Abstract
Background Good quality primary care is essential for the assessment and treatment of headache but there is evidence that primary care for headache is suboptimal.
Aim To identify the international evidence on the assessment and treatment of headache in adults in primary care.
Design & setting A scoping review of the published literature following Preferred Reporting Items for Systematic reviews and Meta-Analyses extension for Scoping Reviews (PRISMA-SCR) guidelines, and a narrative review of the evidence.
Method An electronic search of MEDLINE and Embase (1974–2024) was undertaken. Studies meeting the eligibility criteria were included. Results were grouped by study type and were reported narratively.
Results In total, 1125 articles were screened, 43 articles underwent full-text review, and 28 articles were included in the final review. Six studies used comparative methods, of which n = 3/6 investigated educational interventions. The educational interventions found positive effects on learning, and on outcomes such as diagnosis rates, but the only randomised controlled trial (RCT) did not show any benefits of the intervention. Other comparative studies showed satisfaction with GP with an extended role (GPwER) headache services, benefits from direct access to magnetic resonance imaging (MRI), and benefits from a nurse-led headache service. Twenty-two studies used non-comparative methods, such as surveys and interviews, and investigated approaches to assessment, diagnosis, referral rationale, decision making for prescribing prophylactic medications, educational initiatives, direct access to neuroimaging, GPwER, and nurse-led interventions.
Conclusion Despite the availability of high quality clinical guidelines on the assessment and management of headache, the evidence shows that its implementation in primary care is problematic and educational interventions are a common focus of published studies. Further research is required to assess the quality of the current evidence and to develop, refine, and deploy interventions that have a signal of efficacy.
How this fits in
Headache is highly prevalent, and it is a major cause of disability. Primary care has an important role in the assessment and treatment of headache, and by some interpretations of the clinical guidelines, most cases of headache should be exclusively assessed and managed in primary care and referral to a specialist should be rare. To deliver good patient outcomes, the primary care workforce needs sufficient skills and capacity, but the evidence shows that many people with headaches receive suboptimal primary care with poor symptom control, under-use of key treatments, such as migraine prophylaxis and triptans, and high rates of medication overuse headache. There have not previously been any attempts to review the evidence on primary care headache management for effective interventions nor to identify areas that require more research. This scoping review addresses that knowledge gap.
Introduction
Headache affects around 47% of people globally1 and it is among the top 10 causes of disability, according to the World Health Organization (WHO).2 In the UK, migraine, which is the most common type of headache, affects 10 million people, that is, one in seven adults. Primary care is the first point of contact for people seeking medical care for headaches, accounting for one in 10 consultations.3 By some interpretations of clinical guidelines, most cases of headache should be managed exclusively in primary care and referral to a specialist should be rare. Waiting times for specialist clinics increased from 15–29 weeks between 2021 and 2023 in the NHS and are likely to be even higher now. Only 62% of integrated care systems (ICS) in England have a specialist headache clinic.4
Despite the importance of headaches to patients and to the health service, a 2014 report highlighted insufficient education and training resources for non-specialists in the NHS.5 Primary care clinicians sometimes struggle with diagnosis and treatment leading to suboptimal outcomes6–9 and they may benefit from extra education and training, new guidelines, and tools10,11 to improve the care they deliver, to improve the quality of specialist referrals,12 and to reduce unnecessary referrals.13
A recent review of educational initiatives highlighted the need for innovative, evidence-based methods for content delivery, knowledge assessment, and evaluation,14 with the aim of enhanced patient outcomes, and improved cost-effectiveness.15–17 Several studies and reports have explored optimal care pathways18 and innovations such as providing GPs with direct access to magnetic resonance imaging (MRI) scans.4,19 However, there have not been any attempts to review the evidence pertaining to primary care for people with headaches, to explore which topics are important for clinicians, to explore interventions and their effectiveness, and to identify areas that require more research. A scoping review is the ideal method to identify the extent and nature of a body of evidence, to identify gaps, and to guide future research and ultimately to improve patient care. Therefore, we set ourselves the aims of undertaking a scoping review of the published literature and producing a narrative review of the evidence that we found.
We looked for international evidence to ensure that we captured the best possible evidence from across the world, despite the potential for limited applicability between some countries. And we chose to focus on adults, excluding studies on children because there are significant differences in the differential diagnosis in the two groups; access to neuroimaging for children is usually restricted to specialists, and the threshold for referral is lower in children.
