Oral surgery

The specialty of oral surgery deals with the diagnosis and management of pathology of the mouth and jaws that requires surgical intervention. Oral surgery involves the treatment of children, adolescents and adults, and the management of dentally anxious and medically complex patients. Oral surgery care is provided by oral surgeons and by oral and maxillofacial surgeons, as the clinical competencies of these two specialties overlap. Oral surgery is a recognised specialty of dentistry, whereas the UK General Medical Council recognises ‘Oral and Maxillofacial Surgery’ as a medical specialty.¹

NHS England’s Commissioning Guide for Oral Surgery and Oral Medicine¹ describes the provision of oral surgery within the English NHS. The guidance describes three levels of case complexity, known as Levels 1, 2 and 3.¹ Oral surgery complexity is generally assessed based on both the type of procedure and patient factors. For example, a very simple surgical procedure can be complicated by a patient’s medical history or degree of anxiety. A full list of procedures and conditions that would be treated in each complexity Level can be found in Appendix 1.

In England, oral surgery is typically delivered in one of three settings and by three distinct groups of clinicians:

1. Primary care general dental practice – most surgical procedures are conducted in general practice by general dental practitioners (GDPs). The removal of simple teeth and roots is covered under the mandatory services section of the General Dental Service (GDS) contract.² The extraction of one or more teeth or roots in a single course of treatment attracts a band 2 charge for the patient and results in a ‘payment’ of 3 Units of Dental Activity (UDAs) for the dental practitioner (UDA prices vary between practices, but an average value of £25 is usually used). Patients will typically be free of systemic disease and will not require adjunct sedation, and the procedure will not be technically demanding. Such procedures are known as Level 1 procedures. There is an expectation that all Level 1 procedures will be undertaken in practice unless there are patient factors that complicate management. If a GDP does not feel able to undertake the procedure, they should look within the practice to see if another clinician can assist. Ultimately, it is the provider’s (GDS contract holder’s) responsibility to ensure that Level 1 procedures are undertaken in practice.
2. Intermediate services, dentists with a special interest (DwSpIs) – these services provide Level 2 care, and are typically delivered by a clinician with enhanced skills and experience who may or may not be on a specialist register. Indeed, such services could be provided by a consultant-grade clinician operating and remunerated as a Level 2 provider. It is expected that most Level 2 services will be provided in a primary care setting (where additional equipment may be required) under Any Qualified Provider (AQP), GDS or Personal Dental Service (PDS) contracts, in which case patient’s charges will be levied. Level 2 procedures may be delivered as part of continuing care or, as is most usual, by referral. The basis for the development of the DwSpI services was the 2004 framework document produced by the Department of Health (DH) and the Faculty of General Dental Practice, which was followed in 2006 by guidelines for commissioning such services.^3,4
3. Consultant or specialist care – the commissioning guidance describes Levels 3a and 3b but, for the purposes of this report, Level 3 providers are typically consultant-led services delivered in, and by, NHS hospital trusts under NHS standard contracts. Although Level 3 services are led by consultants, they will typically engage a wider workforce, including specialty and associate specialist-grade clinicians and those in formal training positions. Hospitals delivering oral surgery services at Level 3 include district general hospitals, larger training hospitals (foundation trusts) and dental hospitals that have the additional requirement to train dental undergraduates.⁵

Health needs assessments

Commissioning of primary care services often takes place with little needs assessment or knowledge of where referrals come from and where populations receiving care are based. Detailed knowledge of the population, their needs and treatment preferences are essential to ensure that primary care services can be delivered effectively.⁶

There is no defined methodology for determining the needs of a population for oral surgery services, in contrast to, for example, orthodontic services, where there is a clear and well-defined approach.⁷ The Adult Dental Health Survey, of 2009,⁸ reported that 8% of dentate adults had one or more untreatable teeth (on average 2.2 teeth), but it is difficult to know when or if this need for an extraction is or will be expressed as demand. For example, many decayed roots will be asymptomatic. Data from the NHS Business Services Authority (BSA) may provide indications of activity in relation to band 2 course of treatment provision (which includes tooth extraction, but also includes fillings and root fillings) and the number of extractions provided in primary care, while Hospital Episodes Statistics (HES) can provide similar information for secondary care. None of these methodologies provides information on case complexity; coding and tariff charges in secondary care are not consistently applied and the use of general anaesthesia (GA) or sedation in both settings is poorly understood.^9,10

