse events (2 studies of high quality). There was no clinical difference for other symptoms (2 studies of moderate quality).
Osmotic-controlled Release Oral System (OROS) methylphenidate versus placebo
No new evidence was found. NICE previously identified 12 RCTs. There was no evidence for serious adverse events. There was a clinically important benefit of methylphenidate for the following outcome measures: treatment response (3 studies of moderate quality); ADHD total symptoms – investigator-rated (4 studies of low quality; 2 studies of moderate quality), and self-rated (2 studies of moderate quality, 2 studies of low quality); ADHD inattention symptoms – investigator-rated (2 studies of low quality, 2 of moderate quality) and self-rated (one study of moderate quality); ADHD hyperactivity symptoms (investigator rated; 2 studies of low quality); Clinical Global Impression scale (3 studies of moderate quality); other symptoms (one study of high quality); emotional dysregulation (one study of moderate quality).
There was no clinical difference for ADHD inattention symptoms (investigator-rated; 2 studies of moderate quality), ADHD hyperactivity symptoms (investigator-rated; 2 studies low quality and self-rated; one study of moderate quality), and emotional dysregulation (one study of very low quality). There was a clinically important harm of methylphenidate for discontinuation due to adverse events (9 studies of high quality) or quality of life (one study of high quality).
Immediate-release methylphenidate versus OROS methylphenidate
No new evidence was found. NICE previously identified one study. There was no clinically important difference for ADHD inattention and hyperactivity symptoms (teacher-rated; one study of moderate quality) (parent-rated; one study of moderate quality), Clinical Global Impressions Scale (one study of low quality) and discontinuation due to adverse events (one study of low quality).
OROS methylphenidate versus lisdexamfetamine
New evidence was found in two RCTs in one study (Newcorn et al., 2017). There were statistically significant benefits of fixed-dose lisdexamfetamine over fixed-dose OROS methylphenidate for ADHD total, inattention and hyperactivity symptoms, and on the Clinical Global Impressions Scale (one RCT with low-certainty evidence).
There were no statistically significant differences between flexible-dose lisdexamfetamine and OROS methylphenidate for ADHD total, inattention, or hyperactivity symptoms and on the Clinical Global Impressions Scale (one RCT with low-certainty evidence). NICE previously identified one study. There was a clinically important benefit of lisdexamfetamine for investigator-rated ADHD total symptoms (one study of moderate quality) and clinical global impressions (one study of low quality). There was no clinical difference for discontinuation due to adverse events, academic achievement and other symptoms (one study of low quality).
In addition to the comparisons above, there were also comparisons between methylphenidate and dextromethorphan and piracetam which were not considered to be clinically relevant for the treatment of ADHD (see Technical Report).
Dexamfetamine versus placebo
No new evidence was found. NICE previously identified 3 RCTs. There was a clinically important benefit of dexamfetamine for ADHD total, inattention and hyperactivity symptoms (investigator-rated; 2 studies of moderate quality) and for Clinical Global Impression (one study of moderate quality).
Lisdexamfetamine versus placebo
New evidence was found in one RCT (Weisler et al., 2017) which was a post-hoc analysis from the Adler et al. (2013) study, reporting outcomes as least squares mean difference. There were statistically significant benefits of lisdexamfetamine dimesylate over placebo for all outcomes reported: BRIEF-A global executive composite, behavioural regulation index, and metacognition index, and Conners’ adult rating scale ADHD index, hyperactivity, inattention, impulsivity, and problems with self-concept; all investigator-rated; one RCT with moderate-certainty evidence.
NICE previously identified 3 RCTs. No evidence was identified for serious adverse events. There was a clinically important benefit of lisdexamfetamine for ADHD total symptoms (investigator-rated; 3 studies of moderate quality), ADHD inattention symptoms (investigator-rated; one study of low quality), ADHD hyperactivity symptoms (investigator-rated; one study of low quality), Clinical Global Impression (one study of moderate quality) and other symptoms (one study of low quality). There was no clinical difference for quality of life (one study of very low quality) or discontinuation due to adverse events (3 studies of very low quality).
Atomoxetine versus placebo
No new evidence was found. NICE previously identified 10 RCTs. There was a clinically important benefit of atomoxetine for the following outcome measures: quality of life (5 studies of low to moderate quality); ADHD total symptoms – investigator-rated (10 studies of low to very low quality) and self-rated (2 studies of low quality); ADHD inattention symptoms – self-rated (2 studies of low quality) and investigator-rated (9 studies of low to very low quality); ADHD hyperactivity symptoms – investigator-rated (9 studies of very low quality) and self-rated (2 studies of moderate quality).
There was a clinically important harm of atomoxetine for discontinuation due to adverse events at 24 weeks (one study moderate quality). There was no clinical difference for other symptoms (2 studies of low quality) or discontinuation due to adverse events up to 14 weeks (7 studies of moderate quality).
Guanfacine versus placebo
New evidence was found in one RCT (Iwanami, Saito, Fujiwara, Okutsu, & Ichikawa, 2020) which reported outcomes as least squares mean difference. There were statistically significant benefits of extended-release guanfacine over placebo for ADHD total, inattention and hyperactivity symptoms; executive functioning (BREIF) for inhibit, initiate, and plan/organise, and Global Executive Composite index (investigator-rated; one RCT with moderate-certainty evidence). There was a statistically significant benefit of placebo over extended-release guanfacine for quality of life (productivity).
