Evidence for models of diagnostic service provision in the community: literature mapping exercise and focused rapid reviews

Identification of UK studies

Studies conducted in the UK were identified in three ways:

1. Geographical terms for ‘united kingdom’, ‘uk’, ‘britain’, ‘England’, ‘Scotland’, ‘Wales’, ‘Ireland’ were used to retrieve records from the PubMed database.

2. Searches by technology, with no geographical limitations given its UK focus, were conducted on The King’s Fund database.

3. Similarly, geographical terms for ‘united kingdom’, ‘uk’, ‘britain’, ‘England’, ‘Scotland’, Wales, Ireland, combined with the technology and words such as ‘equipment, logistics, premises, facilities’ were used to retrieve items from Google Scholar (Google Inc., Mountain View, CA, USA). By harnessing full-text retrieval, this search added value over the title-and-abstract-based approach listed above.

Retrieved items were sifted against the inclusion/exclusion criteria listed in the following section. A bibliography of key items was compiled for each technology. This bibliography was used to retrieve full texts when available and to perform citation searches. As additional relevant items were retrieved, further citation searches were performed for these in turn until saturation of relevant items was achieved.

Skills

The complexity of audiological investigation requires that GPs work with a wide range of staff. Some services for adults are now provided in a primary care setting by GPwSIs in otology. Some staff encounter initial difficulties when using the tympanometer or the pneumatic otoscope. However, a wide range of audiological investigations can be performed by community staff not specialising in audiology. In the UK, good practice guidance and standards and quality audit tools have been developed for audiology services.

STEP-UP summary statement With regard to SKILLS, audiology is attributed a MODERATE degree of implementation difficulty owing to initial difficulties in using the tympanometer or the pneumatic otoscope and ongoing concerns about the appropriateness of diagnosis and its implications for secondary care.

Training

Specialist GPs may be trained in the initial diagnosis and management of hearing and balance disorders, particularly in adults. Practice nurses can support delivery of procedures requiring basic skills such as basic hearing aid trouble-shooting, relaxation training for those with tinnitus and vertigo, and in gentle mobility training for those with vertigo. Skills in tympanometry and pneumatic otoscopy can be acquired through short courses.

STEP-UP summary statement With regard to TRAINING, audiology is attributed a MODERATE degree of implementation difficulty owing to the facility of short courses to increase confidence in use of equipment.

Equipment

Equipment most typically used by a practice would include a tympanometer (Table 8) and a pneumatic otoscope (Table 9). Audiology requires a wide range of equipment. Requirements may include flexible fibre-optic nasendoscopy, Hopkins rod endoscopy, a soundproof facility, audiometry, tympanometry and microscopy. Over 75% of all patients will require an additional investigation or procedure. With the introduction of portable and handheld equipment, tympanometry has become more feasible for a primary care setting. Future-proofing of audiology equipment requires provision for otoacoustic emissions technology, initially used for newborn hearing screening but with potential in small children and the aged.

STEP-UP summary statement With regard to EQUIPMENT, audiology is attributed a MODERATE degree of implementation difficulty because, beyond basic equipment, many needs for additional equipment will require follow-up in secondary care.

Premises

Health Building Note 11-01, Facilities for Primary and Community Care Services,46 includes provision for an adult hearing test room and a paediatric hearing test room. Audiological testing must be performed in soundproofed accommodation built to International Standards Organisation (ISO) 8253-1 and ISO 8253-2 standards for acoustic test methods and suitable for children. Despite the growth of purpose-built clinics and the increasing miniaturisation of audiology equipment, the inability of premises to accommodate treatment rooms for audiology services has been a reported obstacle to the relocation of hospital-based provision.

STEP-UP summary statement With regard to PREMISES, audiology is attributed a MODERATE degree of implementation difficulty owing to the specific requirement for a soundproofed room.

User perspective

Generally, GPs found tympanometry and pneumatic otoscopy to be acceptable to carers and children, although some GPs stated that they preferred not to use pneumatic otoscopy as children sometimes found it uncomfortable. 57 Some GPs thought tympanometry was particularly useful for communicating with carers about ear disease, offering tangible proof to parents of the GP’s diagnosis and support of the management plan. 57 Many GPs were unclear about the significance of negative pressure tympanometry57 readings in general practice, which could potentially generate unnecessary GP follow-ups.

General practitioners believed that more training and experience was needed to become confident with pneumatic otoscopy than was needed for tympanometry. 57 GPs could teach themselves how to perform tympanometry after the relatively brief introduction to the equipment, with the main challenge being interpretation of the tympanograms, which required them to refer back to written information they were given during training. Most GPs said that they would choose to continue to use tympanometry.

STEP-UP summary statement With regard to USER PERSPECTIVES, audiology is attributed a LOW degree of implementation difficulty in the light of only minor GP concerns with equipment and the referral pathway and difficulties with investigations involving small children.

Primary–secondary interface

The Good Practice Guide, specifically for adults with tinnitus,69 recommends strategies for tinnitus assessment, management and referral at four different levels of the service: primary care, local community-based tinnitus services, specialist hospital-based centres and supra-specialist assessment centres. Patient routes through the system were to be determined by clinical assessment and specific referral criteria, designed with service efficiency and equity of patient care in mind. A systematic review of outreach services in primary care found a direct counterpoint between the needs of the patient and the needs of the service. 70 Reported disadvantages concerned administrative costs, accommodation costs and inefficient use of specialists’ time. Comparative studies showed that more patients expressed a preference for outreach clinics than for hospital-based clinics, and measures of patient satisfaction and convenience generally were higher for outreach clinics. A self-contained satellite facility within a GP practice sought to enable 5000–7000 people to receive secondary care hearing services in a primary care setting. The project explored the benefits to patients and to the health economy of shifting care closer to home. 71

Traditionally, patients needing a hearing aid would be referred by their GP to the audiology service, with one appointment for assessment and one for fitting. Advances in technology72 mean that some patients can be assessed and fitted at the same appointment.

STEP-UP summary statement With regard to the PRIMARY–SECONDARY INTERFACE, audiology is attributed a MODERATE degree of implementation difficulty owing to dependence on secondary services for further investigation and yet the possibility of streamlined hearing aid pathways.

Conclusion

Provision of audiology services in primary care is complex, in that it not only requires a wide variety of equipment and, optimally, an audiology-friendly environment, but also may necessitate access to other diagnostic services, most notably endoscopy or radiology. Table 10 summarises the overall findings. There is little evidence available to examine the impact of acquisition of audiology-related skills by primary care staff. Studies on provision of specialist outreach clinics (i.e. a hospital specialist operating in a primary care environment) offer supplementary data on equipment utilisation and premises. Current models being favoured in primary care audiology involve direct-access referral clinics or provision of services under contract by commercial (or, alternatively, independent) providers.

Skills

Based on clinical scenarios, GP experts have concluded that natriuretic peptides testing should be the routine test for suspected heart failure where referral for diagnostic testing is considered appropriate. Abnormal natriuretic peptides testing should be followed up with referral for echocardiography. 73 In a survey on barriers to effective management of heart failure, the majority of respondents who diagnosed left ventricular systolic dysfunction (LVSD) used ECGs, chest radiography and clinical assessment, the exception being nurses. Around one-quarter of the nurses and half of the GPs were confident about interpreting the results of an ECG, whereas most cardiologists and general physicians were confident. 74

STEP-UP summary statement With regard to SKILLS, cardiac services is attributed a MODERATE degree of implementation difficulty owing to the range of diagnostic scenarios and investigations.

Training

Education and training in cardiac imaging should meet the present requirements specified by the General Medical Council (GMC) in curricula in cardiology and cardiac radiology, the future demands of revalidation and the requirements of allied professional groups. 75 The British Society of Echocardiography (BSE) (www.bsecho.org), the British Society for Cardiovascular Magnetic Resonance (BSCMR) (www.bscmr.org) and the British Society for Cardiac Imaging (www.bsci.org.uk) have endorsed voluntary accreditation procedures for their members.

A survey of 170 general practices in one cardiac network suggests that training of staff in use of cardiac equipment is variable. 76 Training updates were not provided for 12-lead ECG in 20 practices, static BP in 26 practices, ambulatory BP in 21 practices and Holter monitors in 10 practices. Responding practices indicated that staff require additional training in 12-lead ECG (n = 17), static BP (n = 3), ambulatory BP (n = 11) and Holter monitors (n = 8).

A new extended GP role in cardiology was developed and piloted to enable GPs to diagnose and manage patients with mild to moderate heart failure or atrial fibrillation (AF) and to use diagnostics effectively in primary care. 77 Training entailed GPs participating in a four-session short course with ongoing clinical supervision. A mixed-methods evaluation found that participating GPs perceived the extended GP role as a professional development opportunity with the potential to reduce health-care utilisation and costs, through a reduction in referrals, while meeting the patient’s wishes for the provision of care closer to home. Patient experience of the new GP service was positive.

STEP-UP summary statement With regard to TRAINING, cardiac services is attributed a LOW degree of implementation difficulty owing to the widespread availability of appropriate training and accreditation, although underutilisation of training in basic cardiac equipment remains a concern.

Equipment

A survey of 170 general practices in one cardiac network suggests that the provision of cardiac equipment, and the training of staff in its use, is variable. 76 Service contract providers included local hospitals and equipment suppliers, while practice nurses are the group of staff most likely to be responsible for the equipment on a day-to-day basis. Following the purchase and installation of equipment, there is a need for appropriate maintenance and training to ensure optimal use and patient safety. Guidance and standards regarding equipment maintenance and training within primary care premises is difficult to identify,76 typically being scattered across a range of stakeholder professional association.

With regard to emerging technologies, Table 12 offers a specific framework assessment for B-type natriuretic peptide (BNP). Implementation of BNP will incur initial expenses including the purchase or hire of the analyser. Quality control issues relate to equipment and the purchase of limited shelf-life reagent strips. The cost of each BNP test is likely to be higher when relatively small numbers of tests are required, and will reduce with volume. 80 A key source for an evaluation of the BNP POC testing kit is a 2011 Horizon Scanning report. 81

STEP-UP summary statement With regard to EQUIPMENT, cardiac services is attributed a LOW degree of implementation difficulty owing to developments in POC testing and the increased portability of equipment.

Premises

Most aspects of cardiac services have been accommodated in primary care for many years, typically specified for the ‘near patient testing room’. 46 The implications of equipment purchases such as ECG or echocardiography are reviewed elsewhere (see Electrocardiography and Echocardiography). This review was not able to identify information on the implications of providing an increased range of cardiac testing facilities in primary care (i.e. BNP), specifically in relation to implications for the premises.

STEP-UP summary statement With regard to PREMISES, cardiac services is attributed an UNCERTAIN degree of implementation difficulty owing to the lack of information on the facilities required.

User perspective

A study in UK general practice examined management of suspected heart failure. 82 Some GPs requested a full blood count and urea and electrolytes on all of their patients with suspected heart failure. Very few GPs reported sending a patient for echocardiography to confirm a diagnosis of heart failure. Most GPs mentioned the adequacy of facilities as an obstacle to diagnosing and managing patients with heart failure, specifically lack of access to an echocardiogram. Many GPs mentioned poor access to echocardiography as a barrier to diagnosing patients with suspected heart failure. Those few GPs and practices who admitted to routinely arranging echocardiograms for patients suspected of having heart failure were also more likely to report treating patients with angiotensin-converting enzyme inhibitors.

STEP-UP summary statement With regard to the USER PERSPECTIVE, cardiac services is attributed a LOW degree of implementation difficulty owing to the non-invasive nature of most cardiac tests. However, the non-availability of echocardiology services remains a significant barrier.

Primary–secondary interface

Open-access services (echocardiography, stress testing, Holter monitoring) may simply provide an investigation or add value by providing interpretation and advice. 83 Open-access services may be provided either on hospital premises or in partnership with primary care in a community setting and should be quality assured with appropriate governance arrangements and regular systematic audit. An alternative to a primary care-based approach involves using a local NHS trust laboratory. A NHS Improvement report84 concludes that this is often a more cost-effective option, citing a likely correlation between the number of tests and the cost per test. Points to consider are the length of time for the results to be returned, the courier service required to transport the samples to the provider, transport difficulties for patients and the importance of a robust service-level agreement.

STEP-UP summary statement With regard to PRIMARY–SECONDARY INTERFACE, cardiac services is attributed a LOW degree of implementation difficulty owing to existing pathways for the management of heart disease and heart failure.

Conclusion

Benefits of a BNP service include rapid results, convenience for patients, early diagnosis and commencement of treatment. Centralising in one agreed location allows one GP practice to serve several surgeries, allowing results to be available within a few days or even on the same day. Benefits include that it is convenient for the patient, it is cheaper than doing the test in individual surgeries, and results may still be available within an acceptable time span (depending on the agreement). This type of service may involve the transportation of samples from or to the testing centre, and this must be factored into planning.

