Abstract
Background Management of pharyngitis is commonly based on features which are thought to be associated with Lancefield group A beta-haemolytic streptococci (GABHS) but it is debatable which features best predict GABHS. Non-group A strains share major virulence factors with group A, but it is unclear how commonly they present and whether their presentation differs.
Aim To assess the incidence and clinical variables associated with streptococcal infections.
Design and setting Prospective diagnostic cohort study in UK primary care.
Method The presence of pathogenic streptococci from throat swabs was assessed among patients aged ≥5 years presenting with acute sore throat.
Results Pathogenic streptococci were found in 204/597 patients (34%, 95% CI = 31 to 38%): 33% (68/204) were non-group A streptococci, mostly C (n = 29), G (n = 18) and B (n = 17); rarely D (n = 3) and Streptococcus pneumoniae (n = 1). Patients presented with similar features whether the streptococci were group A or non-group A. The features best predicting A, C or G beta-haemolytic streptococci were patient’s assessment of severity (odds ratio [OR] for a bad sore throat 3.31, 95% CI = 1.24 to 8.83); doctors’ assessment of severity (severely inflamed tonsils OR 2.28, 95% CI = 1.39 to 3.74); absence of a bad cough (OR 2.73, 95% CI = 1.56 to 4.76), absence of a coryza (OR 1.54, 95% CI = 0.99 to 2.41); and moderately bad or worse muscle aches (OR 2.20, 95% CI = 1.41 to 3.42).
Conclusion Non-group A strains commonly cause streptococcal sore throats, and present with similar symptomatic clinical features to group A streptococci. The best features to predict streptococcal sore throat presenting in primary care deserve revisiting.
BACKGROUND
Pharyngitis is one of the commonest presentations in clinical practice and most patients are treated with antibiotics despite a Cochrane Review suggesting modest symptomatic benefit.1 A reasonable strategy in reducing the public health threat of antibiotic resistance is to limit antibiotic use to the minority of individuals with streptococcal infections who are more likely to benefit2–4 and avoid treatment in those unlikely to benefit. Lancefield group A beta-haemolytic streptococci (GABHS) is the most frequent major bacterial pathogen in pharyngitis and clinical scores to predict GABHS have some promise to be useful5–7 including the simple ‘Centor’ criteria — 3 out of 4 of pus, cervical nodes, a history of fever and no history of cough, which are widely advocated in clinical practice guidance.2,4,8–10 However these criteria have low specificity8 leading to high rates of overall antibiotic use.8 Furthermore small studies in typical primary care settings have suggested other features might be useful in refining the criteria; such as, shorter prior duration, severity of pain, and muscle ache.11,7 The issue of which variables most strongly predict streptococcal infections is therefore still not settled.
Rapid streptococcal antigen tests to detect GABHS are used widely in developed countries to target treatment since GABHS is the most common major pathogen,5 and the use of rapid tests may help practitioners reduce prescribing.12,13 Other Lancefield groups (particularly C and G) have hitherto had much less emphasis,14 and there are no rapid antigen tests available for detecting these groups. Although rheumatic fever is probably not caused by group C and G streptococci, the incidence of rheumatic fever has dramatically declined in developed countries1 and antibiotic treatment to prevent rheumatic fever is an extremely inefficient use of healthcare resources.4 The major virulence factors among group A streptococci are shared by group C and G streptococci, particularly the M proteins, peptidase, hylauronic capsule, and streptokinase14,11 and similar numbers of cases of streptococcal septicaemia due to C and G streptococci are regularly reported.15 However, such complications are rare, hence the major benefit of targeting streptococcal infections, which may apply to groups C, G, and A if clinical presentation is similar, is likely to be in limiting antibiotic treatment to individuals who will benefit from more rapid symptom resolution1 and a shorter infective period.2 A small study in two Norwegian practices suggested that Lancefield groups C and G presentations were similar to group A;11 supported by Tiemstra et al.3 Conversely a substantial study concluded that clinical presentations were different.16 Thus both the relevance of C and G streptococci in symptomatic presentation and the clinical predictors require clarification. This study reports new data on the epidemiology of pathogens causing pharyngitis and the predictors of the presence of pathogenic streptococci for patients presenting with pharyngitis in primary care.
How this fits in
This study suggests that it is not just Lancefield group A strains that cause significant illness in primary care: non-group A strains commonly cause streptococcal sore throats, and present with similar symptomatic clinical features to group A streptococci. This suggests that RADTs, which are widely used in many developed countries, are missing clinically important streptococci. The best features to predict streptococcal sore throat presenting in primary care deserve revisiting since features not commonly used in diagnosis (for example, rapid presentation, severity of inflammation) may be useful clinically.
