Diagnostic accuracy of point-of-care tests for acute respiratory infection: a systematic review of reviews

Notes

Permissions

Background

Rates of acute respiratory infection have increased since the COVID-19 pandemic. 1 In response, NHS England has established new acute respiratory infection (ARI) hubs and ARI virtual wards. 2,3 These are intended to reduce pressure on other parts of the health service by providing care for people with respiratory infections. This review was prepared to inform national guidance on the initial assessment and management of acute respiratory infection in people aged over 16 in England, published by the National Institute for Health and Care Excellence (NICE). 4 The NICE guideline (NG237) has a broader scope than the present review and the reader should refer to that document for recommendations from NICE.

Aims and objectives

This systematic review aimed to determine the accuracy of the following tests in adults (>16 years) who present in an acute care setting:

1. symptoms and signs to diagnose bacterial respiratory infections

2. rapid, point-of-care tests to diagnose bacterial or viral respiratory infections

3. rapid, point-of-care tests to diagnose influenza and respiratory syncytial virus (RSV).

Methods

Details of the protocol for this systematic review were registered on PROSPERO and can be accessed at https://www.crd.york.ac.uk/prospero/display_record.php?ID=CRD42023427097. There were no changes to the protocol during the review process. The principal approach used was an overview of systematic reviews.

Results

Systematic reviews

Results of the search

The systematic search for potentially relevant systematic reviews found 4450 references. The full texts of 163 articles were retrieved for closer inspection; 23 of these studies met the inclusion criteria for this review (see Appendix 2, Table 2 for a summary of these studies). The Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) flow diagram is shown in Figure 1.

FIGURE 1.

Identification of relevant systematic reviews. ^aSystematic review that searched only one database, or did not provide an assessment of methodological quality for included studies.

Six relevant systematic reviews were identified as being most aligned with the scope of this overview. Details of these six reviews are reported in Table 1, and the summary risk of bias assessment is shown in Figure 2. Details of all reviews and publications excluded after full text assessment – along with the main reason for exclusion – are given in Report Supplementary Material 1, Table S1.

TABLE 1 - Included systematic reviews

COPD, chronic obstructive pulmonary disease; CT, computed tomography; MRI, magnetic resonance imaging; MxA, myxovirus resistance protein; PCR, polymerase chain reaction; RT PCR, real-time polymerase chain reaction.

Reference	Population	Clinical features	Setting	Target condition assessed	Index tests	Reference standard
Carlton 202132	Adults and children. Different analyses included different populations	People presenting with symptoms of acute respiratory tract infection	Primary, emergency or secondary care	Bacterial respiratory tract infection and viral respiratory tract infection	Combinations of biomarkers (at least two included). TRAIL, IP-10 and CRP (ImmunoXpert) CRP and MxA (FebriDx) CRP and neopterin	Any reference standard, including expert consensus, clinical algorithms and microbiology
Gentilotti 202233	Adults and children. Where possible, summary (subgroup) estimates were extracted which relate to adults only	Symptoms consistent with acute respiratory infection	All included studies relating to primary/emergency care settings, including primary care, emergency department, outpatient clinics and long-term care facilities. Where possible, summary (subgroup) estimates were extracted to show the effect in these different settings	Bacterial pneumonia and influenza	Symptoms and signs, host biomarkers (CRP and procalcitonin) and single pathogen tests for influenza	Any reference standard was permitted including X-ray, bacterial or viral culture, PCR, rapid antigen tests, lung ultrasound, composite analyses, expert opinion, microbiological diagnosis and rapid influenza tests
Minnaard 201734	Adults	Suspected lower respiratory tract infection	Primary healthcare, ambulatory care or emergency department settings	Pneumonia	Combination of symptoms and signs plus CRP measurement	Chest X-ray
Onwuchekwa 202335	Adults and children. Extracted data relate to adults only	No information provided	Primary care, emergency care and hospitalised participants	RSV	Direct immunofluorescence and rapid antigen tests	RT PCR
Pazmany 202136	Adults with COPD	Presenting with an acute exacerbation of COPD	Primary care, emergency care and hospitalised participants	Bacterial acute exacerbation of COPD	Presence of purulent sputum	Microbiological culture
Schierenberg 201737	Adults	Immunocompetent adults who self-referred with an acute or worsened cough or lower respiratory tract infection	Primary care, ambulatory care or emergency departments	Pneumonia	Combinations of symptoms and signs (clinical prediction models)	Chest X-ray, CT or MRI

FIGURE 2.

ROBIS assessment for included systematic reviews. ^aRated at high risk of bias for synthesis and findings as some prespecified subgroup analyses were not reported, no information was available regarding heterogeneity, and the high/unclear risk of bias in many primary studies was not addressed in the synthesis. ^bA single author was involved in sifting studies for eligibility.

Symptoms and signs for the diagnosis of bacterial pneumonia

Four recent systematic reviews were identified which assessed the accuracy of individual symptoms and signs or combinations of symptoms and signs in diagnosing bacterial pneumonia. 33,34,36,37 The total number of studies included in each review ranged from 834,37 to 421. 33 The reviews were all considered to be at low risk of bias overall, although we had some concerns regarding the synthesis for one review33 (as the high risk of bias in the primary studies was not addressed in the synthesis, there was no information on heterogeneity and some subgroup analyses were not reported) and some concerns over the identification of studies for two other reviews34,37 (as a single author was involved in sifting studies). Of note, none of the reviews explicitly stated that case-control studies were excluded – the inclusion of such studies may result in overestimates of diagnostic test accuracy.

The reviews included primary studies of participants with symptoms of acute respiratory infection,33 cough or lower respiratory tract infection34,37 or an exacerbation of COPD. 36 One review included both adults and children33 but presented some subgroup data for adults and children. Where possible, we extracted data which related exclusively to adults. All of the studies specifically included participants in appropriate settings (primary, ambulatory or emergency care settings), although one review also included some hospitalised participants. 36 The target condition was pneumonia for two studies,34,37 bacterial pneumonia for one study33 and a bacterial exacerbation of COPD for the final study. 36

Individual symptoms and signs

Overall, the estimated accuracy of symptoms and signs in the diagnosis of bacterial pneumonia was poor (Figure 3). Data were available from a single review by Gentilotti and colleagues, for symptoms closely associated with acute respiratory infection (such as cough, sore throat and a runny nose), generic symptoms (including myalgia and diarrhoea) and a variety of clinical signs (including tachycardia, hypotension and low oxygen saturation). 33 The certainty of the evidence ranged from very low to moderate certainty (see Appendix 3, Table 3). Concerns were predominantly due to the risk of bias in the primary studies, and wide confidence intervals that crossed our prespecified thresholds for ‘accurate’ or ‘very accurate’ tests.