In some parts of the UK, the introduction of BNP has led to increased numbers of echo tests being performed without an increase in true positives. This means that if the test is used indiscriminately, false-positive results may outnumber true positives. A local agreement may help to decide who should order the test or when access to ordering a BNP test is appropriate. A further consideration when calculating costs is that a percentage of those with positive BNP results will require an echocardiogram. The cost of the BNP test will need to be added to the cost of an echocardiogram. Many studies have identified savings with an increase in the number of appropriate referrals once those testing negative had been set on alternative pathways of care. N-terminal pro B-type natriuretic peptide (NT-proBNP) concentrations increase with age in the normal population. Some areas, through collaboration and agreement have introduced age-related ‘cut-off’ values when deciding which patients to refer for further investigations.

The NHS Improvement Heart Improvement Network has identified examples in the UK where BNP testing has been used as a means of ‘rule-out’ for echocardiography. 80 Some documents include economic information and cost analysis for the projects. NICE guidelines on the initial diagnosis of chronic heart failure and referral for echocardiography recommend use of BNP in combination with clinical assessment. 85 NICE guidelines recommend measurement of serum natriuretic peptides in patients with suspected heart failure without previous myocardial infarction, and those with previous myocardial infarction should be referred for an urgent echocardiogram. 85 Although several hospital laboratories carry out BNP testing, few return results within a day. POC BNP testing can considerably reduce turnaround time and could lead to earlier initial treatment, more timely referral and less uncertainty for patients. Using POC BNP levels to quickly rule out heart failure could allow more rapid initiation of investigation of other causes of dyspnoea.

A UK HTA carried out a comprehensive systematic review and meta-analysis of all studies comparing the diagnostic accuracy of BNP testing with clinical examination by cardiologists in heart failure in all settings until 2006. 79 The review included 20 studies on the accuracy of BNP for the diagnosis of clinically defined heart failure. Four studies took place in primary care, one used the POC test and three used laboratory-based tests. These demonstrated a slightly lower sensitivity (84%) but similar specificity (73%). Eight of 16 studies reporting data on the accuracy of NT-proBNP for the diagnosis of clinically defined heart failure were conducted on samples from patients presenting in primary care (pooled sensitivity of 90%, a specificity of 60%). The HTA report also highlighted that the utility of BNP testing will depend on the pre-test probability of chronic heart failure in the patient. 79 BNP testing would contribute important diagnostic information, as a negative test would reduce the post-test probability. A US technology assessment gave similar results. 86

A systematic review and meta-analysis of diagnostic accuracy of BNP testing for heart failure in the emergency department, in which 9 of the 11 studies used the Triage system, showed that the ≤ 100 pg/ml cut-off value used in most studies has a sensitivity of 0.93 and a specificity of 0.66. 87 Two other systematic reviews that included the same studies achieved similar results. 88,89 However, most studies on the Triage BNP system were performed in a US emergency setting, so cut-off values for primary care may be different.

Although a study investigating the impact of POC NT-proBNP testing with the Cobas h 232 (Roche Diagnostics Limited, Rotkreuz, Switzerland) has yet to be carried out, feedback on its ease of use has been documented. 90 In this study, the nurses on duty in the coronary care unit, who operated the device, found it simple to learn and handle.

The Horizon Scanning report observes that most published studies have investigated the use of BNP in an emergency department or hospital setting. 81 It concludes that more studies are required on the diagnostic accuracy of POC BNP tests in primary care.

Skills

Wireless ECG equipment is easy to use. The challenge with ECG devices of any type lies in interpretation. A systematic review of ECG interpretation accuracy studies found that both physicians and computer software frequently made errors, compared with expert electrocardiographers; however, there was also frequent disagreement in interpretation between experts. 92 When interpreting ECG studies, it is important to distinguish the effect of using a simpler ECG from the effect of using a non-expert to interpret the trace. Thus, relatively poor results of single-lead ECGs interpreted by GPs were similar to results obtained for 12-lead ECGs when read by GPs. 93 In contrast, two studies found high sensitivity and high specificity when a bipolar thumb ECG was read by a cardiologist and a bipolar ECG was read by an experienced GP, respectively.

STEP-UP summary statement With regard to SKILLS, ECG is attributed a LOW degree of implementation difficulty owing to the fact that a large majority of practices have the equipment, although interpretation is variable.

Training

The interpretation of readings from the ECG is included in standard medical and other health-care professional (HCP) training. NICE guidance94 notes that skill can be lost in general practice if it is not used frequently, and commissioners should ensure that training on interpretation is offered, linked to local need (the national training cost is quoted to be £105 per person). Some practices may feel that they are too small to develop expertise in ECG interpretation, which may present opportunities for local practices to work together rather than refer to secondary care. In a survey of cardiac equipment use in a cardiac network, a higher proportion of registered nurses than GPs had received formal training in the use of ECG machines, although the interpretation of ECG recordings is predominantly undertaken by GPs rather than by registered nurses. 76 Most ECG machines have interpretative software, but combining interpretative software with GP interpretation does not improve the sensitivity of diagnosis significantly. The evidence suggests that GPs can detect AF on ECGs accurately, given the appropriate training. 95

STEP-UP summary statement With regard to TRAINING, ECG is attributed a LOW degree of implementation difficulty owing to the availability of training provision, although many training needs are not fulfilled.

Equipment

A new ECG machine with installation and initial training costs approximately £2000, including value-added tax (VAT). 94 Increasingly portable ECG machines have become an attractive option. Handheld ECG devices allow ECG readings to be stored on a portable device for review at a later time. Unlike 12-lead ECGs, which require electrodes to be placed on the patient and connected to the ECG, handheld devices have integral electrodes. The Horizon Scanning report96 documents devices that can be used very quickly in primary care for screening or in acutely unwell patients, as well as a monitoring device over a period of time. Two devices could be used for screening for AF in the general population. 97 One is a finger probe similar to that used in general practice for pulse oximetry; the other is a modified BP monitor as used by patients to monitor their BP at home. 98–100 The latter could be used either by people monitoring their own BP to self-screen for AF or by primary care professionals to opportunistically screen patients. These devices are able to adjust their sensitivity to optimise their value as screening devices.

The AliveCor Heart Monitor and AliveECG apps (Alive Technologies, Ashmore, QLD, Australia) are, respectively, a pocket-sized ECG recorder and a mobile device application for the analysis and communication of the results. 101

STEP-UP summary statement With regard to EQUIPMENT, ECG is attributed a LOW degree of implementation difficulty owing to the availability of a wide range of relatively cheap and acceptably accurate equipment that can be matched to clinical need.

Premises

Health Building Note 11-01, Facilities for Primary and Community Care Services,46 locates ECG in the ‘examination/therapy room’ (contains public sector information licensed under the Open Government Licence v3.0). In the remaining literature, little detail is available of the premises in which ECGs, portable or not, are to be stored, cleaned or maintained. This corresponds with the findings from the survey of equipment from a UK cardiac network. 76

STEP-UP summary statement With regard to PREMISES, ECG is attributed an UNCERTAIN degree of implementation difficulty owing to insufficient detail of storage requirements.

User perspective

The ease of use, convenience and portable nature of ECG devices means that their implementation could see a better rate of patient compliance than that for a Holter monitor in patients who use the device for home recordings. 102

STEP-UP summary statement With regard to USER PERSPECTIVE, ECG is attributed a LOW degree of implementation difficulty owing to ease of use, particularly of wireless devices.

Primary–secondary interface

Excessive delays for cardiac patients waiting for a hospital appointment led Harrow Primary Care Trust (PCT) to establish a mobile cardiac service delivered in community-based settings: in GP surgeries, in community health-care centres and in patients’ homes. 103 The service used a portable ECG system to automatically store and analyse recorded ECG data using existing computer hardware. The service has reduced waiting times and improved patient care.

STEP-UP summary statement With regard to PRIMARY–SECONDARY INTERFACE, ECG is attributed a LOW degree of implementation difficulty owing to the already-integrated nature of the diagnostic and management pathway.

Conclusion

The overall STEP-UP summary for ECG is presented in Table 13. The increasing portability, ease of use and low prices of ECG equipment has led to almost universal diffusion in NHS primary care; therefore, issues relate to choice of the most appropriate equipment for practice needs, including an optimal mix of static and portable equipment. The reduction in the number of leads in ECG systems makes them easier to use and more acceptable to both patient and clinician. Of particular value is the potential for small ECG devices to replace the 24-hour Holter monitor tests. Organisational issues concerning loans of, and maintenance for, equipment to patients may need to be addressed. Similar concerns will relate to ambulatory BP monitoring.

Skills

Echocardiography is pivotal in the management of heart failure as it can make the diagnosis, determine the aetiology and help plan treatment. 104 Community echocardiography is also indicated for heart murmurs, AF and hypertension. 105 An echocardiography service must be integral to a general local plan for heart failure to avoid isolated open-access echocardiography. A UK study comparing community echocardiography with hospital echocardiography found that community echocardiography gave comparable results with traditional hospital echocardiography for LVSD detection and for significant valvular disease detection. 106

STEP-UP summary statement With regard to SKILLS, echocardiography is attributed a MODERATE degree of implementation difficulty owing to difficulties in diagnosis.

Training

Echocardiography should be performed and reported by operators trained to the standards set by the of BSE, that is, they should hold Accreditation in Adult Transthoracic Echocardiography or Adult Community Echocardiography. The BSE has endorsed voluntary accreditation procedures for their members. BSE accreditation in community echocardiography qualifies a GPwSI in standard echocardiography, being equivalent to the BSE adult accreditation process. The provision of echocardiography services, regardless of their setting, must link with staff training and continuing professional development (CPD), which in turn informs workforce planning. In the UK study comparing community echocardiography with community echocardiography, the community echocardiogaphy was performed by a cardiology trained research fellow. 106 The authors identified a need to assess whether or not other HCPs (e.g. nurses) could be trained to successfully to provide similar accuracy with community echocardiography.

STEP-UP summary statement With regard to TRAINING, echocardiography is attributed a MODERATE degree of implementation difficulty owing to the need for an extended period of certificated training.

Equipment

A standard echocardiogram is required for all patients with suspected heart failure and is the required quality standard for Community Echocardiography. 104 The type of study is mainly determined by the clinical question, but also depends on the level of experience of the operator and the type of machine. Standard machines are usually required for standard and advanced studies. Advances in ultrasound technology have led to the development of smaller echocardiography machines that may be transported more easily to different sites in the community. However, studies have not looked at the performance of these smaller echocardiogram machines in community settings where conditions may be different, for example in terms of suitable couches, lighting and lack of an immediate second opinion.

STEP-UP summary statement With regard to EQUIPMENT, echocardiography is attributed a LOW degree of implementation difficulty owing to the ease of use and portability of acceptable devices.

Premises

Community service settings for echocardiography may include mobile units, community hospitals/diagnostic centres, primary care polyclinics/primary care centres/super surgeries, GP practice surgeries and other appropriate locations such as walk-in facilities or high-street settings. Although the performance of smaller echocardiogram machines in community settings has been found to be comparable with that of larger machines in an acute setting, the conditions may be different, for example in terms of the non-availability of suitable couches and the presence of suboptimal lighting. 106

STEP-UP summary statement With regard to PREMISES, echocardiography is attributed a LOW degree of implementation difficulty owing to the flexibility of locales and settings in which the technology may be administered.

User perspective

A UK qualitative study reports numerous barriers to use of echocardiography in primary care. These include lack of availability of electrocardiography, chest radiography and echocardiography, and lack of confidence in interpreting results. GPs are keen to have access to echocardiography, and, where access is provided, they use it appropriately. 82 In another qualitative UK study, cardiologists and nurses were more confident in using and interpreting echocardiography reports than GPs107 but expressed frustration that patients were not routinely referred for echocardiography testing and about the poor quality of referral information.

STEP-UP summary statement With regard to USER PERSPECTIVE, echocardiography is attributed a LOW degree of implementation difficulty owing to the ease of use and lack of an invasive procedure.

Primary–secondary interface

The organisation of services should be based around a BSE-accredited department but organised as a network, avoiding barriers between community and hospitals. Services should extend between hospital and community bases as part of a continuum of care, irrespective of organisational barriers. The service should complement existing hospital-based provision and be integral to the delivery of heart failure services in a health community. The service should be linked to the hospital-based service for second opinions and clinical back-up and to ensure quality assurance.

