The study selection process is shown in Fig. 1. Reasons for excluding trials after full-text assessment are provided in Supplementary Table S2. In total, 48 studies published between 1987 and 2017 met the inclusion criteria, of which 6 were qualitative, 2 mixed methods and 40 quantitative. The quantitative studies were all based on surveys and questionnaires with the exception of one free-listing exercise; while the qualitative studies were based on interviews (n = 4), focus groups (n = 2), no observational studies were identified. Characteristics of each study and their review themes are given in Table 1.
A range of countries were represented with most of the studies originating from the US (23 studies), UK (eight), Australia (3), Canada (4), Netherlands (2), South Africa (2) and 1 each from Iran, Brazil, Finland, France, Pakistan, Switzerland and Singapore.
Data methodological quality
Results of study quality are reported in Table 1. Following Kmet, Lee and Cook’s guidelines , an original quality score from 0 to 1 was calculated for each study. Scores were then classified into low (0–0.44), moderate (0.45–0.69) and high (0.70–1.00). Study quality was assessed by two reviewers (BF and DD). Agreement between reviewers was 88% overall (92% for quantitative studies, 85% for qualitative studies).
The studies showed some variation in their quality. The average quality score was 0.73 with 36 studies receiving a high-quality rating, 10 a moderate rating and 2 a low rating, (quantitative studies mean score of 0.75; qualitative studies mean score of 0.70). The two studies that received a low-quality rating were not used to inform our review results and conclusions.
Data extraction and summary of results
Need for education
The main theme highlighted by this review related to the need for education on ADHD. Thirty-seven papers identified issues related to a lack of education on ADHD, representing a wide range of countries, 14 from the US [30,31,32,33,34,35,36,37,38,39,40,41,42,43], 7 from the UK [3, 11, 16, 44,45,46,47], 3 from Australia [48,49,50], 2 from the Netherlands [51, 52], 2 from South Africa [53, 54], 2 from Canada [55, 56] and 1 each from France , Singapore , Pakistan , Finland , Brazil , Iran  and Switzerland . These papers highlighted both a lack of initial training, as well as inadequate training on ADHD. They also highlighted a lack of awareness, experience, understanding and knowledge of ADHD. Most PCPs also reported a lack of confidence about ADHD and in treating ADHD. These factors often hindered general knowledge and understanding of ADHD.
Lack of training on ADHD was recorded by many studies. This included studies reporting a lack of training especially at undergraduate and postgraduate medical education levels [3, 16, 30, 38, 45, 53, 57] with studies suggesting that 1%  to 28%  of PCPs received specific training about ADHD. In a quantitative study from the UK, Ball  found that only 6% of 150 PCPs surveyed received formal training on ADHD and 80% reported wanting further training. This was more prominent for adult ADHD with two studies reporting a greater lack of education [30, 53]. Even when studies reported training on ADHD, the training was often considered by PCPs to be inadequate [11, 31, 59] with up to two-third of PCPs feeling inadequately trained to evaluate children with ADHD . While a strong need for training on ADHD in general was observed, three studies also highlighted the importance of updated training incorporating new knowledge [37, 39, 42]. This lack of education affected many aspects of the primary care experience of patients from referral and diagnosis [44, 59] to management  of ADHD.
The need for education was also highlighted through the lack of general awareness, experience, understanding and accurate knowledge of ADHD. While two studies directly reported a lack of knowledge and experience [36, 52] as a barrier to dealing with people with ADHD, eight investigated these concepts through knowledge of the DSM criteria or clinical guidelines [31, 33, 34, 47, 64,65,66,67]. Accurate knowledge of guidelines and procedures for identification of ADHD were low; for instance, only 20%  —27%  of PCPs were using DSM criteria and only 20% were using official guidelines . One study from the UK , using a mixed-method approach, found that 75% of PCPs could not identify ADHD DSM criteria correctly and all PCPs were unsure of ADHD prevalence and diagnostic procedures. Two studies investigated these concepts through questions about several factors including treatments, prevalence and symptoms, reporting different levels of knowledge and awareness throughout, with inaccurate beliefs such as believing that there is no need for treatment  or that ADHD was not a medical problem . Additionally, very few studies reported a majority of PCPs in their sample being able to accurately identify ADHD characteristics. Two studies reported that PCPs did not know what the acronym TDHA (ADHD in French) stood for  or that they did not know about ADHD even after reading its definition  demonstrating that a lack of general knowledge about ADHD was highly evident.