METHOD
Inclusion
Health professionals in general practices in the south and central areas of England recruited adults or children aged ≥5 years presenting with acute sore throat (<2 weeks), where the sore throat was the predominant clinical feature (or where the clinician felt that the pharyngitis was driving the illness presentation) and with an abnormality on examination of the throat (erythema with or without pus and anterior cervical glands), similar to a previous study in primary care.17 Exclusions were other non-infective causes of sore throat (for example apthous ulceration, candida, or drugs), or unable to consent (for example dementia or uncontrolled psychosis).
Clinical data
Following informed consent baseline clinical data were collected by the recruiting health professional.18,7,19 The clinical proforma collected information on age; sex; current smoking status; past history of quinsy;20 data on symptom severity for the symptoms of sore throat, difficulty swallowing, fever, cough, coryza (‘runny nose’) headache, muscle ache, abdominal pain, diarrhoea vomiting, earache (each symptom was rated by the patient as 0 = no problem, 1 = slight problem, 2 = moderately bad problem, or 3 = severe problem). The doctor or nurse documented examination findings for oral temperature using TempaDOT™ thermometers,21 the severity of tonsillar and pharyngeal inflammation, the presence of cervical glands, tonsillar exudate, fetor and palatal oedema.7,18,19
Throat swabs
At the training session in each practice clinicians were instructed in standard study procedures, including how to take a throat swab. Swabs were taken by the clinician and sent to a central laboratory for culture and sensitivity of all significant pathogens in line with National Standard Operating Procedures.22,23 Mean time between specimen collection and receipt at laboratory was 2.9 days (data incomplete for 13 samples). The swabs were inoculated onto a blood agar plate and staph/strep agar plate (E&O Laboratories Ltd, Bonnybridge) and spread for single colonies. Plates were incubated anaerobically for 48 hours.22,23 Plates were read after 24 hours incubation and negative cultures reincubated for an additional 24 hours. Suspected beta-haemolytic streptococcal isolates were identified via visual analysis of colony morphology and Lancefield grouping (PathoDx® Strep Grouping Kit, Oxoid, Basingstoke), in accordance with the National Standard Operating Procedures.22,23 Antibiotic sensitivities were conducted using disc diffusion techniques.24
Sample size
The study was also designed to assess the accuracy of rapid antigen detection tests (not reported here) and the RADT element of this study limited the sample size: assuming sensitivity of 90%,25 and that 25% of individuals have streptococcus (based on the study’s piloting), to estimate the sensitivity with 95% confidence intervals (CIs) of +/-5% 139 patients with streptococci were needed: 556 patients in total. To estimate the predictive value of clinical variables a subgroup of 139 patients with a clinical presentation not associated with streptococcal infection would provide estimates of a negative predictive value of 90% with 95% CIs of +/-5%, and a subgroup of 93 individuals with streptococcal infection would provide estimates of a positive predictive value of 60% with 95% CIs of +/- 10%. The study estimated that a sample size of 455 patients would be sufficient to detect of an odds ratio of 2 (assuming an alpha of 0.05 and beta of 0.2) for variables with a prevalence of 20–65% among patients without streptococcal infection.
Analysis
Clinical variables were included in a logistic regression model to assess their association with the presence of streptococci. Forward selection was used: variables were included if significant at the 10% level and retained in multivariate analysis if they remained significant at the 5% level, with no evidence of collinearity. All variables significant in univariate analysis were checked again in the final model. Cases with missing data for a particular analysis were excluded. For variables with several levels (for example, sore throat) to facilitate use in a simple clinical score (that is, ease of implementation), a cut off was normally made at or near an odds ratio of 2. Continuous variables were dichotomised using previous cut-offs (age <10 years; prior duration longer than the median of 3 days7). For duration there was a progressive reduced likelihood of infection with group A streptococci with longer prior duration however the study dichotomised at the median for ease of implementation. The study also presents a version of the final model with more categories for each variables (that is, not dichotomised, using ordered categorical variables). Such a model could potentially be used with computerised practices to document more precisely risk of streptococcal infection. Although for non-group A streptococci to date the study assumed Lancefield groups B, D, and also pneumococcus were not to be counted as significant pharyngeal pathogens, given the ongoing debates about this issue3 the study also presents the multivariate analysis when these streptococci are included as potentially significant pathogens. Given the higher asymptomatic carriage rates of streptococci in children this study did not include age in the final multivariate models.
