LEARNING OUTCOMES
After completing this module the paramedic will be able to:
Paramedics are being employed in increasingly diverse clinical areas. One such area is primary care, where general practitioner (GP) surgeries are employing paramedics to undertake urgent work and, in some cases, consult with patients in non-urgent appointments (Primary Care Workforce Commission 2014). The Primary Care Workforce Commission (2014) identifies that paramedics may have a key role to play, as part of a multidisciplinary team, in improving the delivery of primary care in GP surgeries. Ball (2005) highlighted that specialist paramedics were employed in minor injury units, intermediate care teams and in out-of-hours GP services delivering urgent, unscheduled care. It seems that this area of paramedic practice is growing rapidly and, as such, many paramedics now need to be aware of common illnesses among the population to tailor their responses to a changing healthcare system.
Acute cough is one of the most common illnesses in the UK with an estimated 48 million cases per annum (Morice et al, 2006). It is defined as a cough lasting no longer than 3 weeks, and is one of the most common reasons to seek medical advice in primary care. While the majority of acute cough presentations are benign and self-limiting, viral and bacterial infections of the respiratory tract have the potential to develop into more serious conditions such as community acquired pneumonia (CAP). The majority of patients with CAP can be safely managed in the community; however, a significant proportion will require more intensive treatment, and mortality rates are estimated to be as high as 7% (Cilloniz et al, 2011). In recent years, there has been a notable trend to admit elderly patients to hospital who are diagnosed with CAP. Early identification of the disease is therefore considered important for targeted management in the community (Woodhead et al, 2011). This presents a challenge for clinicians who have to differentiate between benign self-limiting respiratory illness and more complex bacterial infection within a primary care setting.
While the precise aetiology is unknown, it is believed that the majority of acute coughs are initially caused by viral nasopharyngitis, otherwise known as the common cold (Woodhead et al, 2011). In a healthy adult, this condition is considered to be a benign and self-limiting virus, therefore requiring minimal healthcare intervention at first presentation. In vitro studies have shown, however, that the initial viral infection creates favourable conditions for bacterial growth, making secondary bacterial infection more likely (Peltola and McCullers, 2004). The relationship between viral and bacterial respiratory tract infections (RTIs) is further emphasised by Cilloniz et al (2011), who identified that 29% of pneumonias were of mixed viral and bacterial aetiology.
It is therefore reasonable to assume that RTIs with mixed aetiology will present with a combination of viral and bacterial symptoms, making clinical diagnosis, and therefore targeted management, more problematic. This theory is well supported in the literature. Wipf et al (1999) and Metlay et al (1997) highlight the difficulty clinicians have in accurately diagnosing chest complaints with clinical signs and history alone. Additionally, Hopstaken et al (2005) and Huijskens et al (2014) conclude in microbiological studies that there is significant overlap of symptoms for both viral and bacterial presentations of lower respiratory tract infection (LRTI). From these studies, we can surmise that clinicians' ability to differentiate between bacterial LRTI, viral LRTI and bronchitis using clinical assessment and history-taking alone is insufficient to accurately formulate a targeted management plan. This conclusion is supported in a comprehensive review commissioned by the European Respiratory Society (ERS) and The European Society for Clinical Microbiology and Infectious Diseases (ESCMID) (Woodhead et al, 2011).
Acute cough and antibiotics
Since 1964, antimicrobial resistance (AMR) has been identified as one of the major threats to modern medicine prompting national governments and scientific communities to design strategies to combat it. In 2013, the UK Department of Health and Social Care (DHSC) released a 5-year plan, which identifies optimising prescribing practice as one of their main seven points of action (DHSC, 2013). These sentiments are echoed on a continental and global level by the European Commission (2011) and the World Health Organization (WHO, 2011), highlighting the global threat posed by AMR. Furthermore, Genomics to Combat Resistance against Antibiotics in Community Acquired Pneumonia (GRACE) has been specifically commissioned to focus on the management of patients with suspected LRTI.