Referrals from primary care

In 2006, a new dental contract was introduced in England that replaced a fee-for-item service with a banded course of treatment approach. UDAs are awarded based on the type of care delivered, for example 1 UDA for a check-up with preventative care, 3 UDAs for any number of extractions (although most frequently one), any number of fillings or root filings plus any treatment included in band 1, and 12 UDAs for work requiring laboratory input, such as crowns and dentures, plus any treatment included in a band 1 or 2 course of treatment. Co-payments or patient charges are levied according to the three bandings: band 1, £19.70; band 2, £53.90; and band 3, £233.70. These charges apply to services provided under GDS and PDS contracts, but do not apply to services provided under hospital contracts. An anomaly in the contract required dentists to claim the UDA tariff for the procedures that they were referring for and collect the appropriate patient charge revenue. This created a perverse incentive to refer: dentists would be paid the same fee to refer a patient as to undertake the procedure. This resulted in the NHS paying twice for activity: once in primary care, and then again in secondary care. The contract incentive was, in part, responsible for the increase in referrals seen since 2006. Other factors contributing to the seeming relentless increase in referrals,¹¹ which has been mirrored in medicine,¹² include a lack of oral surgery experience at the undergraduate level among junior GDPs,¹³ and the increasing proportion of older patients retaining their teeth but presenting with complex medical histories and polypharmacy. Despite the 2006 contract being causally linked to the increase in referrals, it had been recognised for some time that the capacity in oral surgery services was under pressure.³ Kendall, in an assessment of English oral surgery services, demonstrated that in a 3-year period from 2004, referrals doubled from a monthly average of 182 to 364.¹⁴

Reasons for referral from primary to secondary care vary, but a questionnaire completed by dentists in Greater Manchester¹⁵ found the following:

• anticipated surgical difficulty (69% of cases)
• medical history issues (49% of cases)
• require a second opinion (32% of cases)
• practitioners do not undertake surgical procedures (29% of cases)
• practitioners lack appropriate facilities or staff (28% of cases)
• patients require emergency management of pain, swelling or haemorrhage (11% of cases).

Reasons for referral were not mutually exclusive.

The costs of providing oral surgery in secondary care are substantial. In 2009/10, in the north-west region alone, the total cost amounted to £53,864,857. In addition to the cost element, the increase in referrals has caused issues around workforce insufficiency and capacity, and has negatively impacted on 18-week referral to treatment targets. Although some trusts have welcomed the increased activity, others, especially those departments in district general hospitals (DGHs) staffed by oral maxillofacial surgery (OMFS) services, have found that the oral surgery referrals deflect activity from their core offering and that many are inappropriate.

Primary care (Level 2) services

As with many service developments in the NHS, formal evaluation and published reports on primary care oral surgery services are sparse. Kendall provided one of two such descriptions of how a Level 2 provider might work¹⁴ and this is described below, while Bell describes a retrospective audit of a primary care scheme designed to address issues in provision of services to remote areas.¹⁶

Utilising two GDP practices in the Croydon area, combined with a simple referral management system, all non-urgent referrals for oral surgery procedures were captured and then subjected to a two-stage triage process. The first was an administrative check of the paper referral form and the second was a clinical assessment of a patient’s suitability for primary care treatment. The scheme reported no reduction in the total number of referrals received, but the offer of a primary care service did not appear to stimulate demand or increase Level 1 (work that should be performed by a GDP) referrals from GDPs. Of the 3117 non-urgent referrals, 36% were sent to secondary care and 59% to primary care (data were missing in 5% of cases). No referrals were returned to the GDP as being unsuitable – that is, at Level 1.