There were no statistically significant differences between extended-release guanfacine and placebo for quality of life total, psychological health, life outlook and relationships (one RCT of low certainty); executive function (BRIEF-A) for shift, emotional control, self-monitor, behavioural regulation, working memory, task monitor, and organisation of materials; metacognition index or adverse events (one RCT with low- to moderate-certainty evidence).
NICE previously identified one RCT (Taylor & Russo, 2001) of moderate quality. There was a clinically important benefit of guanfacine for investigator-rated ADHD total, inattention and hyperactivity symptoms.
Guanfacine versus dexamfetamine
No new evidence was found. NICE previously identified one RCT. There was no clinical difference in ADHD total, inattention or hyperactivity symptoms (investigator-rated; one study of low to moderate quality).
There is a paucity of evidence for the effectiveness of medications in children under 5 years of age. As such, no recommendation about medication use is made. Instead, we recommend that an expert in child development and treating ADHD in young children be involved in assessment and treatment decisions. The evidence showed that:
This was supported by the GDG’s experience that stimulants have a more rapid onset of therapeutic effect than non-stimulant agents such as atomoxetine and guanfacine.
The GDG considered the evidence, their experience and Australian prescribing regulations to recommend methylphenidate or dexamfetamine/ lisdexamfetamine as a treatment for children aged 5 years and over, adolescents and adults, given the minimal difference in efficacy and tolerability in these agents.
The GDG debated whether to recommend methylphenidate for children and adolescents and dexamfetamine or lisdexamfetamine for adults, as the first-line treatment.
The GDG concluded that the evidence was not certain enough to support this more restrictive recommendation. The GDG discussed the clinical reasons for starting either short- or long-acting medications and provided clinical guidance on this via a practice point.
If short-acting stimulants are effective and well tolerated but a longer-acting preparation is more convenient or is preferred, lisdexamfetamine or long-acting methylphenidate could be offered. It was noted that some people have intolerances such as gluten, which might influence first-line medication choice, given some short-acting stimulant preparations contain gluten.
The GDG agreed that, if stimulants cannot be tolerated or are ineffective, atomoxetine or guanfacine should be offered. If these are not tolerated or effective, other medications could then be trialled. Atomoxetine and guanfacine were the non-stimulant drugs with the most convincing evidence.
The GDG acknowledged that there was very little evidence on medication choice for people with ADHD and most co-occurring conditions. The GDG agreed that neither the available evidence nor their experience justified a different choice of ADHD medication for people with ADHD and coexisting conditions, but there should be careful baseline assessments and consideration of drug interactions, slower titration and more careful monitoring of adverse effects, and regular contact. The GDG noted that, rarely, stimulant medications can induce psychosis and recommended that ADHD medication should be stopped in people experiencing a psychotic episode.
Recommendations were based on the evidence review, the GDG’s expertise and clinical experience, and adaptation of the NICE recommendations to the Australian context. The recommendations were also informed by a systematic review that met the selection criteria and provided data from network meta-analyses (Cortese et al., 2018), allowing us to more definitively comment on medication choice and sequence across age groups.
Of the newly identified studies, several evaluated medications are not available in Australia. These included clinical trials of mixed amphetamine salts, methylphenidate plus dextromethorphan, methylphenidate plus piracetam, dasotraline, and viloxazine. These were not included in the recommendations.
Pharmaceutical Benefits Scheme (PBS) restrictions for subsidisation of ADHD treatments differ according to the age at which the person received the diagnosis.
Guanfacine and atomoxetine is subsidised only for those with a diagnosis between the ages of 6 and 17 years, while subsidy for long-acting methylphenidate, lisdexamfetamine and atomoxetine is restricted to those with a diagnosis between the ages of 6 and 18 years (retrospective diagnosis permitted for lisdexamfetamine).
Age restrictions do not apply to PBS listings for dexamfetamine and for methylphenidate short-acting formulations. These restrictions may result in increased costs for people for whom ADHD was not diagnosed before the age of 18 years.
The GDG noted that prescribing laws in Australia differ between states and territories as noted by AADPA (AADPA, 2022). It is hoped that, over time, all jurisdictions will reach greater uniformity in prescribing laws that reflect best practice.
The current recommendations are based on evidence reviews and clinical consensus. It is noted that some current PBS restrictions may not fully align with the current recommendations. For example, PBS subsidy for long-acting methylphenidate is restricted to people who have already used short-acting methylphenidate.
Clinicians need to be aware of any cost implications of current PBS restrictions before selecting a stimulant type and duration of action and discuss these with the person with ADHD and their carers/parents/families. As noted in the principles section, prescribers need to be aware of any regulatory requirements when prescribing stimulants.
Adults will generally need to access a private psychiatrist as there are no publicly funded services for adults with ADHD. This can result in significant out-of-pocket costs and also significant delays, due to limited access to specialist adult ADHD psychiatrists, or people not receiving treatment. Children and adolescents may also face significant delays in accessing publicly funded paediatricians and child/adolescent psychiatrists, and may instead access clinicians in the private sector, resulting in significant out-of-pocket costs. Workforce development would ensure that health inequity impacts are minimised.