The favoured model of care is probably the one-stop heart failure clinic, offering diagnosis and initial treatment while liaising with GPs and nurses. 108 A community echocardiography service must always be considered in the context of the whole patient pathway for those with suspected heart failure. The Rapid Access Heart Failure Clinic locates echocardiography within a heart failure one-stop shop linking secondary care-based staff with primary care staff for ongoing care and tertiary staff for specific specialist advice.

STEP-UP summary statement With regard to PRIMARY–SECONDARY INTERFACE, echocardiography is attributed a MODERATE degree of implementation difficulty owing to the need for secondary care support in training, support and referral.

Conclusion

Echocardiography remains a skill that continues to challenge significant numbers of primary care staff, including GPs. Although the administration of the technology is comparatively easy, the challenge lies in the interpretation of the readings. Some GPs may welcome the opportunity to pass on the responsibility for definitive diagnosis and resultant communication to secondary care. Substantive personal and organisational barriers remain in integrating the use of echocardiography into an integrated pathway that spans both primary and secondary care. The STEP-UP summary for echocardiography is presented in Table 14.

Skills

Diabetes is a complicated, multisystem condition. It typically proves a complex condition to integrate into diagnostic testing. However, general practice, in the UK and elsewhere, has extensive experience in managing the condition in a community setting. It is not a considerable extension to include diabetes testing in a primary care-orientated pathway of care. Dipstick tests for microalbuminuria are convenient, but their accuracy is uncertain. An Oxfordshire study compared results of single dipstick tests and sequences of dipstick and laboratory tests with a clinical testing strategy based on current guidelines to assess the accuracy and estimate costs of testing. 109 Testing strategies involving dipstick and laboratory measurements or dipstick tests had similar accuracy. The costs of using dipstick tests were, overall, lower than laboratory-based testing. The authors concluded that dipstick testing in this study did not reliably identify diabetes patients with microalbuminuria. Although dipstick testing would decrease testing costs, it could either fail to diagnose most patients with microalbuminuria or increase the numbers of patients retested, depending on the dipstick used.

STEP-UP summary statement With regard to SKILLS, diabetic services are attributed a LOW degree of implementation difficulty owing to the existence of recognised pathways and guidance.

Training

Biochemical tests can be diagnostic and often necessary for monitoring metabolic and endocrine diseases, so it is important for GPs to know which tests are useful in a primary care setting and how to interpret these tests and understand their limitations. A GMC core competency is to understand the use and main limitations of tests commonly used in primary care to investigate and monitor metabolic or endocrine disease, for example fasting blood glucose, glycated haemoglobin (HbA_1c), urinalysis for glucose and protein, urine albumin-to-creatinine ratio, near-patient testing for capillary glucose, lipid profile and thyroid function tests, and uric acid tests. 110 Blood glucose meters are a good example of POC testing, and central to result quality is high-quality training, robust internal quality control, external quality assurance schemes and effective process management. 111

STEP-UP summary statement With regard to TRAINING, diabetic services are attributed a LOW degree of implementation difficulty owing to the wide availability of training provision.

Equipment

In patients with existing diabetes, HbA_1c monitoring is usually performed every 3–6 months. It typically involves a nurse visit or phlebotomist for venepuncture, with follow-up 1–2 weeks later to discuss results. POC testing could provide more immediate therapeutic decisions and result in fewer patient visits.

Typically, the POC HbA_1c device uses a finger-stick drop of blood applied to a reagent cartridge, which is then inserted in a desktop analyser, where the analysis is performed, and HbA_1c is reported. A prospective controlled trial comparing POC testing and standard laboratory testing in an urban primary care clinic showed that POC testing availability resulted in more frequent intensification of therapy when baseline HbA_1c was ≥ 7.0%. HbA_1c fell significantly in the POC testing group but not in the standard care group. 112

STEP-UP summary statement With regard to EQUIPMENT, diabetes services are attributed a LOW degree of implementation difficulty owing to the increasing availability of POC tests and other portable equipment.

Premises

Health Building Note 11-01, Facilities for Primary and Community Care Services, locates near-patient testing services in the near-patient testing room. 46 No specific detail was identified relating to the implications of storage and maintenance of increasing numbers of POC tests.

STEP-UP summary statement With regard to PREMISES, diabetic services is attributed an UNCERTAIN degree of implementation difficulty owing to insufficient detail on storage and stock maintenance requirements.

User perspective

Diagnostics for diabetes can include POC testing, screening for diabetic retinopathy and a variety of tests for specific diabetes-associated disorders. A survey by the MaDOx (Monitoring and Diagnosis in Oxford) team found that the majority of UK clinicians would like to use POC devices for HbA_1c testing,113 but the reasons for their low uptake are unclear. In England, routine diabetes care and diabetic retinopathy screening (DRS) are principally managed in primary care, whereas treatment for retinopathy takes place in secondary care. Patients appear to confuse routine retinal photography at optometry practices during eye examinations with DRS. 114 The same study observed differences between patients screened at GP and those screened at optometrist practices, identifying that ease of making an appointment, including its time and navigating home after the mydriasis drops, etc., appeared less problematic at GP practices.

STEP-UP summary statement With regard to USER PERSPECTIVE, diabetic services are attributed a LOW degree of implementation difficulty owing to accepted pathways and standards of care.

Primary–secondary interface

Initiatives to break the reported primary–secondary divide in diabetes care include a West Sussex initiative in which services have been redesigned so that the primary care diabetes specialist nurses are employed by the PCT and work with nurses and GPs in general practice, but are based at the secondary care diabetes centre. 115 This model of care has enabled the primary care diabetes specialist nurses to function as part of the wider diabetes team, which encompasses all primary and secondary care clinicians, while not being isolated from their secondary care colleagues.

STEP-UP summary statement With regard to PRIMARY–SECONDARY INTERFACE, diabetic services are attributed a LOW degree of implementation difficulty owing to the existence of integrated pathways across primary–secondary care, notwithstanding the continuing ‘divide’ in some circumstances.

Conclusion

Diabetic services has a long tradition of being delivered as a two-tier service across the primary–secondary interface. The distinction is becoming increasingly blurred, as is the case for cardiac services (as described earlier). In examining these technologies it is most important to first define the optimal management pathway and then decide where to locate the various diagnostic interventions. A key consideration, therefore, is the existence of guidelines that make the basis for intervention, and for specific roles, manifestly clear for both patients and clinicians. Table 15 presents the STEP-UP summary map for diabetic services.

Skills

Many GPwSIs in gastroenterology will probably have developed skills or expertise in endoscopy during their hospital training. Competent GPs should be able to demonstrate that they have sufficient skills to ensure the safe and effective practice of adult endoscopy. Endoscopy procedures in primary care are typically performed on a selected population of otherwise fit patients, making comparison with secondary care unhelpful and irrelevant. A 2015 systematic review of the safety, competency and cost-effectiveness of nursing staff providing GI endoscopy services, although not examining a primary care context, reported the collective experiences of nurses working in metropolitan areas under the strict supervision and guidance of a specialist gastroenterologist. 117 Nurse endoscopists were less cost-effective per procedure at year 1 than services provided by physicians, due largely to the increased need for subsequent endoscopies, specialist follow-up and primary care consultations. However, the studies clustered between 1999 and 2001, making applicability problematic.

STEP-UP summary statement With regard to SKILLS, endoscopy in primary care is attributed a HIGH degree of implementation difficulty owing to the range of skills and competencies required, the risk of serious consequences and the time taken to acquire requisite competencies.

Training

Training for endoscopy is regulated by the Joint Advisory Group on Gastroenterology (JAG). All training must take place at JAG-approved units. Initially, all practitioners have to attend a formal training course (a JAG-compliant course) in that particular procedure. In addition, the GP should have knowledge of the 15 endoscopy workforce competencies that pertain to a primary care setting. In 2002, 96% of GPs performing endoscopy had undergone training in, and continued to work in, a consultant-led endoscopy unit. 118 The same survey concluded that simple diagnostic endoscopies could be performed safely in the primary care setting, leaving secondary care units to concentrate on patients requiring sedation, those who are acutely ill and those who require therapeutic procedures. 118 Primary Care Society for Gastroenterology guidelines suggest that no unit should offer fewer than 200 cases of any flexible endoscopy per year and the endoscopist should have continuing association with a secondary care unit for a wider exposure to case mix. Although the accreditation of a practitioner as fit to provide endoscopy services is outlined by JAG, providing endoscopy services goes beyond this and requires fulfilment of the formal GPwSI accreditation process. Support for endoscopy in primary care may be achieved through active membership of a professional society, such as the Primary Care Society for Gastroenterology or the British Society of Gastroenterology.

STEP-UP summary statement With regard to TRAINING, endoscopy in primary care is attributed a MODERATE degree of implementation difficulty owing to the availability of courses, the clear definition of a curriculum, the existence of specialist groups and the variety of training methods available that mitigate, to some degree, the wide range of skills and competencies required.

Equipment

Endoscopy ‘incurs significant cost through its high usage rate of consumables coupled with significant capital costs (eg, endoscope systems)’. 119 When endoscopy is performed in a setting outside a hospital unit, such as a cottage hospital, diagnostic and treatment centre or GP surgery, the facilities, staffing and equipment should be of the same quality. In 1994, a specification was produced for an endoscopy unit. 120 This included upper GI endoscopes and a fibre sigmoidoscope, a light source, a procedure trolley, an endoscopy trolley, three suction machines, an automated washing machine and three pulse oximeters. Anaesthetic and resuscitation facilities must be provided on site for the management of sedated patients. Adequate equipment must be available, together with safe facilities for its cleaning and maintenance. Cleaning and decontamination of endoscopes should comply with British Society for Gastroenterology (BSG) and Control of Substances Hazardous to Health (COSHH) guidelines. 121 The BSG has published recommendations in a report on aldehyde disinfection and health in endoscopy units. 122 Glutaraldehyde is most commonly used for disinfection of flexible GI endoscopes, and it is toxic, an irritant and allergenic.

STEP-UP summary statement With regard to EQUIPMENT, endoscopy in primary care is attributed a HIGH degree of implementation difficulty owing to the capital costs of purchase, the expense of consumables, the precautions required for cleaning and maintenance and the risks associated with glutaraldehyde.

Premises

Requirements for an endoscopy unit, typically based in an acute hospital, are outlined in Health Building Note 52123 and in the Users Guide to Achieving a JAG Compliant Endoscopy Environment. 124 Similarly, guidelines from the World Endoscopy Organization specify design principles for a digestive disease endoscopy unit. Health Building Note 11-01, Facilities for Primary and Community Care Services, suggests that ‘most invasive procedures and certain procedures using rigid endoscopes can take place in a generic treatment room’ (contains public sector information licensed under the Open Government Licence v3.0). 46 Endoscopy needs to be practised in a purpose-designed area with adequate space and easy accessibility. Instruments, their accessories, monitoring equipment and all the necessary drugs need to be readily to hand. The standards of care in purpose-built, well-equipped primary care premises need not fall noticeably short of those in a general hospital, particularly when staff are well trained and experienced in endoscopy. Endoscopy should be performed in a unit that complies with the JAG guidelines for safe endoscopy. In general practice any endoscopy facility is likely to be used for multiple purposes, but due consideration must be given to the care, storage and security of endoscopy equipment.

STEP-UP summary statement With regard to PREMISES, endoscopy in primary care is attributed a HIGH degree of implementation difficulty owing to the high capital costs of specific accommodation for equipment and the requirements to provide a safe environment for operation of the equipment and for cleaning.

User perspective

Consumer preference is frequently cited as a major driver for a move towards primary care endoscopy services. However, the public is relatively poorly informed of the risks, precautions and the importance of staff training and experience as they contribute to safe endoscopy. Endoscopy procedures should be performed to recognised levels of safety and accuracy, ensuring patient comfort and satisfaction. The lack of severe symptoms, fear of pain, concerns of sedation, comorbidity and competing life demands have been reported by patients as barriers to performing an endoscopic investigation. 125

In a qualitative study of UK ethnic groups, focus groups were used to explore barriers to the uptake of flexible sigmoidoscopy screening among UK ethnic minority populations. Common barriers across all ethnic groups were anxiety regarding the invasiveness of the test, the bowel preparation and fear of a cancer diagnosis. 126

STEP-UP summary statement With regard to the USER PERSPECTIVE, endoscopy in primary care is attributed an assessment of UNCERTAIN, given that attitudes of patients and staff to endoscopy in primary care are not well explored.

Primary–secondary interface

Hospital-based endoscopy depends largely on highly trained specialists. These specialists have acquired a wider clinical knowledge of gastroenterology over and above simple endoscopic interpretation. Conflicts of access and perceived vested interest are best resolved by improving communication between GPs, specialists, patient groups, purchasers and provider units. Alternatives to a GP-based community service include greater open access to hospital facilities by GPs and outreach clinics performed by hospital-based consultants in community hospitals or individual/group practices. 127

STEP-UP summary statement With regard to PRIMARY–SECONDARY INTERFACE, endoscopy in primary care is attributed a MODERATE degree of implementation difficulty given existing working relationships with secondary care but the need to refer more serious or complex cases to an acute hospital setting.