While the ambiguity of symptoms associated with acute cough presentations is considered a major factor in the over-prescription of antibiotics (Woodhead et al, 2011), rising expectations of the patient have also been cited in the literature to contribute significantly (Coenen et al, 2006). Coenen et al (2006) report that GPs felt it was less appropriate to not prescribe antibiotics if the patient had requested them, and McNulty et al (2013) demonstrated that 97% of patients who requested antibiotics from their GP received a prescription for their complaint. Additionally, 23% of these prescriptions were prescribed without any further discussion about their illness with the doctor. While these are isolated studies, they raise questions about current management strategies used by GPs when faced with demanding patients.
Evidence suggests that patients are not well informed to make appropriate decisions regarding antibiotic prescriptions for acute cough symptoms (Cals et al, 2007; Coenen et al, 2013). It is therefore essential to implement a more robust decision-making strategy for primary care clinicians to manage patients presenting with acute cough.
CRP testing as a solution?
The problematic diagnosis of pneumonia, increasing concern over AMR, and high incidence of side effects associated with antibiotics have prompted research into how modern healthcare systems can safely reduce antibiotic prescriptions for acute cough presentations. One of the methods being considered in the UK is the use of point-of-care testing (POCT) for detecting raised C-reactive protein (CRP) in the patient's blood. CRP is an acute phase reactant protein synthesised by the liver. It is produced in response to inflammation, infection or trauma as part of the immunological response (Riodan et al, 2009). Levels of CRP in blood serum begin to increase 4–6 hours after the initial insult or infection, and peak after 36–50 hours (Riodan and McWilliam, 2009). The technology to provide POCT for CRP is a relatively new development in primary care and has been shown to be as accurate as laboratory analysis by microbiologists (Seamark et al, 2003; Kotani et al, 2014).
CRP testing has been used in hospital settings for the last 50 years to monitor patients with acute infections and their response to treatment (Cooke et al, 2015). It seems prudent to explore the sensitivity and specificity of CRP in diagnosing CAP.
Falk and Fahey (2009) conducted a comprehensive diagnostic accuracy systematic review of CRP in relation to CAP for patients both assessed in the community and emergency departments. They concluded that CRP measurements alone were insufficient to rule out CAP in a primary care setting, although they do concede that if the patient displays symptoms relating to CAP, CRP may have some diagnostic value. This is further qualified by Vugt et al (2013), who demonstrated that patients with a CRP of under 20mg/litre had a 3% chance of having radiographically confirmed pneumonia in primary care. This subgroup of patients was more likely to be on long-term steroids, which is a factor known to reduce CRP readings (Vugt, 2013).
In terms of ruling in CAP with CRP, Falk and Fahey (2009) report that their findings were less clear in their meta-analysis. Many of the studies that met the inclusion criteria were deemed to have heterogeneous data and therefore require further research before drawing robust conclusions. Individual studies such as Almirall et al (2004), Muller et al (2007), Bafadhel et al (2011) and Espana et al (2012) do, however, demonstrate high sensitivities and specificities in diagnosing patients with CAP presenting to emergency departments with varying cut-off CRP levels.
Guidelines published by the GRACE consortium and the National Institute for Health and Care Excellence (NICE) indicate that they concur with these conclusions, specifying that CRP of over 100 mg/litre implies a high probability of CAP (Little et al, 2013; NICE, 2014). Despite these promising studies and acceptance by large institutions, it is of note that the majority of evidence for ruling in CAP has been conducted in a hospital setting. It may be argued that patients presenting to emergency departments will have more severe symptoms than patients presenting in primary care, increasing the overall probability of having CAP.
This is highlighted in one of the only studies to be conducted in primary care; Vugt et al (2013) demonstrated that only 35% of patients with CRP of over 100mg/litre were later radiographically confirmed to have pneumonia. Further research is therefore needed in patients presenting in primary care to assess the validity of POCT CRP for ruling in CAP.
POCT for CRP has already been introduced into primary healthcare systems in the Netherlands and Nordic countries to aid clinicians in their decision making in relation to acute cough symptoms (Cooke et al, 2015). The majority of the research from these healthcare systems has focused on how effective CRP tests are at reducing antibiotic prescriptions in primary care.