FIGURE 3.

Sensitivity and specificity of individual symptoms and signs to diagnose bacterial pneumonia. Pooled estimates from random effects meta-analyses. 33

Subgroup analysis: People with chronic obstructive pulmonary disease

We identified one review which provided some data on the presence of purulent sputum to identify those with bacterial exacerbations of COPD. 36 In keeping with data for the general population, the estimated sensitivity and specificity were poor [71%, 95% CI 42 to 90.3 studies, 259 participants (very-low-certainty evidence due to risk of bias and wide confidence interval) and 51%, 95% CI 30 to 73.3 studies, 259 participants (moderate-certainty evidence due to a risk of bias), respectively, see Appendix 3, Table 3]. We did not identify any additional information on the subgroups of interest in this review.

Combinations of symptoms and signs

Schierenberg’s37 review included an analysis of clinical prediction models used to detect bacterial pneumonia – including combinations of symptoms and signs. Across the six models considered, the area under the curve (AUC) ranged from 0.53 to 0.79. This was considered very-low-certainty evidence due to heterogeneity between the individual estimates which ranged from not useful to useful, and a risk of publication bias (see Appendix 3, Table 3).

The same authors34 performed a separate review which investigated the addition of CRP to the models. When using a combination of CRP, symptoms and signs, the AUC was found to increase by 0.075 (95% CI 0.044 to 0.107). They also reported the accuracy of combinations of symptoms and signs plus CRP for diagnosing bacterial pneumonia at two risk thresholds: 2.5% and 20% (i.e. where individuals with a predicted risk of either ≥ 2.5% or ≥ 20% were classed as having bacterial pneumonia). Across the eight studies included in the review, the lower threshold was estimated to have high sensitivity (97%, 95% CI 95 to 98, moderate-certainty evidence) but the estimated specificity was poor (36%, 95% CI 34 to 37, moderate-certainty evidence). Raising the risk threshold to 20% resulted in much higher estimated specificity (90%, 95% CI 89 to 91, moderate-certainty evidence), but the estimated sensitivity dropped to 70% (95% CI 66 to 73, low-certainty evidence) (see Appendix 3, Table 3).

Biomarker point-of-care tests to detect bacterial or viral respiratory tract infection

Two recent systematic reviews were identified which investigated the use of host biomarkers. 32,33

C-reactive protein

Data were identified from one review. 33 Sensitivity and specificity of CRP to detect bacterial infection varied across the different thresholds assessed. These ranged from 10 mg/l [4 studies, 944 participants; estimated sensitivity 92% (95% CI 56 to 99) very-low-certainty evidence; estimated specificity 43% (95% CI 22 to 66) moderate-certainty evidence] to 100 mg/l [6 studies, 4418 participants; estimated sensitivity 52% (95% CI 31 to 72) moderate-certainty evidence; estimated specificity 91% (95% CI 79 to 97) low-certainty evidence] (see Appendix 3, Table 4).

Procalcitonin

Accuracy of procalcitonin to detect bacterial infection was assessed at three thresholds, by a single review. 33 Test accuracy varied across these thresholds. At > 0.1 mcg/ml the sensitivity and specificity were 74% (95% CI 38% to 93% and 36% to 94%) respectively; 4 studies; 1092 participants; very-low-certainty evidence). At > 0.25 mcg/ml the sensitivity was 44% (95% CI 14% to 79%) and specificity was 89% (95% CI 50% to 98%; 5 studies; 4019 participants; low- and very-low-certainty evidence). At > 0.5 mcg/ml the sensitivity was 44% (95% CI 19% to 73%) and specificity was 93% (95% CI 43% to 100%; 4 studies; 1195 participants; low- and very low-certainty evidence).

TNF-related apoptosis-induced ligand, IP-10 and CRP (ImmunoXpert)

The diagnostic accuracy of ImmunoXpert for bacterial infections had an estimated sensitivity of 85% (95% CI 75% to 91%) and estimated specificity of 86% (95% CI 73% to 93%; 4 studies; 1291 participants). 32 However, the evidence was again considered very low certainty (see Appendix 3, Table 4).

C-reactive protein and MxA (FebriDx)

FebriDx had an estimated sensitivity of 84% (95% CI 75 to 90; low-certainty evidence) and estimated specificity of 93% (95% CI 90 to 95; moderate-certainty evidence; 4 studies; 598 participants)32 (see Appendix 3, Table 4).

Other host biomarkers

The Carlton review32 identified one study (198 participants) which examined the combination of CRP and neopterin to diagnose bacterial infection. This was shown to have an estimated sensitivity of 80% (95% CI 71 to 86) and estimated specificity of 82% (95% CI 71 to 89), though there was very low certainty in the evidence for both estimates (see Appendix 3, Table 4).

Single pathogen tests for influenza and RSV

Influenza

The Gentilotti33 review assessed the accuracy of various single pathogen tests for influenza. However, of the six single pathogen tests included, only immunochromatography had an estimate for an adult-specific population. Estimates for the remaining tests are taken from studies that included both adults and children, and we therefore had serious concerns about indirectness in the GRADE assessment. This should be taken into consideration when interpreting the findings.

Most of the tests that rely on direct antigen detection (immunochromatography, direct immunofluorescence, optical immunoassays and MariPOC) showed adequate sensitivity (ranging from 56% to 82%) and high specificity (range 88–99%), although the certainty of the evidence was considered to be very low or low. This was due to concerns over the potential for bias, indirectness for some estimates (where analyses included children) and wide confidence intervals. Tests using a chemiluminescent neuraminidase assay showed adequate sensitivity (81%, 95% CI 51% to 94%, 787 participants, 4 studies) and specificity (82%, 95% CI 65% to 91%, 787 participants, 4 studies), but the certainty of the evidence was also very low due to the risk of bias and the inclusion of children in the analysis (see Appendix 3, Table 5).