Conclusion

The potential to offer endoscopy in primary care has been long recognised and it may make an important potential contribution to the disease management pathway. It no longer holds the high level of risk that it was once perceived to have. Developments in more flexible diagnostic tools have helped to make it easier to conduct. It continues to occasion fear among patients, particularly those who are older and female. The physical requirements for the premises, for example for disinfection and storage of the equipment, may be prohibitive in an ordinary-sized general practice. However, these requirements may more easily be accommodated in purpose-built clinics or community hospitals. Table 16 presents the overall STEP-UP summary map.

Skills

Skills for genetic testing relate to understanding which patients to test, what tests to order and how to interpret the results of these genetic tests. Importantly, they also require being able to handle the counselling and communication issues associated with testing and the test result. 129 A Dutch survey130 of knowledge of genetics and genetic tests revealed general levels to be poor, with GP knowledge being poorer than that of paediatricians and gynaecologists.

STEP-UP summary statement With regard to SKILLS, genetic testing is attributed a HIGH degree of implementation difficulty owing to inadequate knowledge of genetics, challenges in interpretation and the need for skills in communicating and handling the outcome of the test.

Training

Generally, primary care practitioners receive minimal training in clinical genetics. Various methods are required to support practitioners to develop these genetic skills and knowledge, including changes to the undergraduate curriculum. The empirical evidence of the learning needs of practitioners in relation to genetic testing reveals widespread deficits in knowledge and skills, and low confidence levels. Skills-based training, clinical scenarios and assessment have all been identified as factors that can promote learning.

Primary care providers are interested in learning more about who should receive genetic testing and what tests are available. Educational efforts should build on primary care providers’ prior knowledge base, highlight the clinical relevance of genetic medicine to primary care practice and emphasise red flags: cues to alert primary care providers to a potential genetic contribution. 131 Nurses are on the front line and are expected to recognise patterns of disease that may indicate a possible genetic link, educate the family about the implications of a potential genetic susceptibility and refer the family for counselling. Many studies of nurses’ knowledge of genetics reveal gaps in professional competence and/or education. 132 The authors suggested that each nurse should, thus, acquire a minimum basic knowledge of genetics. 133

STEP-UP summary statement With regard to TRAINING, genetic testing is attributed a MODERATE degree of implementation difficulty owing to the need to receive significant training and ongoing support in offering genetic services.

Equipment

Genetic testing is expensive and time-consuming, given the sheer scale of the genes that need to be examined. However, the development of direct-to-consumer testing is likely to reduce this financial burden. A systematic review of genetic tests in primary care was published in 1999. 134 This mapping review has been unable to find a more recently published review. Specific searching for such a systematic review is recommended before considering whether or not to commission further work.

STEP-UP summary statement With regard to EQUIPMENT, genetic testing is attributed a LOW degree of implementation difficulty owing to increasing numbers of direct-to-consumer POC tests.

Premises

The mapping review was unable to identify any items that specifically described requirements to house genetic testing facilities in general practice or community premises. Health Building Note 11-01, Facilities for Primary and Community Care Services, locates near-patient testing services (such as blood and gas) in the ‘near-patient testing room’ (contains public sector information licensed under the Open Government Licence v3.0). 46 This Health Building Note does not provide detailed design guidance on specific rooms and spaces. It is unclear what tests should be provided and what the implications are with regard to the management of stock.

STEP-UP summary statement With regard to PREMISES, genetic testing is attributed an UNCERTAIN rating for implementation difficulty owing to the shortage of studies describing how tests might be accommodated. Co-ordinated approaches with regard to storage may relate to POC tests.

User perspective

Concerns are expressed about potential harm from the inappropriate use of genetic testing. It is believed that primary care practitioners are reluctant to adopt responsibility for genetic testing. In a survey of family physicians, respondents felt that genetic tests would be more useful for breast cancer and haemochromatosis than for Alzheimer’s disease, heart disease or diabetes. Individuals who believed themselves more familiar with genetic tests were more likely to anticipate that genetic testing would impact significantly on their future practice. 135

The findings from qualitative studies of patient viewpoints report misunderstandings concerning genetic tests, for example that genetic tests are more predictive than they actually are, and that they are predictive of behaviours for which no markers have in fact yet been discovered. 136 A qualitative study of genetic testing for risk of coronary heart disease found that the test was acceptable. 137 The findings from qualitative studies of provider viewpoints report that cost of testing consistently appears as the most frequently cited barrier to genetic testing. 138–140

STEP-UP summary statement With regard to USER PERSPECTIVES, genetic testing has LOW implementation difficulty for patients and MODERATE difficulty for clinicians, because patients already have access to direct-to-consumer genetic tests but GPs are cautious about the ramifications of interpretation.

Primary–secondary interface

Studies in the UK have evaluated community genetic counsellors acting as outreach workers134 from the genetics clinic to liaise with local general practices. Such practitioners are either genetic nurses141 or genetic counsellors, and they can fulfil a dual role of filtering referrals to the geneticist and providing basic genetic information. 142

Genetic nurse counsellors, specifically in a cancer genetics context, were found to perform not significantly differently from conventional cancer genetic services in two RCTs in Scotland and Wales. 143 The development of primary care specialists, possibly working in conjunction with community genetic counsellors, has been suggested to offer an intermediate point of referral between general practice and specialist genetic clinics. 142

STEP-UP summary statement With regard to PRIMARY–SECONDARY INTERFACE, genetic testing is attributed a MODERATE degree of implementation difficulty owing to a high dependence on secondary care expertise, support and follow-up provision.

Conclusion

Because of the uncertainties around the benefits and harm from genetic testing, the clinical implications of a specific genetic test will require careful evaluation, including information about cost-effectiveness, before widespread adoption is recommended. 142 With thousands of tests being developed it is important to ensure that these are matched to meaningful intervention. Rather than focusing on the organisation and service delivery of specific individual tests, the priority is to organise services that can adapt to emerging technologies, their interpretation and their implications for patient support and counselling. Table 17 presents the STEP-UP summary map for genetic testing.

Skills

To provide a MRI service requires that staff undertake sufficient current MRI practice and a sufficient number of examinations to maintain competency in every area of MRI. Staff interpreting the images and providing a clinical report should fulfil the following requirements:145

• Be a UK-registered radiologist on the GMC Specialist Register undertaking sufficient current clinical practice in that modality. A consultant radiologist must have undertaken planned regular clinical MRI sessions in their current job plan.

• Be a radiographer currently registered with the Health and Care Professions Council and have performed regular sessions of MRI examinations within the last 12 months.

• Have a minimum of 1 year’s experience.

• Maintain their CPD in accordance with their professional body guidelines.

• Meet the specification set out in the National Occupational Standards for Imaging (RD5: produce MRI images for diagnostic purposes). 145

STEP-UP summary statement With regard to SKILLS, MRI is attributed a MODERATE degree of implementation difficulty owing to its requirements in terms of experience and training.

Training

The BSCMR has endorsed voluntary accreditation procedures for their members. These have no statutory role but are provided to allow both medical and non-medical practitioners to demonstrate a specified level of experience in an appropriate educational environment. In the case of MRI, the BSCMR and the British Society for Cardiac Imaging support the criteria developed by the US-based Society for Cardiovascular Magnetic Resonance and do not separately accredit individuals.

STEP-UP summary statement With regard to TRAINING, MRI is attributed a MODERATE degree of implementation difficulty owing to the ongoing need for professional development and updating.

Equipment

Magnetic resonance imaging has entered common use as part of the standard equipment portfolio over the last 10–15 years. The following are commissioned requirements for MRI equipment that should be met or exceeded: fixed or mobile units shall contain one full-body MRI scanner with a magnetic strength of at least 1.5 Tesla; complies with the Safety Guidelines for Magnetic Resonance Equipment in Clinical Use146 as updated; and is superseded and replaced from time to time.

STEP-UP summary statement With regard to EQUIPMENT, MRI is attributed a HIGH degree of implementation difficulty owing to the prohibitive cost of an installation.

Premises

There is no specific provision for MRI in Health Building Note 11-01, Facilities for Primary and Community Care Services. 46 Commissioners will need to consider mobile or static sites. All facilities, including mobile units, are required to have a minimum of a patient reception and waiting area either on the unit or nearby, access to a toilet and access to appropriate levels of security. We found no specific guidance on housing MRI units in primary care.

STEP-UP summary statement With regard to PREMISES, MRI is attributed a HIGH degree of implementation difficulty owing to the exacting requirements for facilities and supplies.

User perspective

Because MRI equipment is typically housed in secondary care, we were unable to find qualitative research relating to attitudes to MRI specifically in a community setting. Items on perspectives of MRI provided by acute hospitals from GPs and their referred patients would indirectly inform this issue. However, these will require a supplementary search strategy.

STEP-UP summary statement With regard to USER PERSPECTIVE, MRI is attributed an UNCERTAIN degree of implementation difficulty owing to the shortage of published experience on MRI in a primary care context.

Primary–secondary interface

A MRI service needs to work with other providers to offer an integrated service. Key to the successful integration of MRI services is consideration of the management pathway, irrespective of which facilities are located in the primary or secondary sectors of local health service provision. The London NHS Diagnostic Service was established by the InHealth Group in 2007 following competitive tendering. 51 Local PCTs and federated general practice groups utilise the service, which charges for imaging services including ultrasound, plain film and MRI at standard tariff costs. This popular model of service delivery avoids the heavy capital expenditure required for purchasing the MRI equipment.

STEP-UP summary statement With regard to PRIMARY–SECONDARY INTERFACE, MRI is attributed a MODERATE degree of implementation difficulty owing to the dependence on secondary care support facilities and/or commercial providers.

Conclusion

Ultimately, MRI provision did not meet the criteria for this review in terms of services and staff managed in a primary care setting. The most common model was of service provision under contract by an external provider. This sidesteps many of the considerations from the STEP-UP framework, for example training and equipment, except from a commissioner and quality assurance customer perspective (Table 18). The implications of offering MRI services have not been explored and represent a potential line of future exploration.

Skills

To practise competently, users of POC testing equipment must have the knowledge and skills for safe and effective practice when working without direct supervision; they must recognise and work within the limits of competence, keeping knowledge and skills up to date, and take part in appropriate learning and practice activities that maintain and develop their competence and performance. 148

Standard operating procedures would include clinical governance, risk management and also implementation of risk reduction strategies. This would be achieved by careful patient selection, internal and external quality control procedures, a regular audit process, and ensuring relevant, complete and accurate documentation of the clinic process.

The internal quality control supplied by the manufacturer should be performed at the start of each clinic, and external quality assurance, provided either by collaboration with a hospital laboratory or through a quality assurance scheme, such as the National External Quality Assessment Service, should be performed regularly.

STEP-UP summary statement With regard to SKILLS, POC tests are attributed a MODERATE degree of implementation difficulty owing to the exacting requirements for quality assurance.

Training

Training of health professionals to manage POC testing should include an understanding of the specific POC device used, setting up and using the device, recording of results and quality assurance materials, health and safety, disposal of sharps, and the COSHH regulations. 121

STEP-UP summary statement With regard to TRAINING, POC tests are attributed a MODERATE degree of implementation difficulty owing to the need to identify which to use, what they mean and how to integrate them into existing clinical practice.

Equipment

As with all in vitro diagnostic equipment, new POC equipment should be thoroughly evaluated prior to clinical implementation to ensure that reliable results are consistently obtained. 149 In the UK, this is performed on a voluntary basis at the manufacturers’ discretion by the Medicines and Healthcare products Regulatory Agency (MHRA) of the Department of Health. Calibration should be performed by the manufacturers, using the same procedure as conventional laboratory systems. Calibration as performed by the manufacturers typically cannot be altered by the operator. New POC coagulometers are currently under evaluation with the MHRA and will have further field evaluations in the primary care setting. 150 A Cochrane review concluded that a POC biomarker to guide antibiotic treatment of acute respiratory infections in primary care can significantly reduce antibiotic use. 151

An illustrative STEP-UP framework assessment, for C-reactive protein (CRP), is given in Table 19.

STEP-UP summary statement With regard to EQUIPMENT, POC testing is attributed a LOW degree of implementation difficulty owing to the ease of testing and ready availability of equipment and supplies.