The removal of a substantial amount of activity from a single trust (nearly 60% of referrals) could cause concern over the stability of the service. Kendall states, however, that, rather than destabilising the unit, the reduction enabled a balance to be obtained, a reduction in waiting times and a re-focus on the core provision of oral maxillofacial procedures rather than oral surgery.¹⁴

The service was revisited in both 2011¹⁷ and 2012.¹⁸ Kendall reported that, after 2.5 years of service, the background referral numbers continued to rise but the proportion of referrals directed to the primary care (Level 2) service had also increased, from 60% in 2004 to nearly 80% in 2010. With so many referrals being appropriate for Level 2 services, this suggests that much of this work was being undertaken in GDS and, with the provision of a service to deliver this, it is now being referred outwards. The number of Level 1 referrals remained low, at 1.1%.¹⁷ Kendall reports a basic economic analysis with referral management costing around £7 per referral and a £600 saving per case seen in primary care. However, a system-wide economic appraisal was not undertaken and a formal health economic evaluation of costs and effects was not possible from the data available.

Referral management

The recognised increase in referrals from both GDPs and general medical practitioners (GPs) into hospital services has initiated several approaches to management of the problem. A Cochrane systematic review divided these approaches broadly into three main groups: (1) financial, (2) managerial or (3) professional education.¹⁹ Each of the main groups involves varying degrees of active interruption to the referral process:

1. Professional education involves the production, dissemination and support of clear referral guidance, often using harmonised referrals forms to encourage collection of appropriate data, and often supported by targeted continuing professional development initiatives.
2. Managerial systems include the use of referral management services, clinical assessment services, clinical assessment and treatment services and as ‘in-house’ second opinions or peer review.
3. Financial approaches (at least those of relevance to the NHS) are based on fundholding by referrers and, hence, they incentivise care provided within the practice or referred to lower-priced primary care facilities.

The Cochrane review concluded that research into the management of referrals was limited.¹⁹ Preliminary findings suggested that passive systems, such as the introduction of referral guidelines, were unlikely to change referral behaviours. The use of structured referral forms has some potential, but informatics support would be needed to make such forms useable in a practice environment (i.e. to force adherence to completion of mandatory fields). Financial methods risked the application of unselective reductions in referrals and negative impacts on patient care.¹⁹ None of the studies examined a formal referral triage and capture service, such as that employed by Kendall.^14,17

The King’s Fund reviewed referral management systems,²⁰ recognising that such systems can be as simple as a referral guideline through to active interventions in the referral pathway. It summarised that not all referrals were needed, but some patients who needed a referral did not receive one. The review found that the quality of referral letters was often poor, and appropriate primary care treatment or investigations had often not been undertaken prior to referral.

Focusing on capture and assessment referral solutions, the report found a range of strengths and weaknesses (including the filtering of inappropriate referrals, improving quality of referrals and providing commissioning intelligence), but also potentially increasing costs, delay to a patient’s journey and the creation of barriers between primary care and secondary care colleagues.²⁰ Of interest was the reported belief by primary care trusts (PCTs) that their referral management systems were reducing activity despite the fact that the data from acute trusts did not support this supposition.

A study by Cox et al.²¹ highlights this anomaly. Using data from NHS Norfolk and using an active referral management centre (RMC) approach, Cox et al.²¹ found that, in all cases, the use of the RMC approach increased referrals rather than decreased them. The authors concluded that the RMC approach, as the most expensive management option, was the least effective.²¹ The authors’ retrospective time series design looked only at decreasing attendance, that is, reducing overall referrals, and there were no primary care redirection services. Their approach, therefore, would reduce only those referrals that evidence clearly showed were inappropriate. Referrals that were incomplete or poor quality, although initially returned, would be corrected and resent, thus increasing the number of referrals. These findings are consistent with those of Kendall’s work in dentistry, which revealed very low rates of Level 1 referrals – that is, those that might be considered inappropriate.¹⁴