Overall, the diagnostic accuracy of tests based on nucleic acid amplification (both PCR-based and non-PCR-based) appeared higher than those based on antigen detection, with sensitivity ranging from 91% to 95.1% and specificity from 97.5% to 98%. However, the certainty of the evidence was again considered low or very low due to the risk of bias, inclusion of children in the analyses and confidence intervals that crossed our pre-specified threshold for a useful test (90%) (see Appendix 3, Table 5).

Respiratory syncytial virus

A recent systematic review was identified which investigated the use of single pathogen tests for diagnosing RSV. 35 The vast majority of included studies were focused on children. However, we did identify two studies38,39 within this review that considered an adult population, and assessed tests that could be used in a point-of-care setting (direct immunofluorescence and rapid antigen testing). Both of these tests showed high specificity, but poor sensitivity for the detection of RSV, and the certainty of the evidence was low and very low (see Appendix 3, Table 5).

Discussion

The evidence identified in this review shows limited diagnostic accuracy for symptoms and signs of bacterial infection and for point-of-care tests that rely on a single biomarker (such as CRP or procalcitonin). Point-of-care tests that include multiple biomarkers may have slightly higher diagnostic accuracy. However, the evidence was predominantly assessed as low or very low certainty, due to limitations which include the risk of bias in primary studies, indirectness of the evidence and imprecision of the effect estimates.

We identified several tests used to diagnose influenza in an adult population, including tests that detect the presence of influenza antigens and those that detect nucleic acids. Diagnostic accuracy appeared highest for nucleic acid amplification tests – either those that test exclusively for influenza or multiplex tests (capable of diagnosing additional pathogens). The evidence was again considered to be predominantly low or very low certainty. The available data on RSV was very limited – the majority of primary studies were conducted in children, and therefore not applicable to this review. Consequently, we are unable to draw conclusions about the accuracy of direct antigen tests for RSV. The specificity of multiplex tests for RSV is probably high. However, the sensitivity may be lower, and the evidence was low certainty.

We used rigorous methods and extensive searches to ensure that all relevant evidence was identified for this review. Nonetheless, the majority of the evidence identified was considered to be low or very low certainty when assessed with the GRADE framework. In part, this was due to concerns over the potential for bias in the primary studies, and some concerns over indirectness in the included populations (where analyses included children, or some participants who were hospitalised). However, many of the concerns were due to imprecision in the effect estimates – as the confidence intervals crossed thresholds that we considered to represent an accurate or very accurate test (taken to be a sensitivity or specificity of 75% and 90%, respectively). We acknowledge that these thresholds are arbitrary and that readers, or different authors, may consider different thresholds to represent a useful test. This would impact on the certainty in the estimates. Furthermore, we noted that most systematic reviews did not report any information on heterogeneity in the primary studies included in their analyses. Consequently, we were unable to assess inconsistency when applying GRADE, and our assessment of evidence certainty may be considered optimistic.

In this review we primarily sought evidence about the accuracy of tests to distinguish between viral and bacterial causes of ARI that take no more than 45 minutes to yield a result. Symptoms and signs of infection are part of routine clinical assessment and therefore would add no time to the decision-making process. Many of the diagnostic tests identified give results within 10 to 15 minutes, making them suitable for use in a primary care setting or emergency department. However, multiplex tests typically require more time, with many taking up to 1 hour or longer. The extent to which such tests would fit into routine clinical practice needs careful consideration. However, the clear benefit of the multiplex platforms is the possibility of testing a single sample for multiple viral and bacterial pathogens, as well as the apparent increase in diagnostic accuracy.

It should be noted that most of the evidence identified in this review looked at the diagnostic accuracy of tests in isolation – that is the accuracy of a single test to determine the cause of a respiratory infection. In reality, clinical diagnosis involves assessment of a constellation of symptoms and signs, as well as the results of specific tests. We identified only one review that assessed the incremental benefit of assessing CRP in conjunction with symptoms and signs of infection. 34 The addition of CRP showed a small increase in diagnostic accuracy as compared to symptoms and signs alone. Due to a lack of published evidence, it is currently unclear whether this is also the case for tests that examine other biomarkers.

Symptoms and signs of respiratory infection are often used to determine eligibility for studies of test accuracy. For example, many studies will enrol individuals with a fever or cough for further testing. Consequently, it is possible that estimates of accuracy for individual symptoms and signs could be artificially high – as the prevalence of these symptoms is high in the study population. Nonetheless, this situation does reflect routine clinical practice, where healthcare professionals are using these features to determine who requires further testing.

We accepted any reference standard to diagnose bacterial infection. This was partly because there is no agreement on what constitutes an ideal reference test. Microbiological testing may be regarded as an essential component of determining a viral or bacterial cause of an infection. However, these tests are likely to detect the presence of commensal organisms and are known to produce false-negative results (due to inadequacy of sampling technique or culture methods). 56,57 Consequently, a variety of reference standards were used in the studies included in this review – ranging from radiological imaging, microbiological assessment (such as culture and/or PCR) and consensus opinion of an expert panel.

For some tests, there were similarities between the index test and the reference standard used. In particular, a number of pathogen-specific tests used PCR techniques as both the index test and the reference standard. Given the similar methods used, the results of these tests are likely to be correlated, and therefore the accuracy of the index tests may be overestimated. 58

Cost-effectiveness was not assessed as part of this review. However, the cost of different types of tests varies and some may be prohibitively expensive for use in a primary care setting. This will need careful consideration before implementing a new testing strategy.

This review focused only on the diagnostic accuracy of tests. For patients and clinicians, the most important questions are likely to be about the impact of using these tests on health outcomes. For example, does testing for bacterial infections result in better health than relying on clinical judgement alone? Will more people avoid side effects from the prescription of unnecessary antibiotics? Will hospital admissions be reduced, or people suffer fewer complications from severe bacterial infections? Assessing these outcomes requires studies that consider the implementation of these tests followed by clinical management based on the test results.

It is recognised that prescription and use of antibiotics may be affected by factors commonly associated with health inequalities, such as age and ethnicity. 59 Testing to help determine who needs antibiotics could help to reduce these inequities in health care, but only if the tests themselves are used appropriately. CRP is one of the only tests for which there is evidence in relation to equity of use. Despite their limited diagnostic accuracy, the use of CRP tests for people with ARI may reduce antibiotic prescribing without increasing negative health outcomes,60 in part because they may enable clinicians to communicate a ‘no antibiotic’ treatment decision more easily. 59,61 A study from Denmark (where CRP tests are widely used) found that clinicians were less likely to use a CRP test when prescribing antibiotics for those who were unemployed or receiving disability pension, immigrants or children of immigrants. 62 It is not clear why this happens, but consequently these groups may still be more likely to be prescribed unnecessary antibiotics. For these tests to help address (rather than reproduce) inequities, there needs to be clear guidance and monitoring of use with respect to underserved groups.