In a relatively large cluster randomised control trial (RCT) in the Netherlands, the use of CRP POCT was shown to reduce antibiotic prescriptions by 22% (Cals et al, 2009a). This study excluded patients seen at home or in nursing homes, which resulted in a sample with a lower mean age than would normally have been expected. Although one could consider that this subgroup of patients at home is more likely to be prescribed antibiotics due to their age and comorbidities, the study still demonstrated a significant reduction of prescriptions for those attending the surgery.
Little et al (2013) conducted a multinational study across several countries in Europe, including Great Britain, on behalf of the GRACE consortium. They were able to demonstrate a 15% reduction in antibiotic prescriptions after CRP POCT measurements were introduced into primary care. Further studies were included in a large meta-analysis by Huang et al (2013), showing a mean reduction of 18.9% in antibiotic prescriptions for patients assessed with CRP measurements. This is by far the largest meta-analysis conducted to date, but the review has received criticism for several methodological flaws, and therefore, may not be sufficiently valid to draw any robust conclusions (Aabenhus et al, 2014).
A majority of these studies were conducted in European healthcare systems; hence, their antibiotic prescription practices differ from those in the UK. It is important to note however, that no significant adverse events, increase in mortality or admissions were reported in any of the trials, indicating that the use of CRP POCTs and reduction in antibiotic prescriptions may be safe in primary care. Conversely, Engel et al (2011) question whether the evidence in primary care is valid enough to draw this conclusion, with several of the studies being methodologically flawed.
Clinicians' perspectives
These studies show a significant reduction in antibiotic prescriptions, but clinicians have voiced concern about several aspects of CRP testing in primary care. Wood et al (2011) conducted a multi-country qualitative study to gather clinician and patient views surrounding POCT for CRP. The primary concerns for clinicians included questionable accuracy of tests and over-reliance on CRP results, thereby detracting from the clinician's clinical skills.
Clincians were concerned about some of the limitations identified in the literature. CRP increases during an infection, but it also rises in response to inflammation and trauma, causing false positive results (Pepys and Hirschfields, 2003). Chronic diseases like Crohn's disease, osteoarthritis and rheumatoid arthritis are common examples where CRP can be chronically raised, making interpretation of CRP results even less specific to acute cough (Pepys and Hirschfields, 2003). Moreover, CRP levels within the blood do not peak until 36–50 hours after initial infection, which could result in false negative results on presentation.
False negatives are a concern as bacterial infection could be dismissed by the clinician on the basis of the CRP result only for a bacterial infection to develop later without an antibiotic cover (Wood et al, 2011). Despite these shortfalls, clinicians working in primary care with POCT for CRP held positive views about using CRP to guide clinical decision making (Wood et al, 2011). This was further reported by Cals et al (2009b) and Anthierens et al (2014), where clinicians felt that the CRP test decreased clinical uncertainty and supported non-prescribing decisions.
Patients' perspectives
Patients were generally positive when asked about the inclusion of POCT for CRP in primary care. Wood et al (2011) report that patients felt that the test gave the clinician a better chance of accurate diagnosis and, therefore, a more accurate management plan. These sentiments were echoed by Jones et al (2013), where clinicians described enhanced relationships with the patients and a more inclusive decision-making process. Findings from Cals et al (2013) suggest that enhanced patient understanding and stronger patient-clinician relationships as a result of POCT for CRP may contribute to a reduction in follow-up visits post LRTI. The authors concede that this trend is not statistically significant; however, it stands to reason that patients will gain a better understanding of their own illness in the context of their CRP result. Further research is needed to explore whether or not CRP testing is influential in patient behaviour, specifically over patients' decisions to seek a consultation.