Two recent comprehensive reviews have examined demand management.^22,23 The key findings of these reviews provide the current context for our study, and, rather than duplicate the reviews in a formal literature review chapter, the key findings of these two large reviews are summarised here. The first, by Blank et al.,²² sought to examine interventions related to referrals from primary care to specialist services. The work focused exclusively on referrals from GPs and excluded dentistry. The systems of referral were described as complex because of the interplay of local factors, such as waiting times, the directory of services and access to specialists. The review found stronger evidence to support interventions that involved peer review, improved the quality of referral information, offered specialist contact prior to referral, electronic referrals and the provision of community specialist services.²² It found weaker and conflicting evidence over the use of gatekeeping systems and alterations in remuneration. The current work reported here addresses the issues raised in the review, apart from specialist contact prior to referral. By incorporating a standard referral form with mandatory fields, referral quality is improved, an electronic-only submission route can be implemented and a primary care service for the delivery of appropriate oral surgery procedures is introduced. The referral management and triage process can be considered a ‘gatekeeper’ with the potential to divert referrals and reject those considered ‘inappropriate’.

The second review, by Winpenny et al.,²³ focused on the effectiveness and efficiency of moving hospital services (outpatients) into primary care, and examined 184 studies, some of which included dental settings. They found that minor surgical procedures could be carried out in primary care safely and effectively, but that provision of such services could stimulate demand by addressing previously unmet need. The cost-effectiveness of these services was likely to depend on local contracting, and this also applied to general practitioner with a special interest (GPwSI) services, which also demonstrated evidence of supply-induced demand.²³ The review found that direct access to specialist services in some cases (such as audiology for hearing tests) offered obvious benefits, but that in other cases (such as musculoskeletal services) it risked generating a substantial increase in demand. The review considered referral management services as a substudy group and included a qualitative study element with individuals working, commissioning or implementing such services. The group identified four emerging themes from their interviews:

1. the lack of clarity relating the aims of functions of referral management services
2. the challenge of stakeholder adoption and buy-in
3. practical and administrative difficulties
4. the impact of perceived effectiveness of the aims and priorities of such services.

The group recommended that future schemes should have clarity of aims and defined indicators of success. In addition, the group identified a research need in the evaluation of clinical effectiveness and cost-effectiveness of RMCs.

Electronic referral management systems

The NHS ‘Choose and Book’ system, now known as the NHS e-Referral Service (eRS),²⁴ is an example of a large, national electronic referral system. Choose and Book has reduced the administrative burden associated with appointment booking and may have reduced the non-attendance rate at secondary care clinics.^24,25 However, this system may not be appropriate for all specialties and in all contexts. Prior to and following the introduction of Choose and Book, independent electronic referral management systems were developed. For example, Maddison et al.²⁶ evaluated electronic referral management with central triage and an adjunct specialist primary care service for uncomplicated musculoskeletal conditions. They found that, although the number of referrals greatly increased following the introduction of referral management, waiting times fell. In addition, duplicate referrals disappeared and a high degree of patient satisfaction was reported. However, Kim et al.²⁷ found that, although electronic referral management improved access to care, there were some barriers to implementation. Some referring clinics reported that multistep login procedures and a lack of computer access and reliable internet connection contributed to electronic referrals taking longer to complete, which was associated with lower satisfaction with overall clinical care. Again, this finding highlights the importance of considering the context in which the referral management interventions are implemented.

Consultant triage

When GDPs are aware that their referrals are being scrutinised by a peer with a specialist training, their referral behaviours may alter around who and how they choose to refer to secondary care. Studies of peer-reviewed interventions, in which referral quality has been judged by consultants and fed back to GPs, have resulted in some improvement in the quality of referral information and a reduction in the number of overall referrals into secondary care,^28–30 although it may not lead to permanent changes in practice.³¹ In addition, electronic referrals directly from GPs to consultant triagers prior to making Choose and Book appointments were found to be associated with shorter waiting times for appointments than paper referrals.²⁵ There is some evidence to suggest that the consultant triage element may improve the quality and appropriateness of dental referrals; however, GDPs may feel that their clinical autonomy is compromised by examination of their referrals during the triage process.