Method
This review was conducted in line with the Joanna Briggs Institute (JBI) methodology for scoping reviews and is reported according to the Preferred Reporting Items for Systematic Reviews and Meta-Analysis extension for Scoping Reviews (PRISMA-SCR) statement.20 The protocol was set before conducting the review; it was not registered or published.
Eligibility criteria
The review was structured using the Population–Context–Concept (PCC) framework;21 see Table 1 for details.
Information sources, search strategy, and article selection
We searched MEDLINE and Embase from 1974–24 May 2024. The full MEDLINE and Embase search strategies are outlined in Appendix S1.
The search results were uploaded to Rayyan22 and duplicates removed. Two reviewers screened the title and abstract for eligibility, retrieving full-text articles when necessary. In instances where the title and abstract were ambiguous, full-text articles were retrieved. Records were included if they met the inclusion criteria and none of the exclusion criteria as agreed on by two reviewers (AK and WCSC). Conflicts were resolved by a third reviewer (DH) through discussions or meetings. We did not critically appraise study quality but used study design as a proxy for evidential quality.
Data extraction, data items, and narrative review
A standardised data extraction form was developed. Two reviewers (AK and WCSC) worked independently to extract study details, and an additional reviewer (DH) resolved any conflicts. For all studies, we extracted data on the country of origin, setting, publication type, study design, and type of headache treated. For comparative studies discussing interventions, we extracted information on the intervention and comparisons used, tools for measuring outcomes, and findings. For non-comparative studies, we gathered information on the findings, themes, and the authors’ recommendations.
The results are presented in traditional narrative form.23,24 We did not undertake a formal narrative synthesis, instead we aimed to summarise the studies as a body of evidence while preserving their idiosyncratic and unique nature. This allowed us to accommodate the different research questions, designs, and contexts of individual studies, which are presented in tabular summaries.
Results
Selection of sources of evidence
Initial database searches identified 1125 records after the removal of duplicates (see Figure 1). Forty-three articles fulfilled the criteria using the title and abstract. The full text of these articles was retrieved and assessed. Eight articles were excluded at this stage for focusing on the following: secondary care perspective of headache referrals (n = 3); the prevalence of headache in primary care (n = 1); patients’ perspectives (n = 3); and not being primary research (n = 1). This left 35 articles, of which seven articles were reporting similar results to another already included study and so were excluded.25–31 In total, 28 unique studies were included in the final review.
Characteristics of sources of evidence
Twenty-five studies were conducted in Europe (n = 25),3,4,8,32–53 two studies in Africa (n = 2),54,55 and one study in Australia (n = 1).56 Among European countries, 10 studies were conducted in the UK (n = 10),3,4,33,43–46,48,51,52 three studies each in The Netherlands (n = 3)8,34,35 and Norway (n = 3),40,50,53 two studies were conducted across multiple European countries (n = 2),39,42 and one study each in Germany (n = 1),49 Italy (n = 1),36 Denmark (n = 1),47 Switzerland (n = 1),38 Spain (n = 1),41 Russia (n = 1),37 and Estonia (n = 1).32
We classified the studies methodologically as follows: comparative studies (n = 6)4,32–35,53 and non-comparative studies (n = 22).3,8,36–52,54–56 Among the non-comparative studies, 17 used quantitative methods (n = 17)3,36–48,54–56 and five used qualitative methods (n = 5).8,49–52
Twenty-three studies included in this review were publications in peer-reviewed journals and five were conference proceedings (n = 5).37,41,46,48,55
The majority of studies investigated patients with headaches of all causes and a minority investigated specific headache types; for example, migraine, or tension-type headache. In this article we use the phrase ’headache (all causes)’ to denote the former group.
See Table 2 for the full list of study characteristics.
Comparative studies
There were six comparative studies that are summarised in Table 3.
Three of the studies investigated patients with migraine, and three investigated headaches of all causes. Four studies were trials (randomised controlled trial = 1; non-randomised controlled trials = 3), one was an observational study, and one was a retrospective cohort study. Three studies looked at educational topics, and three studies looked at non-educational topics.
Most of the educational studies reported positively on their effects. Schjott et al 53 reported positive self-perceived learning from a medical educational on treatment of migraine. Braschinsky et al 32 reported higher diagnosis rates, reduced investigations, and more initiation of treatment from a 2-day educational course, but the study did not show improvements in patient satisfaction, or reduction in referrals. Smelt et al,34 the only randomised controlled trial that we found, did not show any benefit of an educational intervention, and concluded that psychological distress among the study population was an important confounder.