The final aspect relating to the need for education highlighted PCPs’ lack of confidence about ADHD, most specifically in treating ADHD. This review and the included studies focus principally on the recognition of ADHD but a few studies also raised the issue of a lack of confidence, encompassing treatment and management. In these studies, it is unclear whether the lack of confidence is solely around recognition or all aspects of ADHD management; therefore, it was important to include this aspect in our findings. While two studies reported a general lack of confidence [30, 49], three [36, 55, 63] reported low confidence and competence in diagnosis and management of ADHD. Some studies reported a lack of confidence toward treatments, with PCPs reporting being uncomfortable with medication for ADHD. Goodman et al.  reported that 38% of PCPs had no confidence in treating ADHD, Alder et al.  also highlighted a considerable lack of confidence in treating adults with ADHD and Ball  reported that 11% of PCPs were not willing to prescribe medication at all due to lack of knowledge, while 88% of PCPs wanted further training in the drug treatment of ADHD.
While the need for education underpinned many barriers and issues towards the overall understanding and knowledge of ADHD, a few positive outcomes were also observed. In contrast to our overall findings where a lack of knowledge and confidence was evident, three studies reported above-average ratings of confidence and high knowledge of ADHD [41, 42, 60] and Evink et al.  reported that all paediatricians in their studies used DSM criteria. The reasons for these different findings are unclear but could be due to the high number of paediatricians in the studies who might have received specialist paediatric training.
Despite the overall lack of training and awareness of ADHD, studies reported that PCPs had a keen interest in wanting to gain more knowledge , with strategies such as asking colleagues, self-education, and online enquiries. [16, 40, 53, 57, 62]. Two studies [52, 56] explored the benefits of educational programs for PCPs and reported an increased awareness and confidence in ADHD after taking part; the first study  focused on an educational program for prescription and monitoring of ADHD medication, while Ward et al.  evaluated a 1-day course which aimed to teach PCPs to manage ADHD and observed a significant difference between pre-test and post-test knowledge. Furthermore, Wolraich et al.  reported a marked increase in the use of APA guidelines between 1999 and 2005 by PCPs (13%–50%), suggesting an increased interest and awareness in ADHD.
Misconceptions and stigmas
Linking to the previous theme, misconceptions and stigmatisation surrounding ADHD were often strongly present in the literature. This notion was explored by different studies, either directly reporting the experience of stigma within primary care or reporting inaccurate facts about ADHD; reflecting gender biases (‘it only happens in boys’), misleading causes of ADHD (due to high sugar level or lead poisoning) or most prominently, that ADHD is primarily caused by bad parenting. Seventeen studies discussed elements related to misconceptions, five from the US [36, 38, 67,68,69], four from the UK [3, 11, 46, 47], two from Australia [49, 50], and one each from the Netherlands , Iran , Brazil , Singapore , France  and South Africa .