Despite the apparent lack, or contradictory nature, of evidence to support active referral management systems, by 2009, 91% of PCTs had some form of referral management system in place for GPs.²¹ These systems seek to influence either the decision to refer, the destination of the referral or the quality of the information provided in the referral. At the time of writing there are several referral management systems in place across NHS England Area Teams (ATs) for dentistry – largely resulting from the guidance issued in the Dental Commissioning Guides in which RMCs are central to the process of directing referrals into Level 2 services.¹

Quality of referrals from general dental practitioners

In addition to managing waiting lists, reducing costs and improving overall patient satisfaction, referral management has the potential to improve the quality of referrals from primary care, which may improve triage efficiency and the overall diagnostic accuracy of the content of referrals.²⁰ Qualitative work around quality and appropriateness of referrals from GPs assessed by senior NHS clinical and managerial staff in five PCT areas in England³² found that important attributes of appropriate referrals were:

1. Necessity – should the patient be referred based on clinical examination, National Institute for Health and Care Excellence (NICE) guidelines or their own medical history?
2. Destination – could and should the patient be treated in an intermediate setting rather than in secondary care?
3. Quality – is the information contained in the referral relevant and thorough, and have the necessary investigations taken place?

This is congruent with the NHS Quality, Innovation, Productivity and Prevention (QIPP) agenda around promoting quality while making efficiency savings, and offers evidence to suggest that referral management interventions do have the potential to be cost-effective while supporting improvements in quality. In 2011, a commentary on the implications of QIPP for dentistry identified the development of centralised assessment and triage services and establishment of primary care-based specialised services as innovations that could contribute to the QIPP agenda.³³

Many audits have assessed the quality of referrals from GDPs into all specialties. The results are usually poor, with ‘Dear Sir’ letters still being commonly employed, which contain little in the way of clinical detail, rationale for treatment, results of special investigations or the provision of radiographs.^34,35 Increasing the quality of referrals facilitates the provision of triage, informs commissioning (if data are appropriately captured) and increases the efficiencies of primary and secondary care services by enabling appointment scheduling and clinical staff allocation to be appropriate to complexity.

The King’s Fund describes the financial challenges facing the NHS, stating that difficult ‘trade-offs’ will be required.³⁶ Cost-saving measures in the NHS are rarely welcomed and often viewed as reducing quality and impacting on patient choice. However, beyond the clear need for dentistry to contribute to savings the NHS must make, there is the possibility that referral management allied to primary care diversion has much to offer in terms of quality enhancement, for example:

• care closer to home
• more convenient appointment times and extended opening hours
• reduced waiting times (Kendall reports 6 weeks in primary care vs. 18 weeks in secondary care¹⁴)
• single ‘see and treat’ appointments reducing opportunity costs for patients
• greater productivity leading to increased capacity.

Concerns and consequences

Despite the apparent ‘easy win’ that centralised triage and primary care-based oral surgery services offer for patients and the NHS, concerns have been voiced. The removal of ‘simpler’ cases from secondary care is a potentially destabilising move, and the resultant shift of case mix to more complex patients is a perceived risk to the training of both undergraduates and specialist trainees.¹³ Indeed, the use of a referral management system could, by reducing undergraduate training experiences, result in an increase in referrals from a population of graduates with no oral surgery skills. Hospital trusts may argue that the current tariff arrangements are based on the assumption that hospitals treat a wide range of cases, with payment for simpler procedures helping to generating revenue to offset the higher costs incurred in treating more complex cases that cannot be fully recovered from the tariffs.²²

Clinicians in secondary care have argued that, although care can be delivered in primary care, there is not always a compelling reason why it should. The reduced governance in primary care (for example, wrong site surgery reporting) combined with the single-handed nature of oral surgery provision in primary care, compared with a team approach in hospital, threatens, it is argued, the quality of the care provided. If patients experience complications, these will largely need to be managed by secondary care and, hence, savings are lost, patients experience poor outcomes and the system fails.²³

A further concern is that implementing primary care specialist services adds another service to the system without adequately managing the supply side, that is, without ensuring that there is corresponding downsizing of secondary care services. So instead of substitution, supplementation occurs, producing an overall increase in costs, which is a significant risk for a financially strapped NHS. A good example of possible pitfalls is described by Richardson et al.³⁷ in the context of developing skill mix by introducing nurse practitioners.