Premises

Health Building Note 11-01, Facilities for Primary and Community Care Services,46 locates near-patient testing services (such as blood and gas) in the ‘near-patient testing room’ (contains public sector information licensed under the Open Government Licence v3.0). This document does not provide detailed design guidance on specific rooms and spaces and refers to the following for guidance on generic rooms and spaces: Health Building Note 00-03, Clinical and Clinical Support Spaces. 153 It is unclear what tests should be provided and what the implications are with regard to the management of stock.

STEP-UP summary statement With regard to PREMISES, POC testing is attributed an UNCERTAIN rating for implementation difficulty owing to the shortage of studies describing how tests might be accommodated. Co-ordinating approaches with regard to storage may relate to genetic tests.

User perspective

Paradoxically, the greater the ease with which tests might be severed from existing secondary care domination of supply by primary care services, the correspondingly greater the likelihood of these tests bypassing primary care altogether, by going direct to consumer. GPs with different levels of experience of using diagnostic POC tests had similar perceptions that they would help to reassure patients and lead to more effective targeted treatment without alienating or upsetting patients. 13 Overall, primary care clinicians believed that POC tests increase diagnostic certainty, help target treatment, educate and empower patients, and improved the relationship between clinicians and patients by enhancing communication and shared decision-making. 13 Clinicians were also concerned about cost, over-reliance (in that POC tests could undermine clinical skills) and limited usefulness. Some issues are not unique to POC tests, but also apply to laboratory testing in general. In addition to the barriers to POC test use identified here, other reasons for lack of widespread use may include lack of needs assessments of primary health-care clinicians, resulting in discordance between the tests that they want/would use frequently, and those that are produced. A published review13 focused on blood POC tests in primary care, so further research is needed to confirm whether or not the issues raised here apply to other types of POC tests.

STEP-UP summary statement With regard to USER PERSPECTIVES, attitudes to POC testing between patients and clinicians differed, resulting in a MODERATE rating for implementation difficulty for clinicians and a LOW rating for implementation difficulty for patients.

Primary–secondary interface

In seeking to reduce variation, some PCTs are commissioning laboratories to provide a satellite service from the central laboratory to primary care sites. Testing in primary care reduces the dependency of primary care clinicians on hospital-based services, many of which are outside their own control.

STEP-UP summary statement With regard to PRIMARY–SECONDARY INTERFACE, POC testing is attributed a MODERATE degree of implementation difficulty, as it may facilitate earlier diagnosis but has a potentially unpredictable effect on demand for secondary care services.

Conclusion

Evidence so far indicates that POC testing systems have great potential for primary care with clear quality-of-life benefits for patients. Table 20 presents the overall STEP-UP summary map. The implementation of POC testing technologies requires collaboration between manufacturers, pathology laboratories and general practice, as well as adherence to a recognised external quality assurance scheme. Diagnostic technologies need to improve both accuracy and precision.

The increased availability of a test is likely to increase the usage of that test. Increased use may indicate previously unmet need, constitute an inappropriate response or simply reflect a growth in demand. In practices where desktop analysers are introduced, the rate of testing increases, but these extra tests do not inform changes in diagnosis or management. 154 However, such studies often fail to acknowledge the effect of tests in reducing the uncertainty experienced by both doctor and patient.

From the perspective of GPs, likely benefits of introducing POC tests include increased diagnostic certainty, more efficient care and fewer (re)consultations. 13 Barriers to the implementation of POC tests must be addressed, some by primary care and others elsewhere. The accuracy of POC tests in primary care populations must be addressed by manufacturers. Policy-makers and clinicians should carefully consider the role and impact of POC tests in primary care in relation to GP roles. Against a backdrop of reductions in health service funding, attention must be paid to how POC tests are to be funded. 13

Skills

Radiology is a well-documented profession with well-developed education, training and curricula. For example, the Royal College of Radiologists (RCR) produces a Specialty Training Curriculum for Clinical Radiology. 156 Training must include the equipment and its intended application. It will also include related topics such as health and safety, radiation protection and control of infection. 157 RCR guidance for clinical radiographers emphasises quality assurance issues. It would be challenging for a small radiological unit based in primary care to achieve the adherence to standards and quality control required by a fully-established radiological department.

STEP-UP summary statement With regard to SKILLS, radiology/X-ray is attributed a MODERATE degree of implementation difficulty due to the need to satisfy the expectations created by the curriculum and quality assurance procedures.

Training

Patients should have timely access to appropriate imaging provided by a service with proven clinical governance structures in place, by appropriately trained staff using appropriately maintained equipment, with underlying compliant ‘24/7’ information technology (IT) support. Comprehensive clinical imaging services in the NHS provide patients, GPs and other referrers and commissioners with direct access to essential core services. The quality of any clinical service can be judged by patient outcomes, patient safety, patient and user experience and efficiency. The RCR and the Society and College of Radiographers (SCoR) have developed the Imaging Services Accreditation Standard (ISAS) and the associated accreditation scheme. Patients, GPs, referrers and commissioners can be assured that ISAS-accredited services are delivering the highest quality of services.

STEP-UP summary statement With regard to TRAINING, radiology/X-ray is attributed a HIGH degree of implementation difficulty owing to extensive and specialist training requirements.

Equipment

One study in primary care reports that patient management by the GP changed in 60% of patients following chest X-ray. Chest X-ray substantially reduced the number of referrals and initiation or change in therapy, and more patients were reassured by their GP. This confirms that chest X-ray is an important diagnostic tool for GPs. 158

STEP-UP summary statement With regard to EQUIPMENT, radiology/X-ray is attributed a LOW degree of implementation difficulty owing to established practices and procedures for using X-ray equipment.

Premises

An imaging cluster will require such facilities as X-ray rooms, image control/reporting rooms and changing rooms. These facilities should be clustered together, alongside other dedicated imaging rooms, where provided. 159 Rooms have to meet the requirements of the Ionising Radiations Regulations 1999,160 which have much stricter limits than earlier regulations on radiation exposure to the public from man-made sources of radiation.

STEP-UP summary statement With regard to PREMISES, radiology/X-ray is attributed a MODERATE degree of implementation difficulty owing to the need for accommodation with radiation protection and shielding.

User perspective

When ordering radiological tests, clinicians in primary care consider such factors as the potential impact on the clinical outcome for the patient, the probability of significant findings based on the clinical picture and the sensitivity/specificity of the test. Urgent cases in which the report has an impact on the immediate management of the patient need to be reported at the time of examination and to be made available to the referrer. All reports should be communicated electronically via a reporting system that can push both key images and the report directly into the patient management system. The timeliness of test, and then the result, is thus a key question for patients when they are referred for an imaging test.

STEP-UP summary statement With regard to USER PERSPECTIVES, radiology/X-ray is attributed a LOW degree of implementation difficulty owing to the familiarity of X-ray arrangements.

Primary–secondary interface

The Department of Health has been seeking to facilitate imaging tailored to the needs of primary care, and thus combat a perceived secondary care imaging agenda. When GPs have access to diagnostic testing from primary care and clear referral guidelines, they are likely to use imaging resources as efficiently as hospital doctors. If implemented correctly, improving imaging access should shorten the patient pathway, facilitate better patient care and produce savings across the health-care economy. Traditional providers of NHS clinical imaging based in secondary care hospitals must seek to balance demands from both secondary and primary care for their services. A London-based group explored the effect on patient management of direct access to diagnostic imaging. 161 Three core components supported appropriate GP referral guidelines, structured referral forms and clinical triage with telephone feedback to GPs suggesting alternative tests or contraindications to testing. Thirty-two per cent of patients referred for echocardiography were found to have an abnormal report but only 29% of these were referred to secondary care, as the majority were managed in primary care.

STEP-UP summary statement With regard to PRIMARY–SECONDARY INTERFACE, radiology/X-ray is attributed a MODERATE degree of implementation difficulty owing to uncertainties about the new provider arrangements.

Conclusion

Imaging services are facing what has been described as a paradigm shift. Although the increase in competitiveness and a market orientation ultimately holds the prospect of higher standards and expectations, this is being realised on only a small local scale at present. There is considerable uncertainty in this particular area of diagnostic provision (Table 21 gives a STEP-UP summary), making it very difficult to predict what will happen in the short term.

Skills

A systematic review of 30 primary care studies from across the world162 evaluated the diagnostic ability of GPs in relation to respiratory diseases, such as acute respiratory infections, tuberculosis, asthma and COPD. In relation to asthma and COPD, studies show either overdiagnosis or underdiagnosis. With increasing affordability and recognition of its clinical applications, there is an increasing interest of the role of the oximeter in primary care. Data concerning the influence of pulse oximetry on patient management and on the extent of oximetry use in the general practice setting are scarce. Several studies identify the role and potential of the oximeter as a screening tool in assessing hypoxia in primary care. 163–166

STEP-UP summary statement With regard to SKILLS, respiratory tests are attributed a LOW degree of implementation difficulty owing to the existence of well-established testing technologies.

Training

A draft working document outlines possible roles for a GPwSI, using respiratory medicine as a model. It envisages the role of a GPwSI in respiratory medicine as primarily one of leadership and service development. A modified draft document proposes a hybrid framework that combines a generic Royal College of General Practitioners (RCGP) framework with a respiratory disease-specific framework. 167

STEP-UP summary statement With regard to TRAINING, respiratory tests are attributed a LOW degree of implementation difficulty owing to the ease of use of most equipment.

Equipment

Table 22 offers an illustrative framework assessment for pulse oximeters. Pulse oximeters are increasingly used during endoscopy and other diagnostic procedures, and as part of pulmonary function testing. Data on the role of pulse oximeters in detecting hypoxia in general practice are limited. A minority of GPs reported that they used a pulse oximeter to measure pulse rate or to assess respiratory status. 2,177

Pulse oximeters are available, highly portable and increasingly becoming less costly to purchase. 178 If pulse oximeters are being considered for a practice/community setting, it may be worth discussing this with colleagues who use pulse oximeters regularly. The NICE Centre for Evidence Based Purchasing has published an evidence review, market review and buyers’ guide for pulse oximeters, providing a list of pulse oximeters with evidence of their accuracy and performance. 179

STEP-UP summary statement With regard to EQUIPMENT, respiratory tests are attributed a LOW degree of implementation difficulty owing to portability, increasingly cheap equipment and ease of use.

Premises

Little detail is available on the implications of respiratory tests for primary care premises. If patients are being assessed in an area with a high level of artificial light (e.g. operating theatre fluorescent lighting), this can falsely reduce the readings. 180,181

STEP-UP summary statement With regard to PREMISES, respiratory tests is attributed an UNCERTAIN degree of implementation difficulty owing to lack of detail on storage and stock space requirements.

User perspective

Interviews with family practitioners174 found that they considered pulse oximetry especially valuable when they were on an out-of-office-hours shift, as a supportive tool to decide whether to send a patient to an emergency department or to refer to a medical specialist. They considered pulse oximetry an adjunct in diagnostic assessment of patients, not a full diagnostic tool by itself.

To reduce the likelihood of inaccurate readings in patients undergoing pulse oximetry, HCPs should always ensure that a patient’s nail varnish is removed, if present, that a patient’s hand is warmed, if it is cold on presentation, and that the probe is correctly positioned and clean.

STEP-UP summary statement With regard to USER PERSPECTIVES, respiratory tests are attributed a LOW degree of implementation difficulty, as they tend to be well understood and well tolerated.

Primary–secondary interface

Several countries have developed respiratory assessment units to improve the diagnosis of respiratory diseases, such as asthma and COPD, and to overcome problems with misdiagnosis. A community respiratory assessment unit was established in London to optimise diagnosis of respiratory disease by providing focused history-taking, quality-assured spirometry and evidence-based guideline-derived management advice. 182 Based in a secondary care hospital, this was a nurse-led facility, staffed by two specialist respiratory nurses with extensive experience of caring for respiratory diseases in both hospital and the community. A 4-year review found that one-third of suggested diagnoses of COPD by the GP were incorrect, resulting in the inappropriate prescribing of inhaled therapies. The authors identified significant financial, ethical and safety implications and highlighted a need for either diagnostic centres (community respiratory assessment unit) or alternative peripatetic practice-based services operating to quality-controlled standards.

STEP-UP summary statement With regard to the PRIMARY–SECONDARY INTERFACE, respiratory tests are attributed a MODERATE degree of implementation difficulty owing to dependencies on follow-up secondary care diagnosis and treatment.

Conclusion

The overall STEP-UP summary map for respiratory tests is presented in Table 23. The evidence base to support pulse oximetry is limited, but there are several areas in which further work may demonstrate benefits from the application of this technology in primary care. 183 Quantitative studies involving the rate of admission to hospital of acute respiratory illness might evaluate this device further in acute illness scenarios. Further qualitative studies could examine GPs’ experiences of the use of portable oximeters. It is conceivable that the oximeter could be used in health promotion, for example in encouraging patients to give up smoking. Although no clinician would base treatment solely on the oximeter’s readings, there is some evidence for the usefulness of pulse oximetry in general practice. It is not yet clear whether or not its use has any effect on diagnosis or patient-defined outcomes.