Ridsdale et al 33 showed that patients were more satisfied with a GP with a special interest (GPwSI) service than a hospital neurologist service, and that the costs of the GPwSI service were lower. Taylor et al 4 showed that direct access to MRI for GPs led to high patient satisfaction, cost reductions, and no difference in the findings of the scans between the groups. Veenstra et al 35 showed reduction in referrals and reduced headaches for the nurse-led intervention compared with management by a GP, but no significant difference in patient satisfaction.
Non-comparative studies
Quantitative
There were 17 non-comparative studies that utilised quantitative methods. The study population, study design, and the focus of each study is summarised in Table 4.
Eight studies investigated assessment strategies used by GPs (n = 8).40,42–45,47,54,56 Seven of these studies looked at assessment strategies involving the use of imaging (n = 7).40,42–45,54,56 and four explored the use of patient headache diaries (n = 4).40,42,47,54 Three studies reported the use of guidelines and recommendations.40,42,54
Eleven studies explored GPs’ behaviour and choices in prescribing acute treatments (n = 11)3,36,38–42,47,48,54,55 and five studies examined prophylactic treatments (n = 5).40–42,47,54
Fourteen studies investigated GP referrals (n = 14).3,4,38–40,42–44,46–48,54–56 Of these, five studies reported that GPs referred patients to specialists (n = 5),39,40,42,48,54 four studies involved referrals to both specialists and imaging services (computed tomography [CT] and MRI scans) (n = 4),43,44,47,56 two studies referred patients to neurology clinics (n = 2),3,46 and one study involved referrals to imaging services only (n = 1).38 The most common reasons for these referrals were better treatment options for patients (n = 6),39,40,42,48,54,56 diagnosis or diagnosis confirmation (n = 3),42,54,56 diagnostic uncertainty (n = 2),46,57 and seeking expert advice (n = 1).38
Five studies investigated training and education for GPs. Of these, one study reported on available continuing medical education (CME) for GPs (n = 1)36 and four studies focused on continuing training or learning needs (n = 4).37,38,41,55 Two studies found a significant proportion of GPs desired additional education on headache management through practice-oriented workshops (n = 1)38 and postgraduate courses (n = 1).55 One study highlighted common mistakes in headache evaluation and management by GPs, emphasising the need for enhanced training (n = 1).41 Another study indicated that training could lead to a 15–20% increase in headache diagnosis and management efficiency (n = 1).37 One study highlighted the lack of GP awareness of evidence-based medicine (EBM) owing to difficulties in interpreting and accessing relevant information (n= 1).36 Another study explored cost-effectiveness, noting that GPs with direct access to CT scans were cost-saving by reducing unnecessary referrals to secondary care (n = 1).43
Qualitative
There were five non-comparative studies that utilised qualitative interviews; these are summarised in Table 5.
From the three studies on managing all headache types, several themes emerged. One study (n = 1)50 highlighted GPs’ views on the diagnostic approach, including their understanding of patients and their medical history, reliance on intuition, personal experience, and the passage of time.49 One study (n = 1)51 reported disparities in GPs’ confidence in patient referrals and the diversity of referral approaches, considering factors such as identifying life-threatening conditions, tolerance for uncertainty, beliefs about patient entitlement to referrals, perception of referral benefits, availability of local services, including GPwSI in clinics funded by charities. GPs were often compelled to make referrals owing to patient anxiety.51 Another study (n = 1)52 found that GPs used scans to guide management, address uncertainty, and facilitate preventive treatment discussions, even without a perceived benefit in reassuring patients. GPs who received prior teaching and education were more confident in managing patients and interpreting radiology reports compared with those who received no additional education.52
One study (n = 1)8 focusing on the management of migraine highlighted GPs’ decision-making processes in administering prophylactic medication when acute medication provides insufficient relief.
Another study (n = 1)50 discussing medication overuse headache reported the importance of considering patient autonomy, the benefits of reducing patient resistance to medication-induced headaches by formally diagnosing it as ’medication overuse headache’. It also highlighted the significance of building a strong alliance with patients to effectively integrate brief interventions (BIs) into regular consultations for self-management of headaches by constantly reshaping patients’ perceptions of their headaches and medication use.