Aims and objectives

Considering the identified evidence gaps and the pressing need for the NHS to understand the quality and financial impact of referral management services, there is a need for a high-quality, contemporary evaluation of this change in service organisation and delivery.^22,23,38

This project aimed to evaluate the introduction of an electronic referral management system with consultant-level triage and the introduction of a new primary care service for oral surgery within a defined health-care system containing a diverse set of hospital providers. The study design used a mixed-methods approach with ITS design.

In addition, a diagnostic accuracy study of remote clinical triage was undertaken to assess the efficiency of this important stage in the referral management process. The research programme contains the necessary elements to address the research gaps identified by the previous systematic reviews,^20,22,23 primarily the impact on quality of referrals, use of electronic referrals, the provision of community specialist services and the effect of gatekeeping systems.

The main research question to be addressed by this work was the following:

• How does a robust online referral management and triage system, allied to provision of a specialist primary care service, impact on the costs and quality of oral surgery services provided by different providers in different settings in a defined health-care system?

At the highest level, we wanted to know if we can change the behaviour of referring GDPs without destabilising a complex, interdependent acute sector, to ensure that only those who need hospital care are managed in this setting. In order to fully answer this main research question, the following secondary research questions were formulated:

• Chapter 2, Efficiency of remote clinical triage
  • How do remote clinical triage outcomes conducted by an experienced consultant compare with outcomes of face-to-face examination (reference test) performed by the same consultant?
  • How do remote clinical triage outcomes performed by GDPs and different consultants compare with outcomes of face-to-face examination performed by an experienced consultant (reference test)?
  • What are the views of triagers on the benefits and problems of a remote clinical triage system and how can the system be improved based on their experiences?
• Chapter 3, Implementation and health needs assessment (phase 1)
  • What are the practical issues for the NHS in introducing an all-electronic referral system from scratch?
  • What is the effect of all-electronic referral system on
    • the total number of referrals
    • the quality of referrals including an assessment of compliance with national referral guidelines
    • the time taken to complete referrals?
  • What are the views of key stakeholders on the benefits and problems of the electronic referral system and how can the system be improved based on their experiences?
• Chapter 4, Active referral management with consultant and general dental practitioner triage: quantitative findings including economic evaluation (phases 2 and 3)
  • What are the differences in referral numbers, referral quality and the mean cost per referral between virtual management (phase 1) and consultant-led active management (phase 2)?
  • What are the differences in referral numbers, referral quality and mean cost per referral between the year of virtual management (phase 1) and GDP-led active management (phase 3)?
  • How do these findings (differences in referral numbers, referral quality and the mean cost per referral between study phases) differ by the provider of secondary care?
  • Does consultant-led triage offer improved costs over GDP self-determined provider choice (phase 2 vs. phase 3)?
  • How do the views and experiences of patients differ between those using primary and secondary care services?
• Chapter 5, Active referral management with consultant and general dental practitioner triage: qualitative findings (phases 2 and 3)
  • The use of the ITS methodology with robust adjunct and parallel qualitative components enables these issues to be addressed from both a metric and a narrative perspective.
  • What are the issues encountered when establishing a new primary care oral surgery service?
  • What are the views of stakeholders on the development and implementation of the primary care service?
  • What are the views of service users on the quality of service they received from the referral management and triage system?

Public and patient involvement

Public and patient involvement has been a key element of this work, from the design stage, in which consent and patient information sheets were reviewed and revised, through to the extensive involvement of patients (see Appendix 2) in the qualitative component of the research. Patients’ voices are heard and reflected strongly in the current work, as their views and experiences are key to meeting the aims of the research. Service redesign impacts multiple stakeholders and, although professional views are often heard, we have sought to ensure that those of service users in Sefton are recognised and reflected.