At present there is an absence of evidence regarding diagnostic accuracy to support current clinical practice (where symptoms and signs are used to diagnose bacterial infection) or to justify the introduction of microbiological or host-response point-of-care tests. Policy makers should resist seeing point-of-care tests as the ‘silver bullet’ to solve healthcare system pressures until there is adequate evidence to demonstrate they are safe, clinically effective and cost-effective. There are concerns that introduction of point-of-care tests may unintentionally increase healthcare demand, as patients’ illnesses become ‘medicalised’ with attendances for testing.

We recommend further research to define an adequate reference standard for respiratory infection diagnosis. This could be based on a better understanding of the natural history of the microbiology, and/or prognosis, of infections. In addition, it should be established whether point-of-care tests add diagnostic value over and above current practice – the use of symptoms and signs to identify individuals at risk of more severe illness, or who require additional treatment. There is a lack of high-quality evidence regarding the diagnostic accuracy of point-of-care tests for ARI in the community and emergency department setting. The diagnostic accuracy of such point-of-care tests should be assessed specifically in this setting where the population is different – with generally less severe infections and consequently different microbiology and immune responses. This means that data from inpatients cannot necessarily be extrapolated to the outpatient setting, as the diagnostic accuracy of the test may vary according to disease severity. Finally, it will be important to assess if the use of point-of-care tests will medicalise illness and lead to unintended increased demand for NHS care for ARIs.

Conclusion

The majority of the evidence identified in this review was considered to be low or very low certainty, highlighting that future studies may change the overall estimates of accuracy. Nonetheless, from the evidence identified in this review it appears that individual symptoms and signs, or existing clinical prediction models (incorporating multiple symptoms and signs) are unlikely to be sufficiently accurate to distinguish between bacterial and viral infections. Diagnostic accuracy of individual host biomarkers also appears to be insufficient, although certain combinations of biomarkers may have higher sensitivity and specificity. As may be expected, the accuracy of different types of rapid tests for influenza and RSV varied. The highest diagnostic accuracy was seen with tests that rely on amplification of viral nucleic acid (including PCR and non-PCR-based techniques).

Further work is required to determine the optimum reference standard, and whether the introduction of point-of-care tests may add value to current diagnostic pathways. It remains to be seen whether additional testing would improve health outcomes for patients, or simply lead to an increase in healthcare consultations and resource costs.

Additional information

List of abbreviations

AECOPD

acute exacerbation of chronic obstructive pulmonary disease

ARI

acute respiratory infection

AUC

area under the curve

BP

blood pressure

CDSR

Cochrane Database of Systematic Reviews

CI

confidence interval

COPD

chronic obstructive pulmonary disease

CRP

C-reactive protein

CT

computed tomography

DFA

direct fluorescence antigen

DTA

diagnostic test accuracy

EU

European Union

GRADE

Grading of Recommendations Assessment, Development and Evaluation

HSROC

hierarchical summary receiver operating characteristic

ICU

intensive care unit

IP-10

interferon-γ-induced protein-10

IQR

interquartile range

LRTI

lower respiratory tract infection

mPCR

multiplex polymerase chain reaction

MRI

magnetic resonance imaging

MxA

myxovirus resistance protein A

NAAT

nucleic acid amplification test

NICE

National Institute for Health and Care Excellence

PCR

polymerase chain reaction

POCT

point-of-care test

PRISMA

Preferred Reporting Items for Systematic Reviews and Meta-Analyses

QUADAS-2

Quality Assessment of Diagnostic Accuracy Studies version 2

RADT

rapid antigen detection tests

RFT

rapid flu test

RIDT

rapid influenza diagnostic test

ROBIS

Risk of Bias in Systematic Reviews

ROC

receiver operating characteristic

RSV

respiratory syncytial virus

RTI

respiratory tract infection

RT PCR

real-time polymerase chain reaction

SD

standard deviation

SpO2

oxygen saturations

TRAIL

TNF-related apoptosis-induced ligand

URTI

upper respiratory tract infection

WHO

World Health Organization

Notes

1. Supplementary material

Supplementary material can be found on the NIHR Journals Library report page (https://doi.org/10.3310/JLCP4570).

Supplementary material has been provided by the authors to support the report and any files provided at submission will have been seen by peer reviewers, but not extensively reviewed. Any supplementary material provided at a later stage in the process may not have been peer reviewed.