Skills

A study of spirometry carried out by GPs in primary care found that, in 95% of cases, fewer than five trials were required to achieve the highest-quality grade, concluding that spirometries undertaken in general practice are of acceptable quality and reproducible in only 60% of measurements. 185 Although the study involved only small numbers of diagnoses, it was concluded that the practice nurses required more training than they were originally given, again highlighting the importance of high-quality training if spirometry is to be used successfully in primary care. A study conducted with community pharmacists found that 73% of spirometries carried out by pharmacists were of acceptable quality, as judged by lung function experts in an acute setting. 186 Ongoing challenges related to the selection of a suitable spirometric screening test and to maintaining the quality of spirometric tests in primary care. A further study investigating the quality of spirometry in primary care found that > 15% of the tests being sent from primary care to specialists to analyse lacked complete data. 187 The results showed unacceptable quality in the provision of spirometry in primary care for patients with COPD, suggesting that adequate training must be given if spirometry is to be performed appropriately in primary care.

STEP-UP summary statement With regard to SKILLS, spirometry is attributed a MODERATE degree of implementation difficulty owing to the skills required to obtain a valid reading.

Training

Spirometry should be performed only by people who have been trained and assessed to Association for Respiratory Technology & Physiology, or equivalent, standards by recognised training bodies in the performance and interpretation of spirometry. Most COPD training courses include training in spirometry. Spirometry can be performed by any health-care worker who has undergone the appropriate training and who keeps his or her skills up to date. 188 However, a RCT in Australia, in which 26 intervention practices received comprehensive spirometry training and 14 control practices provided usual care, concluded that training in spirometry did not result in any measurable improvement in the use of spirometry, quality of management of asthma or patient outcomes in primary care. 189

A longitudinal study in Denmark demonstrated that improved education of staff enhanced the use of spirometry in hospital outpatients with COPD, indicating the importance of staff training. 190 Further education has also been shown to increase the use of spirometry by GPs191 and also improve their capacity to diagnose clear-cut pathologies.

STEP-UP summary statement With regard to TRAINING, spirometry is attributed a LOW degree of implementation difficulty owing to the success of short training courses for skills acquisition.

Equipment

In its Spirometer Users and Buyers Guide, the National Asthma Council Australia provides a useful list of factors to consider when purchasing a spirometer. 192 Several devices are available on the market, and one study reviewed the technical properties of 10 different spirometers designed for use in general practice. 193 Spirometers should be regularly cleaned and sterilised, as they may become reservoirs of micro-organisms. In a small study of 16 spirometers in South Australia, microbiological contamination was present in three. 194 The frequency of spirometer disinfection did not match the manufacturers’ recommendations, highlighting a need for stricter hygiene measures for spirometer maintenance in general practices. A Dutch study examined 50 desktop spirometers in general practices and found that, on average, they slightly overestimated forced expiratory volume (FEV) and forced vital capacity (FVC) values, with some devices showing substantial deviations. 195 Spirometers, therefore, need to be calibrated yearly and verified before each session.

STEP-UP summary statement With regard to EQUIPMENT, spirometry is attributed a LOW degree of implementation difficulty owing to improvements in ease of use, cost and portability.

Premises

Health Building Note 11-01, Facilities for Primary and Community Care Services,46 locates spirometry testing services in an ‘examination/therapy room’ (contains public sector information licensed under the Open Government Licence v3.0). Initial testing could occur in community locations or surgery waiting rooms, but diagnosis should be confirmed in line with recommendations from NICE and the British Thoracic Society.

STEP-UP summary statement With regard to PREMISES, spirometry is attributed a LOW degree of implementation difficulty owing to the facility to apply portable spirometry in a variety of primary care settings.

User perspective

Evidence suggests that spirometry is underutilised and that guidelines that recommend spirometry to confirm airflow obstruction among patients with suspected COPD are not routinely followed. 196 Reported barriers to the use of spirometry include poorly designed and unduly complex spirometers with too many confusing parameters of limited value, lack of availability of spirometers, poor or no teaching in medical schools and the perceived lack of an evidence base demonstrating the value and cost-effectiveness of spirometry. In a UK context only specialist registrars and GPwSIs in undergraduate education spontaneously cited spirometry as a diagnostic tool. 197 GPs stated that they now felt that they had more access to spirometry than in the past and they were highly aware of the value of spirometry, perhaps reflecting its inclusion as a quality marker in the NHS GPs’ contract. GPs in the study commented on the need for retraining, confirmed by a further study which showed that only 33% of GPs were confident in interpreting spirometry and 58% were confident in using spirometers. 198 Health-system barriers specific to spirometry use were not identified, suggesting that the availability of spirometry was not a perceived barrier.

STEP-UP summary statement With regard to USER PERSPECTIVES, spirometry is attributed a LOW implementation difficulty for clinicians and a MODERATE difficulty for patients. The equipment is becoming easier to use but may require multiple attempts for a patient to make their technique acceptable.

Primary–secondary interface

The increased availability of spirometers in primary care offers the potential for wider use. Newer spirometers are user-friendly and have the capacity for self-monitoring. With the advent of telemedicine and internet transmission of data, many more patients could have access to a diagnostic screening and/or monitoring.

Konstantikaki et al. 199 compared an open spirometry programme with a case-finding programme providing spirometry to high-risk subjects selected by primary care physicians. The proportion of newly diagnosed COPD was 27.9% in the case-finding programme, compared with 8.4% in the open spirometry programme. The average cost for a new diagnosis of COPD was €173 in the open spirometry programme and €102 in the case-finding programme. Thus, a case-finding programme involving primary care physicians was more cost-effective than an open spirometry programme for the identification of new cases of COPD.

STEP-UP summary statement With regard to PRIMARY–SECONDARY INTERFACE, spirometry is attributed a LOW degree of implementation difficulty owing to existing referral pathways for asthma, COPD, etc.

Conclusion

Spirometry equipment has frequently figured in GP practices, alongside other requisite equipment such as ECGs. It is similarly benefiting from the move to miniaturisation, as well as from demand for end-user friendly devices. However, the availability of the technology should not disguise the fact that interpreting spirometry readings, and indeed deciding to use the equipment in the first place, when readily available, remains a significant barrier to the effective utilisation of the technology. The STEP-UP summary for spirometry is presented in Table 24.

Skills

Imaging must be undertaken by trained and experienced practitioners, and, even then, perfect images may not be obtained in every patient. 201 Ultrasound images can be saved or printed for reference, but, unlike MRI scans and computed tomograms, performance and interpretation are much more dependent on the skill and experience of the operator. This limits ultrasound’s use for the skull, chest and abdomen. Ultrasound has no known long-term side effects, but it has a more limited scope than other imaging methods.

STEP-UP summary statement With regard to SKILLS, ultrasound is attributed a MODERATE degree of implementation difficulty owing to its being very operator dependent.

Training

The fact that it is safe to carry out, is relatively inexpensive and can be provided in most clinical facilities makes ultrasound one of the most commonly requested examinations in the field of diagnostic imaging. Ultrasound examinations are undertaken by practitioners from a wide range of professional backgrounds and in many different clinical settings. The National Ultrasound Steering Group of the British Medical Ultrasound Society recommended in 2008 that local clinical governance boards be established to oversee the training, supervision and audit of all providers of ultrasound imaging services. 202 Many factors affect the quality of ultrasound examinations, including appropriate training, experience, the equipment itself, clinical leadership, audit, general support and having sufficient time to undertake the examination and compile a clinically relevant report. Clear, effective clinical leadership is also essential if the ultrasound service provider is to achieve timely, accurate, clinically relevant reports. Ultrasound practitioners for whom statutory registration is not possible can apply for voluntary registration with the Public Voluntary Register of Sonographers, which is administered by the College of Radiographers.

STEP-UP summary statement With regard to TRAINING, ultrasound is attributed a LOW degree of implementation difficulty owing to the widespread availability of training courses and opportunities.

Equipment

The latest portable machines produce images that are almost the same quality as that of the larger machines; they are easy to use, durable and cost as little as £5000. 201 Images obtained with a portable ultrasound machine are usually not stored, and a decision on how to act on the scan is made at the time by the clinician doing the scan. The role of the operator is critical, and matching the operator knowledge and competence level to the equipment features is essential, to include the key machine characteristics, how these characteristics match to clinical need, features to be considered when purchasing a machine, and the associated environmental and organisational systems needed to support the efficient use of the machine. Physicians in many medical specialties are thought to be using POC scanning, but the scale on which this is happening is not yet clear; these scans are not systematically recorded in the same way as those performed by imaging experts. In 2002 a cost analysis and an assessment of quality of GP scans, based on a clinical audit and a postal survey of patients’ preferences in the Grampian region of Scotland, reported that the unit cost of a scan was higher in the practice than at the hospital. 203 However, when all of the costs of a scanning episode were considered, the total and average costs were lower in the practice because of the avoidance of hospital visits.

STEP-UP summary statement With regard to EQUIPMENT, ultrasound is attributed a LOW degree of implementation difficulty owing to its portability, its ease of use and the absence of radiation risk.

Premises

Health Building Note 11-01, Facilities for Primary and Community Care Services,46 locates ultrasound services in a generic treatment room. A key consideration is the size of the machine relative to the size of the room in which it is to be used. Other issues to be considered include the scanning couch and operator seating, the display monitor, the room heating and lighting, hygiene, infection and cleanliness, and electrical and IT provision. 204

STEP-UP summary statement With regard to PREMISES, ultrasound is attributed a LOW degree of implementation difficulty owing to increasing developments in miniaturisation and portability.

User perspective

A patient preference study for 500 patients from a GP list and 250 consecutive patients scanned at a Grampian general practice found that patients preferred to be scanned at the practice. They were prepared to wait up to an extra 5 days to enact their preference. 203 Patients were prepared to accept a reduction in the accuracy of scanning of up to 3.5% in being able to realise their choice. Patients were not concerned about which member of staff actually carried out the ultrasound scan.

Skills and willingness were not sufficient factors in themselves to prompt rural family practitioners to utilise ultrasound. 205 Economic considerations (i.e. equipment cost and remuneration) were seen as more important. 205

STEP-UP summary statement With regard to USER PERSPECTIVES, ultrasound is attributed a LOW degree of implementation difficulty owing to it being well accepted among clinicians and well tolerated among patients.

Primary–secondary interface

The number of ultrasound examinations performed by imaging experts has increased on average by 5.2% every year for the past 10 years, according to Department of Health data, and imaging experts are struggling to keep up with the growing demand. With increasing pressure on ultrasound services owing to the number of requests, changing patterns of service delivery and a shortfall in the qualified workforce, there have been concerns that the quality of some ultrasound examinations has been affected. 201 A 2002 audit of a small series of patients demonstrated that the use of an ultrasound scanner at a Grampian general practice reduced the number of hospital scans, outpatient and inpatient visits, and emergency admissions. 203 Blackpool, Fylde and Wyre Hospitals Foundation Trust extended primary care ultrasound as a NHS improvement initiative. 206 Long waiting times for routine ultrasound examination, accompanied by patients arriving late for appointments due to parking difficulty, and a limited choice for patients in terms of time and location of appointment, required an innovative service response. The trust negotiated the use of a PCT facility, off the main district general hospital site, and acquired a portable ultrasound machine to provide community-based ultrasound. Improvements included 30–40% of routine outpatient work being performed off the main site (more local scanning with better parking), improved patient satisfaction and increased patient choice of scanning venue.

STEP-UP summary statement With regard to PRIMARY–SECONDARY INTERFACE, ultrasound is attributed a LOW degree of implementation difficulty owing to its being comparatively easily implemented in primary care and the likely shifting demand from secondary care.

Conclusion

Ultrasound is one of the fastest growing areas of diagnostic technology application, excluding POC testing. There is substantive concern that this might not be translating into clinical benefit. Ultrasound, notwithstanding its portability and ease of use, carries requirements for quality assurance. These must be monitored and met if the technology is to achieve its potential clinical benefit. Furthermore, there is a need to examine the impact of ultrasound scans on the referral and secondary care pathways. The STEP-UP summary map for ultrasound is presented in Table 25.

Studies comparing/describing service models

The principal service models for diagnostic ultrasound are listed in Table 29. Rapid-access clinics may be hospital based or community based and are relevant only to certain indications, for example suspected heart failure. The other models mentioned could be applied to general diagnostic ultrasound services or services targeted at specific conditions. ‘Traditional’ access to ultrasound, through referral by a hospital clinician, is not considered in this report, which deals only with community ultrasound.