Most studies reported general misconceptions about ADHD. In a mixed-method study in the UK, Salt et al.  reported that over 50% of PCPs agreed on the controversial nature, the strong stigmatisation of ADHD and the disadvantages the diagnosis brought. In a quantitative survey of 380 US PCPs, Kwasman et al.  reported strong misconceptions about ADHD, including: ADHD is “caused by poor diet” (21% agreed), “the child does it on purpose” (15%), “medications can cure ADHD”(10%) and “ADHD medications are addictive” (48%). Many studies reported participant views that sugar levels were a cause of ADHD [58, 62, 69] while others reported a gender misconception that ADHD was only present in boys [38, 68]. Other misconceptions were more surprising with Ghanizadeh and Zarai ; for instance, reporting that 82% of PCPs believed children with ADHD misbehaved primarily because they do not want to obey rules and do their assignments, while Quiviger and Caci  stated that 24% of the PCPs surveyed thought it was a disorder constructed abroad and imported into France.
While it could be expected that PCPs should not hold stigma towards ADHD due to their expected knowledge of the disorder, in a quantitative study in the Netherlands on stigmatisation towards ADHD, Fuermaier et al.  reported no difference in stigmatisation levels between physicians and a control group of non-medical professionals.
Ten studies reported that PCPs believed ADHD was due to bad parenting [11, 36, 46, 47, 49, 50, 54, 62], numbers varying from 15%  to over half [36, 62] of PCPs surveyed believing that dysfunctional families were predominately to blame for ADHD symptom expression. In semi-structured interviews with UK PCPs and parents, Klasen and Goodman  reported that most GPs saw symptoms of hyperactivity (one of the three symptom clusters of ADHD) as an effect of dysfunctional families and many felt that parents’ views of hyperactivity as a medical problem were an attempt to avoid dealing with possible shortcomings in their parenting practices.
Four studies linked the presence of misconceptions with negative media coverage and the use of labels. Klasen and Goodman , for instance, reported that parents felt PCPs were against labels, trying to normalise hyperactive behaviours, while Klasen  reported that 25% of PCPs felt labelling was not useful. Salt et al.  highlighted, through a targeted questionnaire, the influence of the media in the general public’s conception of ADHD, whereas Shaw et al.  argued that negative media coverage and labels affect the representation of medication and had led to labelling bad parenting as ADHD.
Although very few facilitators can be observed within this theme, it is important to note that the concepts of misconception and stigma were only explored in a third of included studies suggesting stigma about ADHD did not emerge from studies as much as might have been anticipated. Studies identifying stigma reported stigma from a wide variety of different countries and cultures, suggesting that stigma surrounding ADHD is not specifically culturally determined.
Internal and resource constraints with recognition, management and treatment
As PCPs are often responsible for the recognition and management of ADHD, a few barriers were observed surrounding these procedures. The first considered the barriers experienced around recognition, referral and diagnosis of ADHD, mainly referring to resource constraints such as time and the need for appropriate screening tools. With regard to treatment options, the main barriers observed included the limited treatment options available and uneasiness around medications. Twenty studies discussed aspects related to recognition and treatment, with eleven studies from the US [30, 31, 34,35,36,37, 41, 66, 69,70,71], three from the UK [3, 44, 46], two from Australia [49, 50] and one each from Brazil , the Netherlands , France  and South Africa .
The resource constraint mainly experienced in the recognition and management of ADHD is in relation to time and the complexity of ADHD. Many studies found that the time necessary to gain all the relevant information was often too demanding [31, 34, 35, 49, 50, 52, 53, 69] especially taking into consideration the complex nature of ADHD [34, 36, 41]. After interviewing 19 PCPs in focus groups in the US, Guevara et al.  reported that limited resources and lack of time to communicate with schools led to limited access to care, whilst Klasen and Goodman  found in their interviews in the UK that information necessary for management and recognition is often conflicting and ambiguous. Five other studies mentioned the need for better assessment tools [3, 35, 54, 71], especially for adults . Finally, one other barrier experienced in relation to time and resource constraints is that PCPs make decisions on assessment for referral based on the child’s behaviour in their office [33, 57, 65], which can lead to potential misdiagnosis as the child might behave very differently at home or school .