Financial feasibility
In a climate of economic uncertainty and budget-cuts to healthcare systems around the world (John and Price, 2013), the cost effectiveness of POCT for CRP must also be considered. In a costing statement, NICE estimated the total initial spend on equipment for GP surgeries in England as £3.8 million. Each subsequent test would cost £13.50 when staff time, reagents and calibration are taken into consideration (NICE, 2014). Hunter (2015) investigated the potential cost effectiveness of CRP POCT in the UK by analysing data collected by Cals et al (2013), Huang et al (2013) and Little et al (2013). The cost of prescriptions, adverse events, training costs, and quality-adjusted life years (QALYs) were all considered over a 3-year period. While the data from these studies have primarily been acquired from European healthcare systems and therefore may not be directly applicable to UK prescription practices, Hunter (2015) concludes that CRP POCT would be cost-effective within the NHS. In the study, the initial costs of CRP testing were outweighed by reduction in prescriptions, reduction in attendances and increased QALY outcomes. The increase in QALY outcomes post-intervention were largely based on a reduction of follow-up attendances for acute cough over a 3-year period (Cals et al, 2013). This reduction may have been caused by patients perceiving that they would not receive any antibiotics due to the new CRP procedure; therefore, they may have chosen not to attend despite having cough symptoms. If this was the case, it would have reduced their QALY scores, thereby affecting the outcome of the study. Additionally, the study uses data from Huang et al (2013) which, as discussed, has been criticised for its data collection techniques in the literature (Aabenhus et al, 2014).
A second cost analysis was conducted in Norway and Sweden on behalf of the GRACE consortium, again using parameters such as QALYs, cost of testing, and cost of prescriptions for analysis (Oppong et al, 2013). The study concluded that using CRP POCTs significantly raised the overall costs of managing patients with LRTI; however, this was deemed acceptable based on an increase in QALYs. The original data used in this study were sourced from Butler et al (2009), of which the primary outcome measure was related to prescribing trends across Europe and not primarily concerned with CRP testing. The practices from Norway and Sweden had already established the use of CRP POCT in primary care and only displayed a reduction in prescriptions of 5% between the CRP test group and the no-CRP test group. Studies based in countries that had previously not had CRP testing have demonstrated a larger reduction of 15–22% of prescriptions (Cals et al 2009a; Little et al, 2013), which is likely to result in greater cost savings.
One aspect that appears to have been overlooked by both of these economic evaluations is the potential reduction in radiological imaging which was demonstrated in a Russian study. Andreeva and Melbye (2014) reported a reduction of referral for chest X-rays of 20% in patients assessed with CRP POCT in primary care, with no adverse outcomes. This could have significant implications for reducing unnecessary radiological exposure, reduced transportation costs to secondary care centres, and the cost of the imaging itself. The studies also concede that it is difficult to quantify the economic value of reducing prescriptions in relation to AMR. In theory, reducing AMR will have significant implications for cost savings. Targeted therapy and early detection of disease may result in reduced admissions, less repeat antibiotic prescriptions and fewer investigations.
Conclusion
The use of CRP POCT within the primary care environment in the UK remains controversial. The accuracy of CRP for ruling in or ruling out CAP in primary care is not well defined at present, although hospital-based studies would suggest that it may be of value in the presence of other clinical signs. The delay of CRP reaching its peak levels, and false-positive or false-negative results are some of the other limitations of the test in primary care. Despite these limitations, current evidence suggests that CRP testing can significantly reduce prescriptions for acute cough presentations within primary care. These results were achieved with a few reported adverse events, which would indicate that CRP POCT is relatively safe to implement. The majority of the research was primarily conducted in European healthcare systems, and therefore, prospective randomised controlled trials are required in the UK to draw any definitive conclusions.
Clinician and patient perceptions of the intervention were primarily positive, and indicate that it could strengthen the clinician-patient relationship. This could have further implications on how often patients choose to seek professional advice for acute cough symptoms, although further research is needed to confirm this. Whether the use of CRP POCT is cost-effective is also debated in the literature.
The studies included in this article did not attempt to quantify the economic benefits of reducing the incidence of AMR and overlooked the benefit of reduction in imaging in their analysis. Taking these factors into account, it is likely that CRP POCT will be cost-effective in primary care, although data acquired from a UK-based study would be beneficial for further analysis.