Studies describing/evaluating service models are those by Everett and Preece,222 Wordsworth and Scott,203 Salihefendic et al. ,229 Pallan et al. 16 (mobile), van Gurp et al. 223 (community open access) and Landers and Ryan234 (community palliative care) and conference abstracts by Gallagher et al. 242 and Singh et al. 244 (echocardiography clinics). Of these, only Salihefendic et al. 229 compared primary care services with and without direct ultrasound access. Unfortunately, poor reporting of the methods and results made this study difficult to interpret and greatly limited its usefulness to the review.

The other two included level 1 studies were non-comparative evaluations of services provided in the UK (England222 and Scotland203). In both studies, ultrasound scanning was carried out at a GP practice/health centre and primarily by GPs. These studies were performed in the late 1990s to early 2000s, so may not reflect current practice. As discussed below (see Table 33), the results of both studies were generally favourable for the community diagnostic ultrasound service model but, given their reported limitations, they must be considered as relatively weak evidence. However, we have not been able to find any more recent and/or rigorous research.

Pallan et al. 16 evaluated a community-based mobile service compared with a hospital-based open-access service. The community service was described as an ‘independent, radiographer-led service’. It appears likely that this service was delivered by a private sector provider rather than a NHS community organisation; this limits its relevance to the review question as originally defined as the study evaluated two ‘comparator’ interventions. Given that the grey literature search located numerous examples of services commissioned from private sector providers, Pallan et al. ’s16 study clearly addressed an intervention that remains relevant and a comparison for which further research is needed.

The conference abstracts by Gallagher et al. 242 and Singh et al. 244 both dealt with the role of echocardiography for heart failure diagnosis in the community. Both abstracts reported limited details of the models of service under study. Singh et al. 244 may be evaluating a hospital-based service and Gallagher et al. 242 may be evaluating community outreach, both of which meet the criteria for comparator interventions rather than the main focus of the review. We have not been able to locate full publications for these studies, limiting the conclusions that can be drawn from them.

Finally, Landers and Ryan234 reported an evaluation of ultrasound in the management of ascites in a community palliative care service. The authors reported that bedside ultrasound allowed scanning and subsequent treatment to be undertaken at home in many cases, which was convenient for the patients and potentially cost saving for the health service. There was an implicit comparison with a model of care based on hospital admission for investigation and treatment, but no comparative data were reported.

Other studies offering insight on service models

Level 3 papers by Robinson et al. 215 and Colquhoun et al. 246 suggest that the hospital open-access model was widely favoured in the UK in the late 1990s. If there was a general acceptance of this model, this may explain the relative lack of research on alternatives. Interestingly, the 1999 thesis by Aitken44 evaluating patient and health professional opinions on community ultrasound services pays little attention to open access as an alternative model of service and concentrates on the implications of transferring a ‘conventional’ hospital-based service to a community setting.

The results of the internet and grey literature search reported below demonstrate the existence of NHS community diagnostic/primary care centres offering ultrasound and other tests, but we have found no evaluations of these in the published literature.

Equipment

Provision of ultrasound scanning in community settings has been favoured by the development of portable devices, culminating in the availability of handheld or ‘pocket’ ultrasound. We located two reviews of handheld devices for echocardiography. 252,255 A third, more up-to-date review examines portable ultrasound devices in other indications, with echocardiography being due for consideration in a separate publication. 45 Although the use of handheld devices appears well established, this review identifies limitations of the evidence base for outcomes other than diagnostic accuracy and identifies an extensive list of research gaps that we have drawn on in our own research recommendations. The Welsh Cardiac Networks Co-ordinating Group also identified areas of uncertainty around the use of handheld scanners for echocardiography. 220

Three primary studies provided additional evidence on the use of handheld devices in community settings (Table 30). 205,247,250 The papers by Doddy247 and Mjølstad et al. 250 provide limited evidence for the effectiveness of handheld scanners in view of their small sample size, lack of a control group247 and reliance on a surrogate outcome measure. 250 There is some overlap with the literature on workforce and training as discussed in the following section. The review also included a conceptual paper on equipment for pre-hospital ultrasound in stroke. 248 The report by the Welsh Cardiac Networks Co-ordinating Group suggests equipment requirements for echocardiography services. 220

Workforce and training issues

Numerous studies examine training for GPs and others to perform ultrasound in the community (see Table 27). Further details are provided in Table 31 for level 2 studies and Table 32 for level 3 studies. The amount and level of training varies widely across studies, from short courses designed for use of ultrasound in specific indications243 to comprehensive training programmes designed for clinicians wanting to offer a wide range of ultrasound services. 245 The available research looks at training of predominantly GPs to use ultrasound technology (a study on physiotherapists236 was an exception). No data were available on training or employment of radiographers in community or primary care settings. A general limitation of many of the included studies is that they look only at one or a small number of people being trained.

Attitudes/barriers to community provision

A small number of papers, all classified as level 3 studies, addressed attitudes or barriers to provision as their main focus. 44,205,215,224,238,246 Studies from the UK/England suggest that most health professionals preferred open access to hospital services rather than facilities based in general practice. 215,246 This seems to be a key finding, although the most relevant study215 (the other source being expert opinion) dates from 1997 and is not high quality, so the findings may not be applicable to current practice. Also in the late 1990s, Aitken’s44 PhD thesis investigated attitudes to transferring ultrasound services (including monitoring of pregnancy as well as diagnostic ultrasound) to community settings. The research involved focus groups and a questionnaire study with relatively large samples of health professionals (n = 353) and patients (n = 495). Overall, health professionals regarded ultrasound as a suitable service to be provided in the community and patients would be happy to use such a service. However, given the broad scope of ultrasound services covered by the thesis, the relevance of the findings to specifically diagnostic ultrasound is uncertain. Several potential problems were identified, particularly by hospital-based clinicians. These included issues related to communication, quality control and training. Radiographers were concerned about possible isolation when working in the community.

A more recent paper by a GPwSI in ultrasound also shows some awareness of barriers to ultrasound by GPs. 247 Barriers mentioned include access to training and funding for providing the service, although the author believes that these can be overcome. Doddy argues that the perceived benefits to patients are such that all general practices should have access to diagnostic ultrasound for appropriate indications.

Another potential barrier to community ultrasound would arise if secondary care professionals displayed a negative attitude towards this type of service. In the Netherlands, Scholten-Peeters et al. 224 found that hospital doctors (radiologists and orthopaedic surgeons) had low confidence in diagnostic musculoskeletal ultrasound performed by physiotherapists or GPs. Scans performed in primary care were often repeated in secondary care, although this was not quantified. It is interesting that the findings of this recent study echo those of Robinson et al. 215 from 1997.

In a non-UK context, respondents to the survey of Siu et al. 205 identified equipment costs and remuneration as the major barriers to ultrasound in the community, followed by issues related to training. These barriers were evidently considerable, as none of the respondents was currently providing a service in this setting, although most recognised the advantages, had relevant training and were keen to do so.

Siu et al. ’s205 findings are broadly in line with the results of a qualitative study (involving interviews and focus groups) of French GPs. 238 Again, none of the eight participants was currently providing an ultrasound service in the community; seven were interested in doing so and five had received training. However, the participants identified numerous barriers: the authors summarised the main themes as lack of experience with ultrasound in general practice; mastering ultrasound scanning techniques; uncertain relevance of ultrasound scanning in general practice; the GPs’ own reluctance to use the technique; and possible legal issues.

Clinical effectiveness (includes diagnostic accuracy/yield)

All three level 1 studies reported outcomes related to clinical effectiveness of primary care ultrasound (Table 33). 203,222,229 Everett and Preece reported that GP ultrasound had a sensitivity of 97% and specificity of 98% for establishing fetal viability in women with bleeding in early pregnancy. The main clinical benefit was that women with a viable fetus could be given a strong reassurance that they were not at high risk of miscarriage. This enabled referral to hospital to be avoided, saving a journey for the patient and the cost of a hospital ultrasound scan for the health service. However, this was not a controlled study and the authors recognised the need for further research to compare different ways of delivering diagnostic ultrasound for women with bleeding in early pregnancy.

The other level 1 UK study reported that having ultrasound scanning available in general practice for abdominal, gynaecological and obstetric problems reduced outpatient referrals, hospital scans and emergency admissions. 203 This study is also limited by the lack of a control group. The results for patients using the GP ultrasound service were compared with a consensus view of how that patient would have been managed had the service not been available.

The only study with a control group that reported any outcomes related to clinical effectiveness of a community ultrasound service was that of Salihefendic et al. 229 These authors compared two primary care clinics in Bosnia-Herzegovina, with and without direct access to abdominal diagnostic ultrasound. The usefulness of this study is limited by unclear reporting but the authors stated that abdominal ultrasound was valuable in guiding patient management. It was unclear how the impact on patient management differed between the experimental and control groups.

Eight studies classified as level 2 assessed the diagnostic accuracy of primary care/community ultrasound but in most cases with some attention to wider clinical implications, making it appropriate to discuss them in this section (Table 34). 230–233,236,237,239 There were contrasting findings across this group of studies, partly explained by differences between different indications. In two studies of musculoskeletal ultrasound, accuracy was found to be poor in community practice in Australia230 and levels of agreement between community physiotherapists and radiologists were slight to moderate for most shoulder pain diagnoses. 236 These findings cast doubt on the value of community ultrasound for these conditions/settings. A further two studies examined abdominal ultrasound. An investigation of scans performed by a gastroenterologist in a US community setting indicated that the results were useful for guiding clinical management in two-thirds of cases and findings were generally supported by subsequent reference tests. 231 However, Heikkinen et al. ,232 in a long-term study of patients with functional dyspepsia, reported that repeat ultrasound rarely changes the diagnosis and hence it was not recommended.

Two studies of ultrasound for diagnosis of sinusitis in Finnish primary care identified problems caused by a high rate of false-positive diagnoses233 but, given appropriate training, GPs could diagnose as accurately as specialists,237 with the potential to reduce unnecessary antibiotic use. Two studies of echocardiography were identified. A study of a community-based open-access service in the Netherlands223 was accompanied by a critical commentary. 253 The study authors concluded that the open-access service reduced referrals and allowed more patients to be managed in primary care, although again there was no control group and comparison was against what the primary care physician reported they would have done in the absence of the service. The author of the commentary criticised the lack of connection between the open-access service and local hospitals. The commentary also stated that the ultrasound technicians might have had insufficient experience and the cardiologists involved in the study were provided with insufficient information. In conclusion, the commentary author argued that open-access echocardiography may lead to a ‘low-budget pseudo-consult[ation] of the cardiologist, whilst the patient thinks he is getting a state-of-the-art treatment’. 253

Overall, these studies indicate that community ultrasound can guide patient management and potentially reduce referrals, at least in some indications/settings. However, the contradictory results and generally poor quality of the studies, particularly the absence of true control groups, make it difficult to draw any firm conclusions.

Conference abstracts by Gallagher et al. 242 and Singh et al. 244 are discussed separately, as they deal with different service models for echocardiography (Table 35). Gallagher et al. 242 studied a community outreach clinic giving GPs access to natriuretic peptide testing followed by echocardiography in the community if required. The main outcome measure in this non-comparative retrospective study was diagnostic yield (see Table 35). The authors concluded that specialist review in the community offers the opportunity for older patients to access specialist care in a timely manner. 242 Singh et al. 244 studied a one-stop diagnostic clinic for heart failure. The setting for this clinic was probably secondary care, although it was run by a primary care physician specialist. As with Gallagher et al. ,242 the main outcome measure for the study was the diagnostic yield and there was no comparison group. The ability to provide alternative diagnoses for most patients without LVSD was seen as an advantage of the one-stop clinic over open-access echocardiography. Unfortunately, neither study appears to have been published in full, which makes it difficult to evaluate them or place them in the context of the review question.

Four studies classed as level 3 evidence reported issues related to clinical effectiveness, although these were limited. Hölscher’s review article assesses the potential use of diagnostic ultrasound in the pre-hospital treatment of people with stroke, but it is described as a concept paper and does not report any hard data. 248 Ottenheijm et al. 251 published the protocol for a randomised trial of diagnostic ultrasound-guided therapy in patients with shoulder pain. The primary outcome is patient-perceived recovery at 52 weeks. Full details of the trial had not been published in time for inclusion in the report.

Finally, two review articles should be mentioned for the sake of completeness. Senior et al. ,252 focusing mainly on accuracy and technical outcomes, compared handheld echocardiography with traditional echocardiography and BNP testing for diagnosis of BNP in the community. A review article in Spanish by Vicente-Molinero et al. 254 searched a range of sources including INAHTA (International Network of Agencies for Health Technology Assessment), PubMed and websites of scientific societies for the years 1996–2006. This broad-ranging review presented few details of the included studies and was used primarily as a means by which to identify references for possible inclusion in our review.