RESULTS
A total of 70 GPs and practice nurses in south and central England recruited 606 patients from March 2007 until January 2008. Recruitment took a year due to the limited duration of recruitment in many practices; the median time spent recruiting was 3 months. However, the median recruitment rate (the number of patients/months recruiting) was 4.7 patients per month; close to the expected rate from national data.26 Sixty seven of 605 patients (11%) were aged <10 years, 106/604 (18%) were smokers, and 109/605 (32%) were male.
Of 606 patients recruited, 592 had microbiology results and 567 had useable baseline clinical data.
Pathogenic streptococci were found in 202 patients (34%): of these 136 were Lancefield group A beta-haemolytic streptococci, and 66/202 (33%) were non Lancefield group A streptococci: mainly groups C (n = 27), G (n = 18), and B (n = 17), but also D (n = 3) and pneumococcus (n = 1).
Patients who had group A beta-haemolytic strains when compared with all other patients were more likely to have short duration of illness (≤3 days), anterior cervical glands, be aged <10 years, have a moderately bad or worse sore throat, moderately bad or worse muscle aches, fever during the last 24 hours, and not have a bad cough (Table 1). Although purulent tonsils were predictive in univariate analysis they did not independently predict the presence of group A streptococci (Table 1).
Table 1 Clinical variables in patients with Lancefield Group A beta-haemolytic streptococci (n = 136) compared with all other patients
Patients with group C and G beta-haemolytic strains presented with similar clinical features to individuals with group A beta-haemolytic strains (Table 2 and Figure 1), with the exception of age, where children were very unlikely to have C and G strains and more likely to have group A strains. Many of the same features associated with group A strains were associated with the presence of C and G streptococci.
Figure 1 Percentage of symptom and signs at presentation for patients with group A Streptococci, group C or G streptococci, and no group A, C or G streptococci.
Table 2 Clinical variables in patients with Lancefield group C and G beta-haemolytic streptococci compared with patients having no growth of group A, C, or G beta-haemolytic streptococci
The independent clinical features associated with combined group A, C, and G streptococci were: prior duration 3 days or less, moderately bad or worse muscle aches, moderately bad or worse sore throat, and the absence of a bad cough, severely inflamed tonsils, aged <10 years old, fever during the last 24 hours, and anterior cervical glands (Table 3). The absence of a ‘runny nose’ (coryza) was also very close to significance in multivariate analysis (P = 0.054). There were too few patients with group B infections to assess with confidence but many had similar features in terms of a severe sore throat (17/17; 100%) purulent tonsils 10/17 (59%), 13/16 (81%) cervical glands, short prior duration 9/17 (53%), the absence of a bad cough 14/17 (82%), and no runny nose 13/17 (76%); however fever 10/17 (59%) was similar to those patients where no streptococci were isolated and only 3/17 (18%) had severely inflamed tonsils. A fuller model — where the variables in Table 3 are not dichotomised — is presented in Appendix 1 which supports the overall findings from the simpler models, with the exception of coryza (which is no longer significant).
Table 3 Clinical variables in patients with non-group A (C and G) and Group A compared with patients having no growth of C,G or A streptococci
There is considerable variation as to how well each indicator performs in helping rule in or rule out the presence of streptococci (Figure 2).
Figure 2 Log univariate likelihood ratios (log of likelihood ratio for a negative test (logLR-) and log of likelihood ratio for a positive test [log LR+]) for individual symptoms and signs at presentation for patients with either group A, C or G betahaemolytic streptococci.
Recruitment bias
Comparing patients from higher recruiting doctors (higher than the median: average 11.8 patients per month) with patients from lower recruiting doctors (an average 2.6 patients per month) there was no difference in the number of features that predicted streptococcal infections in multivariate analysis (Table 3, respectively a mean of 3.3 features and 3.4 features), suggesting little or no recruitment bias based on clinical characteristics.
DISCUSSION
Summary
This study documents that non-group A strains are a common cause of streptococcal sore throat in primary care, have a similar symptomatic presentations to group A presentations, and that the best predictors of streptococcal infection may not include some of the features traditionally used.
Strengths and limitations
The power to detect variables associated with group C and G streptococci was limited, although this is one of the largest studies to assess a broad range of clinical variables and the study found similar features to a smaller study reported previously.11 Missing data was minimal (less than 5% for any analysis), and although consecutive recruitment of cases was difficult to enforce in practice, there was very little evidence of recruitment bias compared with expected rates of recruitment from national samples. Selection bias is a potentially important issue among low recruiting doctors, but the study found no evidence of clinical differences comparing higher and lower recruiting doctors. Overall fewer children than expected from historical data sets27 were recruited which probably reflects reluctance of parents and/or GPs to expose children to a throat swab, but since the study elected not to include age per se in the model, the impact of this should be slight.