References

1. UK Health Security Agency . Surveillance of Influenza and Other Seasonal Respiratory Viruses in the UK, Winter 2022 to 2023 2023. Surveillance of influenza and other seasonal respiratory viruses in winter 2021 to 2022 - GOV.UK (www.gov.uk) (accessed 27 July 2023).
2. NHS England . Combined Adult and Paediatric Acute Respiratory Infection (ARI) Hubs (Previously RCAS Hubs) 2022. BW2064-combined-adult-paediatric-ari-hubs-october-22.pdf (england.nhs.uk) (accessed 27 July 2023).
3. NHS England . Guidance Note: Acute Respiratory Infection Virtual Ward 2022. NHS England » Guidance note: Acute respiratory infection virtual ward (accessed 27 July 2023).
4. National Institute for Health and Care Excellence . Suspected Acute Respiratory Infection in Over 16s: Assessment at First Presentation and Initial Management. NICE 2023. Suspected acute respiratory infection in over 16s: assessment at first presentation and initial management (nice.org.uk) (accessed 12 December 2023).
5. Jin X, Ren J, Li R, Gao Y, Zhang H, Li J, et al. Global burden of upper respiratory infections in 204 countries and territories, from 1990 to 2019. EClinicalMedicine 2021;37. https://doi.org/10.1016/j.eclinm.2021.100986.
6. Safiri S, Mahmoodpoor A, Kolahi AA, Nejadghaderi SA, Sullman MJM, Mansournia MA, et al. Global burden of lower respiratory infections during the last three decades. Front Public Health 2022;10. https://doi.org/10.3389/fpubh.2022.1028525.
7. Meier GC, Watkins J, McEwan P, Pockett RD. Resource use and direct medical costs of acute respiratory illness in the UK based on linked primary and secondary care records from 2001 to 2009. PLOS ONE 2020;15. https://doi.org/10.1371/journal.pone.0236472.
8. Creer DD, Dilworth JP, Gillespie SH, Johnston AR, Johnston SL, Ling C, et al. Aetiological role of viral and bacterial infections in acute adult lower respiratory tract infection (LRTI) in primary care. Thorax 2006;61:75-9. https://doi.org/10.1136/thx.2004.027441.
9. Musher DM, Abers MS, Bartlett JG. Evolving understanding of the causes of pneumonia in adults, with special attention to the role of pneumococcus. Clin Infect Dis 2017;65:1736-44. https://doi.org/10.1093/cid/cix549.
10. Fletcher-Lartey S, Yee M, Gaarslev C, Khan R. Why do general practitioners prescribe antibiotics for upper respiratory tract infections to meet patient expectations: a mixed methods study. BMJ Open 2016;6. https://doi.org/10.1136/bmjopen-2016-012244.
11. O’Connor R, O’Doherty J, O’Regan A, Dunne C. Antibiotic use for acute respiratory tract infections (ARTI) in primary care; what factors affect prescribing and why is it important? A narrative review. Ir J Med Sci 2018;187:969-86. https://doi.org/10.1007/s11845-018-1774-5.
12. Ayorinde A, Ghosh I, Ali I, Zahair I, Olarewaju O, Singh M, et al. Health inequalities in infectious diseases: a systematic overview of reviews. BMJ Open 2023;13. https://doi.org/10.1136/bmjopen-2022-067429.
13. Millett ER, Quint JK, Smeeth L, Daniel RM, Thomas SL. Incidence of community-acquired lower respiratory tract infections and pneumonia among older adults in the United Kingdom: a population-based study. PLOS ONE 2013;8. https://doi.org/10.1371/journal.pone.0075131.
14. Fisk WJ, Eliseeva EA, Mendell MJ. Association of residential dampness and mold with respiratory tract infections and bronchitis: a meta-analysis. Environ Health 2010;9. https://doi.org/10.1186/1476-069X-9-72.
15. MacIntyre EA, Gehring U, Molter A, Fuertes E, Klumper C, Kramer U, et al. Air pollution and respiratory infections during early childhood: an analysis of 10 European birth cohorts within the ESCAPE Project. Environ Health Perspect 2014;122:107-13. https://doi.org/10.1289/ehp.1306755.
16. Almirall J, Serra-Prat M, Bolíbar I, Balasso V. Risk factors for community-acquired pneumonia in adults: a systematic review of observational studies. Respiration 2017;94:299-311. https://doi.org/10.1159/000479089.
17. Nomamiukor BO, Horner C, Kirby A, Hughes GJ. Living conditions are associated with increased antibiotic resistance in community isolates of Escherichia coli. J Antimicrob Chemother 2015;70:3154-8. https://doi.org/10.1093/jac/dkv229.
18. Covvey JR, Johnson BF, Elliott V, Malcolm W, Mullen AB. An association between socioeconomic deprivation and primary care antibiotic prescribing in Scotland. J Antimicrob Chemother 2014;69:835-41. https://doi.org/10.1093/jac/dkt439.
19. Thomson K, Berry R, Robinson T, Brown H, Bambra C, Todd A. An examination of trends in antibiotic prescribing in primary care and the association with area-level deprivation in England. BMC Public Health 2020;20. https://doi.org/10.1186/s12889-020-09227-x.
20. Rose J, Crosbie M, Stewart A. A qualitative literature review exploring the drivers influencing antibiotic over-prescribing by GPs in primary care and recommendations to reduce unnecessary prescribing. Perspect Public Health 2021;141:19-27. https://doi.org/10.1177/1757913919879183.
21. Cabral C, Lucas PJ, Ingram J, Hay AD, Horwood J. It’s safer to …’ parent consulting and clinician antibiotic prescribing decisions for children with respiratory tract infections: an analysis across four qualitative studies. Soc Sci Med 2015;136-137:156-64. https://doi.org/10.1016/j.socscimed.2015.05.027.
22. Wang KY, Seed P, Schofield P, Ibrahim S, Ashworth M. Which practices are high antibiotic prescribers? A cross-sectional analysis. Br J Gen Pract 2009;59:e315-20. https://doi.org/10.3399/bjgp09X472593.
23. Whiting P, Rutjes AW, Westwood ME, Mallett S, Deeks JJ, Reitsma JB, et al. QUADAS-2: a revised tool for the quality assessment of diagnostic accuracy studies. Ann Intern Med 2011;155:529-36. https://doi.org/10.7326/0003-4819-155-8-201110180-00009.
24. Whiting PF, Rutjes AW, Westwood ME, Mallett S, Group Q-S. A systematic review classifies sources of bias and variation in diagnostic test accuracy studies. J Clin Epidemiol 2013;66:1093-104. https://doi.org/10.1016/j.jclinepi.2013.05.014.
25. Lijmer JG, Mol BW, Heisterkamp S, Bonsel GJ, Prins MH, van der Meulen JHP, et al. Empirical evidence of design-related bias in studies of diagnostic tests. JAMA 1999;282:1061-6. https://doi.org/10.1001/jama.282.11.1061.
26. Whiting P, Savovic J, Higgins JPT, Caldwell DM, Reeves BC, Shea B, et al. ROBIS Group . ROBIS: a new tool to assess risk of bias in systematic reviews was developed. J Clin Epidemiol 2016;69:225-34. https://doi.org/10.1016/j.jclinepi.2015.06.005.
27. Pollock M, Fernandes RM, Becker LA, Pieper D, Hartling L, . Cochrane Handbook for Systematic Reviews of Interventions. Version 6.3 (updated February 2022). Cochrane; 2022.
28. Chu H, Cole SR. Bivariate meta-analysis of sensitivity and specificity with sparse data: a generalized linear mixed model approach. J Clin Epidemiol 2006;59:1331-2. https://doi.org/10.1016/j.jclinepi.2006.06.011.