Overall, the evidence base for clinical effectiveness of community ultrasound is very limited and, in particular, there is a lack of controlled studies comparing different ways of delivering ultrasound services.

Evidence on process/system outcomes (e.g. waiting times)

Only one level 1 study provided evidence on process or health system-related outcomes. 203 The study by Wordsworth and Scott203 included an evaluation of the quality of GP ultrasound scans by a consultant radiologist. Fifty of 131 scans could not be evaluated, but overall the quality of scans was sufficient to allow scanning to continue at the practice, subject to the further training of GPs in specific areas. It was also recommended that one GP, rather than two, should perform ultrasound scans to achieve a sufficient volume of scanning. 203 The authors also noted that 6 of the 131 patients scanned received earlier access to treatment as a result of being scanned at the practice.

Five of the level 2 and 3 studies published in English provided evidence on process/system outcomes. Makela et al. 235 reported that, when available, ultrasound was used to diagnose maxillary sinusitis in Finnish primary care in 82–92% of cases. The paper by Pallan et al. 16 compared a radiographer-led community diagnostic ultrasound service with a hospital open-access service and found that waiting times for appointments were shorter for the community service (mean 17.44 vs. 44.53 days). Eighteen out of 23 GPs who used both services stated that waiting times influenced their decision about where to refer a patient. Pallan et al. 16 also followed the pathways of three pairs of patients referred to secondary care following an abnormal scan. The time interval between initial ultrasound investigation and referral ranged from 1 to 168 days, although the reasons for this were not reported. The authors reported that, overall, there were no systematic differences between the two services.

An important finding from the study by Scholten-Peeters et al. 224 was that, according to Dutch orthopaedic surgeons and radiologists, the diagnostic process for people with musculoskeletal problems was rarely focused on the referral diagnosis from primary care. Only 47 (22.1%) respondents stated that it was, compared with 133 (62.4%) answering negatively and 23 (10.8%) not applicable. Combined with the negative opinions of respondents about diagnostic ultrasound performed in primary care, this would suggest that the process of diagnosis may be suboptimal, although it is unclear if the results are likely to be generalisable to the UK setting.

In the Canadian health system, Siu et al. 205 reported, based on a survey with a 44% response rate, that family physicians were not currently using ultrasound in community office settings. This was attributed in part to the availability of ultrasound technicians locally to perform scans following referral. The authors noted that bedside ultrasonography performed by a physician would most often be focused on a particular clinical question and could be used to inform a decision about referral for further investigation (including ultrasonography).

Process and/or system outcomes reported in van Gurp et al. ’s223 study of community echocardiography in the Netherlands included waiting times and GP adherence to the cardiologist’s advice and to Dutch clinical guidelines for heart failure. Waiting time for echocardiography via referral was estimated at 5 weeks, compared with 6 days via the community open-access service. GPs followed specific advice from the cardiologist evaluating the scan in 25 out of 31 cases (81%). In one case the advice was not followed and in five cases the outcome was not known. Adherence to clinical guidelines was considered suboptimal, in that 13 out of 55 patients with suspected heart failure were referred to echocardiography without evidence of an abnormal ECG or BNP result.

A study published as a conference abstract243 and two studies published in Spanish240,241 focused on training GPs for handheld echocardiography241,243 or abdominal ultrasound. 240 The studies are discussed above, but in terms of health system outcomes, they indicate that training programmes can be delivered successfully at different centres and levels of intensity. By contrast, the barriers to GP ultrasound in the French health system identified by Blanchet and Thierry238 included uncertainty about various process- or system-related factors, for example impact on length of consultations and relationships with secondary care specialists.

Patient acceptability

The four studies that reported on patient views of ultrasound in the community all reported positive findings. 16,44,203,247 The studies used different methods to assess patients’ views. Wordsworth and Scott carried out a discrete choice experiment which showed that patients strongly preferred scanning at the GP surgery to the hospital alternative. This study was performed in a rural area of Scotland and the hospital was 30 miles from the surgery, so the results may not be generalisable to more urban areas. Patients showed no strong preferences as to who performed the ultrasound scan (e.g. the GP or the radiographer). 203 In the study by Pallan et al. ,16 patients (and referring GPs) were asked to rate their satisfaction with a community ultrasound service and a hospital open-access service on a 1–5 scale. Eighty-six per cent of patients rated the community service at 4 or 5 (the highest levels of satisfaction) compared with 76% for the hospital service. GPs were markedly more satisfied with the community service than with the hospital service. This was the only study to compare the views of users of different types of service.

Turning to level 3 evidence, the majority of patients who took part in Aitken’s focus groups and surveys indicated that, hypothetically, they were happy to use a community ultrasound service. 44 Percentages of patients responding positively in the surveys were 76% in a pilot survey and 72% in the main survey. Focus groups identified convenience and improved access as key reasons for preferring community ultrasound. Finally, Doddy states in his opinion article that his patients are ‘delighted to be scanned at the surgery’, although no evidence is presented to support this. 247

Costs/resource use

Two included studies reported on costs or resource use. 16,203 Wordsworth and Scott performed a cost analysis of GP versus hospital ultrasound scanning. The main finding was that cost per scan was higher for GP scanning (£36.37 vs. £20.32) but the average cost per episode was lower (£148 vs. £183). Pallan et al. 16 calculated the cost per abnormality detected for community and hospital ultrasound services and found that the cost was higher for the community service [£107.69, 95% confidence interval (CI) £90.61 to £132.71 vs. £77.35, 95% CI £63.76 to £98.30]. The authors argued that reduced waiting times and high patient and GP satisfaction could justify the higher costs for the community service.

Published articles

Two published articles not identified by the database searches234,247 were picked up by the internet search and included in the synthesis below. An ongoing systematic review of ultrasound for musculoskeletal disorders in primary care was identified from the PROSPERO systematic reviews register. 257 Other online articles were retained as useful background papers. The remaining articles were judged not relevant to the review or added little to the evidence available from the peer-reviewed literature.

Companies providing diagnostic ultrasound services

The searches revealed details of services provided by 22 companies, most of which provided a range of diagnostic (and in some cases clinical) services in addition to ultrasound (see Appendix 8). However, one of these appeared to be not mainly focused on ultrasound (InHealth), one was mainly serving hospitals rather than community settings (Independent Vascular Services), and one appeared to be an exclusively private service (The Ultrasound Centre). Details provided by the company websites were highly variable. Commonly reported features include time standards for appointments and report delivery and patient satisfaction ratings. Links to the NHS, including experience of the radiographers and others providing the service and roles of NHS consultants in audit and governance, were emphasised by most of the companies. It appeared that many staff worked part-time for the companies and the remainder of the time for NHS organisations (primarily hospital trusts).

Services were described as being generally commissioned through the ‘any qualified provider’ system. All services are regulated by the Care Quality Commission (our search found one Care Quality Commission inspection report). All of the services appeared to follow basically the same model of care, namely a mobile service operating through GP surgeries and other community sites and staffed by radiographers (and health-care assistants), with access to consultant radiologists as required.

NHS service specification

The internet search also found a standard NHS specification for non-obstetric ultrasound,258 covering the following indications:

• general abdominal: includes assessment of the aorta, biliary tract, gallbladder, inferior vena cava, kidneys, liver, pancreas, retroperitoneum and spleen

• gynaecological: including transabdominal and transvaginal

• renal/bladder/prostate

• scrotal/testicular

• musculoskeletal

• vascular.

There are various exclusions from the service specification, including scans for suspected cancer, ultrasound-guided procedures and obstetric care, as well as parts of the body not listed above. The NHS service specification also provides for scans to be undertaken in secondary care where this is an integral part of a clear clinical pathway. 258

Referral to a specialist

Referrals to specialists by GPs depended on the medical condition and the investigations required for a diagnostic work-up. For example, echocardiography was reported to assist in identifying cases of heart failure. However, Khunti et al. 82 reported a lack of systematic diagnostic work-up among GPs for this condition, perhaps due to a lack of awareness regarding the usefulness of echocardiogram. Many GPs preferred ECG and referral for an X-ray because it was ‘easier to organise’ and ‘less expensive’ than accessing echocardiography, particularly for non-fundholding practices. There was also a reported lack of trust at the interface between primary and secondary care. Some single-handed GPs referred their patients to a rapid-access cardiology unit at the hospital, whereas others requested access through cardiologists. Lack of time to assess patients for heart failure was cited as another barrier in primary care diagnostics, particularly as the population with suspected heart failure are elderly and can require more time to undress for physical examinations.

For patients with breathlessness and accompanying symptoms suggestive of lung cancer, the process of specialist referral was rapid once the symptom had been recognised by the GP, possibly because of greater awareness of recent UK guidance for cancer symptoms. 295,296

Office spirometry

Spirometry was one investigation that could be conducted in the GP office and was particularly associated with the differentiation between a diagnosis of COPD and one of asthma. A number of included studies explored the perceived usefulness of office spirometry from the point of view of the clinicians involved. No included study was identified that looked at patient views and experiences of spirometry in this setting.

In the four qualitative studies exploring GP views about performing spirometry, there were contradictory findings. Joo et al. 198 reported that GPs felt confident in diagnosing COPD without the use of spirometry, particularly if risk factors such as smoking were apparent and trialled treatments for COPD appeared to be working. The results from referred spirometry were reported as confusing for GPs, furthering their confidence in using non-spirometric diagnostic strategies. Practical considerations were also reported, such as lack of time to perform spirometry297 and the patient’s unwillingness to return for follow-up investigations. Goeman et al. 298 found that Australian GPs felt uncomfortable about their ability to use spirometers properly and that few practices had access to a spirometer. These findings highlighted a gap between national guidelines and current practice. However, Roberts et al. 197 found that GPs regarded spirometry as an essential diagnostic tool, albeit with acknowledgement of the need for training and retraining. This study sampled GPs in the UK who may be incentivised by criteria in the new GMC contract. They also had a special interest in medical education, and, therefore, their perceptions of spirometry use might differ from those of the broader UK GP population.

Five included cross-sectional studies assessed the use and perceived usefulness of spirometers in the diagnosis of COPD and/or asthma. Boffin et al. 303 found that although 38% of Flemish GPs had access to a spirometer, 66% had never or almost never used one. One in five GPs was using a spirometer at the time of the study and a similar number had used spirometers but stopped doing so. Seventy per cent were using a peak flow meter, suggesting that GPs recognised the importance of assessing lung function. Reasons for not using spirometers or stopping spirometer use included the time it takes during the consultation, and a lack of knowledge and skill. Eighty-six per cent of GPs responded that training in spirometry ought to be provided. Of users, 71% had attended educational sessions, of which 64% were organised by vendors of spirometers. A study carried out in Italy305 concluded that GPs probably only diagnose the most severe cases of COPD and that 70% of these will have used spirometry. They suggest that this could be improved by offering better educational programmes. Following a workshop that included a presentation and hands-on experience of how to use a spirometer in a US-based study,191 the main reasons given for not carrying out spirometry were reimbursement, time, lack of familiarity with testing and uncertainty with interpretation of results. The authors point towards a lack of awareness or acceptance of the value of lung function testing among primary care practitioners.

By contrast, two studies304,306 found that access to spirometers and perceived confidence in diagnosis using a spirometer was higher in the UK. Over 80% of practices in one study304 reported possessing a spirometer and 160 out of 187 of these used it in every case of suspected COPD. However, 44% rarely used a spirometer. Over half of those using spirometry were confident in its use, and these practices used spirometry more than those who reported less confidence. However, only one-third of using practices were confident about interpreting the results. Most spirometry was carried out by a nurse, with only one-third carried out by a GP. Halpin et al. 306 reported that practice nurses were less confident (55%) in diagnosing COPD than GPs (80%), although GP confidence had increased from 2001 to 2005, whereas that of practice nurses remained the same at the end of this period. Bolton et al. 304 concluded that reported misdiagnosis of COPD could benefit from more widespread use of spirometry. More GPs than practice nurses reported feeling confident in differentiating COPD from asthma.

For patients, diagnosis was received with mixed emotions, and the way that the diagnosis was conveyed by HCPs was important; for example, there were reports of a lack of empathy. GPs might have come across some conditions, such as pulmonary hypertension, only rarely, and this might also be the first time that a patient had heard of the condition. Patients often felt ill-informed, returning home to look for information on the internet, with upsetting consequences. 293 Similarly, Khunti et al. 82 found differing views between GPs within practices about diagnosing heart failure, as well as diagnostic confusion about whether a condition was respiratory or cardiac related. In another study, continuing education for GPs to promote better care was a priority for Australian GPs. 298