Discussion
Summary
We identified 28 studies that met our criteria. Six studies used comparative methods, three of which investigated educational interventions. The educational interventions showed positive effects on learning and patient outcomes, such as diagnosis rates, but the only RCT found no significant benefits. Other comparative studies highlighted satisfaction with GP with an extended role (GPwER) headache services, benefits from direct MRI access, and advantages of nurse-led headache services. Twenty-two studies used non-comparative methods, such as surveys and interviews, exploring assessment and/or diagnosis, referral rationale, decision making for prescribing prophylactic medications, educational initiatives, direct neuroimaging access, and GPwSI and nurse-led interventions.
Despite high quality clinical guidelines for headache assessment and management, implementation in primary care is problematic, with educational interventions often being the focus of studies. There is evidence to indicate that an educational intervention delivered in primary care could improve patient outcomes, improve confidence among GPs, reduce unnecessary investigations, reduce referrals to secondary care, and reduce costs. Further research is needed to assess the quality of current evidence and refine interventions with a signal of efficacy, and to design definitive trials.
Strengths and limitations
As far as we are aware this is the first review in the published literature on the assessment and management of headaches in primary care. It was conducted according to gold standard methods (PRISMA-SCR and JBI) ensuring a transparent, systematic, credible, and replicable approach.20,58 We comprehensively identified the available literature, providing an overview of each article and a narrative review of the evidence, and we have identified knowledge gaps and made suggestions for further research.
Scoping reviews often identify methodologically heterogenous literature, which makes comprehensive and coherent quality assessment across the different methods challenging. Our study was not externally funded, limiting the capacity of the review team. We did not critically appraise study quality, but we reported the design of each study as a proxy of evidential quality. Our review was limited to English-language studies, potentially omitting valuable research in other languages and introducing language bias, resulting in an incomplete reflection of the full body of international evidence. Owing to capacity constraints, our search was confined to two databases meaning that we may have missed articles that were indexed in other databases. Including additional databases, such as the Cumulative Index to Nursing and Allied Health Literature (CINAHL), and sources of grey literature could have provided more comprehensive coverage.
Comparison with existing literature
Although there are many good quality clinical guidelines for headache, most of these do not specify which sector (primary care, secondary care, tertiary care) should provide the elements of care that are recommended. People affected by headache are often living with multimorbidity (physical and psychological) and polypharmacy that requires a generalist whole-person approach, which makes primary care the optimal sector to deliver most care for people with headache. Despite this, the literature is dominated by secondary and tertiary care perspectives, which is not useful in a primary care context, and which creates an epistemic bias. This review redresses that bias and presents the evidence that is relevant to primary care.
Implications for research and practice
The articles in this review provide evidence for GPs, clinicians, commissioners, managers, and policymakers. While we did not formally assess evidence quality, we identified studies, particularly those using comparative methods with outcome data, suggesting that educational interventions in primary care can improve patient outcomes, boost GP confidence, and reduce unnecessary investigations, referrals, and costs. A key implication of this review is the need for formal quality assessment, further research, and the development of effective interventions.
The best design of service reconfigurations or interventions based on the evidence available is open to interpretation. Many of the problems with delivery of care for people with headaches reflect lack of capacity across the whole system and not specific problems with primary care. Many issues with headache care delivery stem from system-wide capacity limitations, not specific problems in primary care. New services must involve whole-system modelling, including health economics, to ensure any additional costs are justified by savings in areas such as emergency care, referrals, and neuroimaging. Clear boundaries must be established between primary and secondary care to prevent the current unproductive disputes that currently prevail. Once boundaries are established, structures should be put in place to encourage strong relationships with specialists who can provide advice, support, and specialist review when necessary.
Notes
Funding
Not externally funded.
Ethical approval
Not required. This study solely involves the evaluation of previously published literature.
Provenance
Freely submitted; externally peer reviewed.
Data
Data is not available.
Acknowledgements
Some of this work was undertaken as part of a master’s degree in public health at the University of Sheffield.
Competing interests
The authors declare that no competing interests exist.
- Received March 25, 2025.
- Accepted April 4, 2025.
- Copyright © 2025, The Authors
This article is Open Access: CC BY license (https://creativecommons.org/licenses/by/4.0/)
References
In this issue
Jump to section
More in this TOC Section
Related Articles
Cited By...