29. Reitsma JB, Glas AS, Rutjes AW, Scholten RJ, Bossuyt PM, Zwinderman AH. Bivariate analysis of sensitivity and specificity produces informative summary measures in diagnostic reviews. J Clin Epidemiol 2005;58:982-90. https://doi.org/10.1016/j.jclinepi.2005.02.022.
30. Harbord RM, Whiting P. Metandi: meta-analysis of diagnostic accuracy using hierarchical logistic regression. Stata J 2009;9:211-29. https://doi.org/10.1177/1536867X0900900203.
31. Guyatt GH, Oxman AD, Vist GE, Kunz R, Falck-Ytter Y, Alonso-Coello P, et al. GRADE Working Group . GRADE: an emerging consensus on rating quality of evidence and strength of recommendations. BMJ 2008;336:924-6. https://doi.org/10.1136/bmj.39489.470347.AD.
32. Carlton HC, Savovic J, Dawson S, Mitchelmore PJ, Elwenspoek MMC. Novel point-of-care biomarker combination tests to differentiate acute bacterial from viral respiratory tract infections to guide antibiotic prescribing: a systematic review. Clin Microbiol Infect 2021;27:1096-108. https://doi.org/10.1016/j.cmi.2021.05.018.
33. Gentilotti E, De Nardo P, Cremonini E, Gorska A, Mazzaferri F, Canziani LM, et al. Diagnostic accuracy of point-of-care tests in acute community-acquired lower respiratory tract infections. A systematic review and meta-analysis. Clin Microbiol Infect 2022;28:13-22. https://doi.org/10.1016/j.cmi.2021.09.025.
34. Minnaard MC, de Groot JAH, Hopstaken RM, Schierenberg A, de Wit NJ, Reitsma JB, et al. The added value of C-reactive protein measurement in diagnosing pneumonia in primary care: a meta-analysis of individual patient data. CMAJ 2017;189:E56-63. https://doi.org/10.1503/cmaj.151163.
35. Onwuchekwa C, Moreo LM, Menon S, Machado B, Curcio D, Kalina W, et al. Under-ascertainment of Respiratory Syncytial Virus infection in adults due to diagnostic testing limitations: a systematic literature review and meta-analysis. J Infect Dis 2023;228:173-84. https://doi.org/10.1093/infdis/jiad012.
36. Pazmany P, Soos A, Hegyi P, Dohos D, Kiss S, Szakacs Z, et al. Inflammatory biomarkers are inaccurate indicators of bacterial infection on admission in patients with acute exacerbation of chronic obstructive pulmonary disease-A systematic review and diagnostic accuracy network meta-analysis. Front Med (Lausanne) 2021;8. https://doi.org/10.3389/fmed.2021.639794.
37. Schierenberg A, Minnaard MC, Hopstaken RM, van de Pol AC, Broekhuizen BD, de Wit NJ, et al. External validation of prediction models for pneumonia in primary care patients with lower respiratory tract infection: an individual patient data meta-analysis. PLOS ONE 2016;11. https://doi.org/10.1371/journal.pone.0149895.
38. Franck KT, Schneider UV, Ma CMG, Knudsen D, Lisby G. Evaluation of immuview RSV antigen test (SSI siagnostica) and BinaxNOW RSV card (alere) for rapid detection of respiratory syncytial virus in retrospectively and prospectively collected respiratory samples. J Med Virol 2020;92:2992-8. https://doi.org/10.1002/jmv.26369.
39. Simpson JL, Moric I, Wark PA, Johnston SL, Gibson PG. Use of induced sputum for the diagnosis of influenza and infections in asthma: a comparison of diagnostic techniques. J Clin Virol 2003;26:339-46. https://doi.org/10.1016/s1386-6532(02)00084-7.
40. Castro-Guardiola A, Armengou-Arxé A, Viejo-Rodríguez A, Peñarroja-Matutano G, Garcia-Bragado F. Differential diagnosis between community-acquired pneumonia and non-pneumonia diseases of the chest in the emergency ward. Eur J Intern Med 2000;11:334-9. https://doi.org/10.1016/S0953-6205(00)00118-7.
41. Holm A, Nexoe J, Bistrup LA, Pedersen SS, Obel N, Nielsen LP, et al. Aetiology and prediction of pneumonia in lower respiratory tract infection in primary care. Br J Gen Pract 2007;57:547-54.
42. Liu YF, Gao Y, Chen MF, Cao B, Yang XH, Wei L. Etiological analysis and predictive diagnostic model building of community-acquired pneumonia in adult outpatients in Beijing, China. BMC Infect Dis 2013;13:1-10. https://doi.org/10.1186/1471-2334-13-309.
43. Gulich MS, Matschiner A, Glück R, Zeitler HP. Improving diagnostic accuracy of bacterial pharyngitis by near patient measurement of C-reactive protein (CRP). Br J Gen Pract 1999;49:119-21.
44. Escarate E, Jones CG, Clarke E, Clark P, Norton S, Bag S, et al. Rapid on-site molecular Point of Care Testing during influenza outbreaks in aged care facilities improves antiviral use and reduces hospitalisation. Aust N Z J Public Health 2022;46:884-8. https://doi.org/10.1111/1753-6405.13307.
45. Farfour E, Yung T, Baudoin R, Vasse M. Evaluation of four fully integrated molecular assays for the detection of respiratory viruses during the co-circulation of SARS-CoV-2, influenza and RSV. J Clin Med 2022;11. https://doi.org/10.3390/jcm11143942.
46. Maignan M, Viglino D, Hablot M, Termoz Masson N, Lebeugle A, Collomb Muret R, et al. Diagnostic accuracy of a rapid RT-PCR assay for point-of-care detection of influenza A/B virus at emergency department admission: a prospective evaluation during the 2017/2018 influenza season. PLOS ONE 2019;14. https://doi.org/10.1371/journal.pone.0216308.
47. Morris TC, Bird PW, Horvath-Papp E, Dhillon JK, May S, Tang JW. Xpert Xpress Flu/RSV: validation and impact evaluation at a large UK hospital trust. J Med Virol 2021;93:5146-51. https://doi.org/10.1002/jmv.26860.
48. Valentin T, Kieslinger P, Stelzl E, Santner BI, Groselj-Strele A, Kessler HH, et al. Prospective evaluation of three rapid molecular tests for seasonal influenza in patients presenting at an emergency unit. J Clin Virol 2019;111:29-32. https://doi.org/10.1016/j.jcv.2019.01.003.
49. Yin N, Van Nuffelen M, Bartiaux M, Preseau T, Roggen I, Delaunoy S, et al. Clinical impact of the rapid molecular detection of RSV and influenza A and B viruses in the emergency department. PLOS ONE 2022;17. https://doi.org/10.1371/journal.pone.0274222.
50. Youngs J, Iqbal Y, Glass S, Riley P, Pope C, Planche T, et al. Implementation of the cobas Liat influenza point-of-care test into an emergency department during a high-incidence season: a retrospective evaluation following real-world implementation. J Hosp Infect 2019;101:285-8. https://doi.org/10.1016/j.jhin.2018.12.008.
51. Boku S, Naito T, Murai K, Tanei M, Inui A, Nisimura H, et al. Near point-of-care administration by the attending physician of the rapid influenza antigen detection immunochromatography test and the fully automated respiratory virus nucleic acid test: contribution to patient management. Diagn Microbiol Infect Dis 2013;76:445-9. https://doi.org/10.1016/j.diagmicrobio.2013.04.029.
52. Hansen GT, Moore J, Herding E, Gooch T, Hirigoyen D, Hanson K, et al. Clinical decision making in the emergency department setting using rapid PCR: results of the CLADE study group. J Clin Virol 2018;102:42-9. https://doi.org/10.1016/j.jcv.2018.02.013.
53. Peretz A, Zadok BS, Azrad M. Performance of the Influ a + B K-SeT(R) assay as compared to two RT-PCR assays for detection of influenza virus. Diagn Microbiol Infect Dis 2020;98. https://doi.org/10.1016/j.diagmicrobio.2020.115097.