Considerable issues were also highlighted around treatment: the lack of available treatment options as well as uneasiness around medication. While studies reported a general uneasiness with ADHD medication [36, 38, 41, 50], this at times led to resistance or refusal to grant prescriptions by PCPs [16, 52, 66]. In a series of interviews with 128 PCPs, Gomes et al.  reported high levels of uneasiness around medication, limited knowledge of treatment options, and a lack of knowledge of the pros and cons of medication and other treatments. This reflected other findings describing confusion around treatment options in relation to professionals’ knowledge of what is available and limited availability of treatment [3, 36, 70].
Despite the constraints explored in these studies, attempts to address these issues were reported in only two studies. After participating in a 1-h educational online course on ADHD medication, Hassink-Franke et al.  found that most PCPs felt more confident and competent about prescribing and monitoring medication. As this was a qualitative study, information was not available on the degree of change of confidence in the participating PCPs. Ward et al.  evaluated a 1-day course which aimed to help PCPs to manage ADHD. Results demonstrated some impact on practice in the form of increased levels of ADHD referrals. However, the study was based on only 34 clinicians, was not controlled and did not verify the appropriateness of referrals.
Multidisciplinary approach: the role of other specialists, teachers, parents and patients
The final theme encompassed the concepts of a multidisciplinary approach. This mainly referred to the role of different specialists and the importance of shared care. However, it also included the role of other parties involved such as the people with ADHD, parents and teachers. Twenty-two studies explored issues pertaining to a multidisciplinary approach, twelve from the US [31, 36,37,38,39,40,41, 64, 66, 69, 70, 73], five from the UK [3, 43, 454, 12, 15], two from Australia [49, 50], two from South Africa [53, 54] and one from the Netherlands .
When discussing the concept of the multidisciplinary approach, many studies explored the communication between specialists, principally between primary and secondary care. With the belief that integrated care pathways and a collaborative approach is essential [15, 40, 44, 64], issues with communication between specialists was expressed as a major barrier [36, 39, 49, 50, 69]. In semi-structured interviews in the US , PCPs reported the importance of involvement of other stakeholders, psychiatrists, and schools in decision-making and over half of the professionals interviewed mentioned difficulties in communicating with other specialists. Furthermore, Ross  reported that only 15% of PCPs surveyed received communication from psychiatrists. Guevara  found similar issues with communication and a need for shared care; however, this paper acknowledged the breakdown of communication between parents, schools and physicians but not from a lack of will or desire, rather as a ‘System failure’—lack of accountability, discontinuity of care, lack of support, limited resources and finger pointing.
Ambiguity about the role of different professionals [11, 40, 41, 52] was also noted as a barrier to access to care. Klasen and Goodman  highlighted that most PCPs were not aware of specialist help available in their area and were not certain of whom to refer to.
Communication with other parties such as schools, parents and people with ADHD themselves was also reported as being a barrier. Four studies, for instance, mentioned that patients failing to turn up for appointments limited the PCPs’ ability to assess them and provide the right care [36, 44, 45, 69]. Other studies found that PCPs experienced considerable difficulties in getting information from parents and schools [31, 39, 73] as well as reporting feeling continued pressure for diagnosis from schools and parents [53, 54, 64]. In a US survey of 723 PCPs, Rushton et al.  found that 55% felt a strong pressure from teachers to diagnose ADHD while 70% felt pressure to prescribe medication. Kwasman et al.  reported that their large sample of school nurses highlighted a lack of multidisciplinary communication between PCPs and school staff and suggested that PCPs and schools would benefit from a greater understanding of the contributing that each could make to an effective ADHD assessment.
An integrated pathway between primary care and secondary care may provide the optional solution for ADHD assessment. Hassink-Franke et al.  in their study of Dutch PCPs highlighted that greater support and more constructive long-term relationships with secondary care enhanced PCP’s confidence about ADHD. Greater support for Dutch PCPs also allowed families of children with ADHD to received care from PCPs with whom they had a long-lasting relationship and allowed care to be provided in a more informal primary care context rather than more formal secondary care.