The time between taking a swab and receipt at the laboratory was slightly longer than expected, but Lancefield groups A, C, and G are not particularly sensitive to transport conditions, and the study found relatively high percentages of streptococci compared to previous literature. Since there was no indication from sentinel practices or microbiology laboratories to indicate a streptococcal epidemic, the high streptococcal percentages and more florid clinical signs compared to previous studies in a similar geographical area28 may indicate changing consultation thresholds. Although the study indicates clinical features not traditionally incorporated in making diagnostic assessments may possibly not be important, further data sets are needed before recommending a key variable set.
The way variables are operationalised may also be important (for example, McIsaac8,29 uses tonsillar swelling or exudate, Centor18 just exudate, the study chose exudate since swelling is not necessarily an acute feature). The study used intermediate cut points where indicated rather than the extremes of each scale; however, since the judgement of intermediate points may be more variable, using the extremes (none or very) may be more sensible for developing a clinical prediction rule.
Comparison with existing literature
Traditionally clinicians have been predominantly interested in group A beta-haemolytic streptococci due to their association with major non-suppurative adverse outcomes, particularly rheumatic fever.5 Hence the clinical predictors of group A infection5–7 — especially pus, cervical nodes, a history of fever and no history of cough30 — have been widely used in clinical guidelines.2,4,31 Historical comparisons tentatively suggest that these variables may identify a group of patients who are more likely to benefit from antibiotics.1 The study confirmed the importance of cervical glands and the absence of a bad cough and of fever.7,8 However, this study documents that although the feature of purulence is associated with the presence of group A streptococci in univariate analysis, in this data set it is not independently predictive – and other features may be important, particularly the severity of both the sore throat and the inflammation, the prior duration (reflecting the more rapid severe onset), muscle aches, and possibly the absence of coryza. Some of these features have been identified previously in studies from typical primary care settings7,11 but previous studies have been limited by lack of multivariate analysis or limited power.
The clinical presentation of group C and G streptococci suggests strongly that not only are these presentations unlikely to be due to commensal carriage, but that they are causing a similar clinical syndrome to group A streptococci. This supports those studies which observed similar symptomatic presentation.11,3 If group C and G streptococci are clinically important, then rapid streptococcal antigen tests (which are targeted at group A beta-haemolytic streptococci only) will miss a significant proportion of streptococcal infection.
Implications for practice
Group C and G streptococci present with symptomatic illness in a similar manner to group A streptococci. Rapid antigen detection tests which are widely used in many developed countries to detect group A streptococci will miss these organisms. The best features to predict streptococcal sore throat presenting in primary care also deserve clarification since features not commonly used in diagnosis (for example, rapid presentation, severity of inflammation) may be useful clinically.
Acknowledgments
The PRISM investigators were: University of Southampton — Paul Little, Ian Williamson, Mike Moore, Mark Mullee, Man Ying Edith Cheng, James Raftery, David Turner, Jo Kelly, Jane Barnett, Karen Middleton, Gerry Leydon; University of Birmingham — Richard Hobbs, Brendan Delaney, Olga Kostopoulou, Richard McManus, Razia Meer-Baloch; University of Oxford — David Mant, Paul Glasziou, Sue Smith, Diane Coulson; Health Protection Agency — Cliodna McNulty, Peter Hawtin. We are grateful for the doctors and patients who gave their time to make this possible.
Appendix 1. Clinical variables in patients with group A, C or G beta-haemolytic streptococci compared with patients having no growth of C,G or A beta-haemolytic streptococci using more levels for variables.
Notes
Funding
This study was funded by the UK Health Technology Assessment Programme, reference number HTA 05/10/01.
Ethical approval
Approved by MREC, reference number 06/MRE06/17.
Provenance
Freely submitted; externally peer reviewed.
Competing interests
CM leads the writing and review of the HPA management of Infection guide for primary care and is a member of the Department of Health Advisory Committee on Antimicrobial Resistance and Healthcare-Associated Infections and the Prescribing Sub-Group and Chair of the Public Education Sub-Groups of this Committee, and was also a member of the NICE self-limiting respiratory tract infection guideline development group.
- Received December 20, 2011.
- Revision received January 31, 2012.
- Accepted June 14, 2012.
- © British Journal of General Practice 2012