54. Tanei M, Yokokawa H, Murai K, Sakamoto R, Amari Y, Boku S, et al. Factors influencing the diagnostic accuracy of the rapid influenza antigen detection test (RIADT): a cross-sectional study. BMJ Open 2014;4. https://doi.org/10.1136/bmjopen-2013-003885.
55. Zuurbier RP, Korsten K, Verheij TJM, Butler C, Adriaenssens N, Coenen S, et al. REspiratory Syncytial Virus Consortium in EUrope (RESCEU) Investigators . Performance assessment of a rapid molecular respiratory syncytial virus point-of-care test: a prospective community study in older adults. J Infect Dis 2022;226:S63-70. https://doi.org/10.1093/infdis/jiab600.
56. Theerthakarai R, El-Halees W, Ismail M, Solis RA, Khan MA. Nonvalue of the initial microbiological studies in the management of nonsevere community-acquired pneumonia. Chest 2001;119:181-4. https://doi.org/10.1378/chest.119.1.181.
57. Ewig S, Schlochtermeier M, Goïke N, Niederman MS. Applying sputum as a diagnostic tool in pneumonia: limited yield, minimal impact on treatment decisions. Chest 2002;121:1486-92. https://doi.org/10.1378/chest.121.5.1486.
58. Trikalinos TA, Balion CM. Chapter 9: options for summarizing medical test performance in the absence of a ‘gold standard’. J Gen Intern Med 2012;27:S67-75. https://doi.org/10.1007/s11606-012-2031-7.
59. Harvey EJ, De Brun C, Casale E, Finistrella V, Ashiru-Oredope D. Influence of factors commonly known to be associated with health inequalities on antibiotic use in high-income countries: a systematic scoping review. J Antimicrob Chemother 2023;78:861-70. https://doi.org/10.1093/jac/dkad034.
60. Smedemark SA, Aabenhus R, Llor C, Fournaise A, Olsen O, Jørgensen KJ. Biomarkers as point‐of‐care tests to guide prescription of antibiotics in people with acute respiratory infections in primary care. Cochrane Database Syst Rev 2022;2022. https://doi.org/10.1002/14651858.CD010130.pub3.
61. Tonkin-Crine S, Anthierens S, Hood K, Yardley L, Cals JW, Francis NA, et al. GRACE INTRO/CHAMP consortium . Discrepancies between qualitative and quantitative evaluation of randomised controlled trial results: achieving clarity through mixed methods triangulation. Implement Sci 2016;11. https://doi.org/10.1186/s13012-016-0436-0.
62. Sydenham RV, Hansen MP, Justesen US, Pedersen LB, Aabenhus RM, Wehberg S, et al. Factors associated with C-reactive protein testing when prescribing antibiotics in general practice: a register-based study. BMC Prim Care 2022;23. https://doi.org/10.1186/s12875-021-01614-6.
63. Bruning AHL, Leeflang MMG, Vos J, Spijker R, de Jong MD, Wolthers KC, et al. Rapid tests for influenza, respiratory syncytial virus, and other respiratory viruses: a systematic review and meta-analysis. Clin Infect Dis 2017;65:1026-32. https://doi.org/10.1093/cid/cix461.
64. Chartrand C, Leeflang MM, Minion J, Brewer T, Pai M. Accuracy of rapid influenza diagnostic tests: a meta-analysis. Ann Intern Med 2012;156:500-11. https://doi.org/10.7326/0003-4819-156-7-201204030-00403.
65. Chartrand C, Tremblay N, Renaud C, Papenburg J. Diagnostic accuracy of rapid antigen detection tests for respiratory syncytial virus infection: systematic review and meta-analysis. J Clin Microbiol 2015;53:3738-49. https://doi.org/10.1128/JCM.01816-15.
66. Engel MF, Paling FP, Hoepelman AI, van der Meer V, Oosterheert JJ. Evaluating the evidence for the implementation of C-reactive protein measurement in adult patients with suspected lower respiratory tract infection in primary care: a systematic review. Fam Pract 2012;29:383-93. https://doi.org/10.1093/fampra/cmr119.
67. Falk G, Fahey T. C-reactive protein and community-acquired pneumonia in ambulatory care: systematic review of diagnostic accuracy studies. Fam Pract 2009;26:10-21. https://doi.org/10.1093/fampra/cmn095.
68. Hill AT, Gold PM, El Solh AA, Metlay JP, Ireland B, Irwin RS. CHEST Expert Cough Panel . Adult outpatients with acute cough due to suspected pneumonia or influenza: CHEST guideline and expert panel report. Chest 2019;155:155-67. https://doi.org/10.1016/j.chest.2018.09.016.
69. Han MY, Xie TA, Li JX, Chen HJ, Yang XH, Guo XG. Evaluation of lateral-flow assay for rapid detection of influenza virus. Biomed Res Int 2020;2020. https://doi.org/10.1155/2020/3969868.
70. Hoult G, Gillespie D, Wilkinson TMA, Thomas M, Francis NA. Biomarkers to guide the use of antibiotics for acute exacerbations of COPD (AECOPD): a systematic review and meta-analysis. BMC Pulm Med 2022;22. https://doi.org/10.1186/s12890-022-01958-4.
71. Htun TP, Sun Y, Chua HL, Pang J. Clinical features for diagnosis of pneumonia among adults in primary care setting: a systematic and meta-review. Sci Rep 2019;9. https://doi.org/10.1038/s41598-019-44145-y.
72. Huang HS, Tsai CL, Chang J, Hsu TC, Lin S, Lee CC. Multiplex PCR system for the rapid diagnosis of respiratory virus infection: systematic review and meta-analysis. Clin Microbiol Infect 2018;24:1055-63. https://doi.org/10.1016/j.cmi.2017.11.018.
73. Lee J, Song JU, Kim YH. Diagnostic accuracy of the Quidel Sofia rapid influenza fluorescent immunoassay in patients with influenza-like illness: a systematic review and meta-analysis. Tuberc Respir Dis (Seoul) 2021;84:226-36. https://doi.org/10.4046/trd.2021.0033.
74. Merckx J, Wali R, Schiller I, Caya C, Gore GC, Chartrand C, et al. Diagnostic accuracy of novel and traditional rapid tests for influenza infection compared with reverse transcriptase polymerase chain reaction: a systematic review and meta-analysis. Ann Intern Med 2017;167:394-409. https://doi.org/10.7326/M17-0848.
75. Nicholson KG, Abrams KR, Batham S, Medina MJ, Warren FC, Barer M, et al. Randomised controlled trial and health economic evaluation of the impact of diagnostic testing for influenza, respiratory syncytial virus and Streptococcus pneumoniae infection on the management of acute admissions in the elderly and high-risk 18- to 64-year-olds. Health Technol Assess 2014;18:1-274. https://doi.org/10.3310/hta18360.
76. Petrozzino JJ, Smith C, Atkinson MJ. Rapid diagnostic testing for seasonal influenza: an evidence-based review and comparison with unaided clinical diagnosis. J Emerg Med 2010;39:476-490.e1. https://doi.org/10.1016/j.jemermed.2009.11.031.
77. van der Meer V, Neven AK, van den Broek PJ, Assendelft WJ. Diagnostic value of C reactive protein in infections of the lower respiratory tract: systematic review. BMJ 2005;331. https://doi.org/10.1136/bmj.38483.478183.EB.
78. Vos LM, Bruning AHL, Reitsma JB, Schuurman R, Riezebos-Brilman A, Hoepelman AIM, et al. Rapid molecular tests for influenza, respiratory syncytial virus, and other respiratory viruses: a systematic review of diagnostic accuracy and clinical impact studies. Clin Infect Dis 2019;69:1243-53. https://doi.org/10.1093/cid/ciz056.
79. Wu MH, Lin CC, Huang SL, Shih HM, Wang CC, Lee CC, et al. Can procalcitonin tests aid in identifying bacterial infections associated with influenza pneumonia? A systematic review and meta-analysis. Influenza Other Respir Viruses 2013;7:349-55. https://doi.org/10.1111/j.1750-2659.2012.00386.x.

Appendix 1

Appendix 2

Appendix 3

Appendix 4

FIGURE 4.

PRISMA flow diagram for white cell differential count.

FIGURE 5.

QUADAS-2 assessments: white cell differential count. ^aIndex test was incorporated as part of the reference standard, and was also not conducted in a point-of-care setting (central laboratory analysis of white cell count). ^bIndex test was not conducted in a point-of-care setting (central laboratory analysis of white cell count). ^cPeople with more severe illness (requiring hospital admission) were excluded from participation. Index test was not conducted in a point-of-care setting (central laboratory analysis of white cell count). Participants with possible malignancy were excluded from analysis. ^dUnclear whether sampling was consecutive/random. Unclear how white cell count was assessed, but not conducted in a point-of-care setting.

Appendix 5

Appendix 6

FIGURE 6.

PRISMA flow diagram for multiplex PCR tests.

FIGURE 7.

QUADAS-2 assessments: multiplex PCR tests. ^aUnclear if a consecutive/random sample of participants was enrolled. Unclear if reference standard results were interpreted without knowledge of the index test. ^bUnclear if a consecutive/random sample of participants was enrolled. Unclear if all participants would have attended for assessment by a healthcare provider if they were not enrolled in the study (care home residents). Some participants did not receive a reference standard and were excluded from the analysis. ^cUnclear whether the index test was carried out at point of care. Some participants were excluded from analysis. ^dUnclear if a consecutive/random sample of participants was enrolled. Participants with multiple conditions were excluded from this study. ^eUnclear if a consecutive/random sample of participants was enrolled. Staff were asked to collect samples ‘as per usual practice’. No guidance on specific inclusion/exclusion criteria for sampling. ^fUnclear if a consecutive/random sample of participants was enrolled. Rapid antigen test was used as the reference standard. Unclear if the index test was interpreted without knowledge of the reference standard. Index test not actually conducted in a point-of-care setting. ^gUnclear if the index test was interpreted without knowledge of the reference standard. Rapid antigen test was used as the reference standard. Index test not actually conducted in a point-of-care setting. ^hNot all participants were included in the analysis. Index test not analysed at point of care. ⁱUnclear if a consecutive/random sample of participants was enrolled. Index tests were a component of the reference standard. Not all samples analysed in point-of-care setting. ^jReference standard results were interpreted with knowledge of the index test results. High number of exclusions. ^kParticipants were undergoing home assessment for respiratory illness as part of a larger study. Many would not have sought medical care had it not been that they were participating in this study. High number of exclusions from analysis.

Appendix 7

FIGURE 8.

Meta-analysis for influenza A.

FIGURE 9.

Influenza A data and meta-analysis results in ROC space.

Appendix 8

FIGURE 10.

Meta-analysis for influenza B.

FIGURE 11.

Influenza B data and meta-analysis results in ROC space.

Appendix 9

FIGURE 12.

Meta-analysis for influenza A or B.

FIGURE 13.

Influenza A/B data and meta-analysis results in ROC space.

Appendix 10

FIGURE 14.

Meta-analysis for RSV.

FIGURE 15.

RSV data and meta-analysis results in ROC space.