Summary

Objective: Sore throat is a common condition in general practice for which antibiotics are frequently prescribed. This study looks at the bacteriology of sore throat and the correlation between the clinical features and a positive bacterial culture.

Design: This is a prospective study done in a general practice by a single doctor who looked at the age, sex, temperature, cervical lymph nodes, tonsil state, and nasal condition of all patients with sore throat as their chief complaint between July 1999 and July 2000. Throat swabs were taken from all these patients.

Subjects: All patients who attended the clinic from 2 July 1999 to 26 July 2000 and who presented with sore throat as their chief complaint.

Main outcome measures: Temperatures above or below 101 degrees F, presence or absence of cervical lymphadenitis, presence or absence of pustular tonsillitis, presence or absence of rhinitis, age and sex were looked at in relation to a positive or negative culture.

Results: 171 throat swabs were taken from 171 patients. 26 (15%) were positive for bacteria. The main pathogens were Group A and Group G streptococcus. Both groups were 100% sensitive to penicillin but a significant percentage were resistant to tetracycline and erythromycin. The only clinical feature that has a statistical relationship with a positive bacterial culture was exudative tonsillitis.

Conclusion: Streptococci cause less than 15% of sore throats when presented as the chief complaint. Group G streptococcus is almost as commonly cultured from these throats as Group A streptococcus and its significance is uncertain. In view of the limited nature of the disease, the resistance of the bacteria to erythromycin and tetracycline, and the low incidence of Group A streptococcus as the cause of sore throat, primary care doctors should exercise restraint in prescribing antibiotics for their patients with sore throats.

Keywords: Sore throat, pustular tonsillitis, Group A beta-haemolytic streptococcus, Group G streptococcus, antibiotic resistance

摘要

目的：咽喉痛是門診常見的疾病，醫生常給予抗生素 治療。本研究的目的在於探討咽喉痛的細菌學、臨床徵狀和陽性細菌培養的關係。

設計：這是一個前瞻性的研究。紀錄從1999年7月至2000年 7月，由同一位門診醫生所診治以咽喉痛為主訴的病人的年齡、性別、體溫、頸部淋巴腺、扁桃腺 和鼻腔的情況，並拭其咽部作細菌培養。

對象：從1999 年7月2日至2000年7月26日為止，所有因主訴咽喉痛而求診的病人。

測量內容：研究體溫高於或低於華氏101度，有無頸部淋巴腺炎、化膿性扁桃腺炎、鼻炎，以及年齡、性 別和這些因素細菌培養結果的關係。

結果：171位病人均做咽部細菌培養。 15%（26個）呈陽性結果。主要致病菌是A型和G型鏈球菌，它們 全部對青黴素有反應，但有相當比例對四環素和紅霉素有抗藥性。唯一有統計學關係的臨床表現是化膿性 扁桃腺炎與陽性細菌培養。

結論：鏈球菌感染引發的咽喉痛少於 15%。咽部細菌培養發現，G型和A型鏈球菌一樣普遍，但無明確意 義。由於疾病的病情輕微，細菌對紅霉素和四環素有抗藥性，而且 A型鏈球菌並不是咽痛常見的原因，基層醫生治療咽喉痛時應儘量少用抗生素。

主要詞彙：咽喉痛、化膿性扁桃腺炎、A型b溶血性鏈球菌、 G型鏈球菌、抗菌素抗藥性。

---

Introduction

Sore throat is a commonly encountered problem in general practice. In community paediatric practices, it is responsible for 11% of all office visits and 24% of all acute infections.¹ On average, in the United Kingdom, a general practitioner sees about 600 upper respiratory tract infections including 100-150 sore throats yearly. In a local morbidity survey done in 1994, upper respiratory infections accounted for 34.6% of visits and tonsillitis for 1%.²

Medical authorities agree that at any age, most cases of pharyngitis (particularly if the milder ones are included) are viral (particularly the adenovirus). Group A beta-haemolytic streptococci are the main bacterial pathogens. Rarely, Neisseria gonorrhoea and Mycoplasma pneumoniae contribute to the aetiology. Pneumococci, Haemophilus species, and occasionally staphylococci are also isolated from throat swab specimens. Most would suggest that in children less than 3 years of age, pharyngitis is almost always viral.¹

In one local study, out of 95 throat swabs, 32 grew Haemophilus influenzae, 18 beta-haemolytic streptococcus, 12 Staphylococcus aureus, 11 Enterobacters, 7 Acinetobacter, 5 Klebsiella, 3 Pseudomonas and 1 E. coli.³ In university students, 15-26% of pharyngitis cases have been noted to be streptococcal.¹ In the local study in 1992, all beta-haemolytic streptococci were susceptible to penicillins or cephalosporins tested. Sixty-one percent were resistant to tetracycline but 17% or less to erythromycin, co-trimoxazole or chloramphenicol.³ There is now an increasing awareness that, for most acute respiratory illness (common cold, sore throat, otitis media, sinusitis, bronchitis) treatment with antibiotics is likely to be of only marginal benefit to individual patients. There is also growing concern about the development of antibiotic resistant organisms. Despite this, 12 million antibiotic prescriptions were made for colds, upper respiratory tract infections and bronchitis in 1992 in USA alone, accounting for 21% of all antibiotic prescriptions to adults in the country in that year. Prescribing antibiotics for upper respiratory infections is therefore a worldwide issue and not restricted to Hong Kong alone.

The correlation of symptoms, throat appearance and even throat swab growth with an infection is, at best, imperfect. The classical triad of fever, exudate and tender cervical nodes is present in only 15% of cases. Although the presence of palatal petechiae and vesicles is suggestive of viral infection, and tonsillar exudate in patients under 15 years of age is more likely to be streptococcal, while in those over 15 it is more likely to be glandular fever, it is generally felt that the aetiology cannot be predicted from appearance with any degree of certainty. Prescribing habits, not surprisingly, vary between the extreme views of "antibiotics for all" and "no antibiotics at all" with many GPs occupying the middle ground and prescribing for various indications of their own choosing (e.g. fever, malaise) which again correlate poorly. Professor Hull suggested that if an exudate is not visible, the infection is almost certainly viral. The presence of pus suggests, but is by no means diagnostic of, bacterial infection. Many physicians would think that the sore throat is viral when there is associated rhinitis or laryngitis.⁴

The majority of upper respiratory infections are viral in origin and antibiotics are not indicated. If Streptococcus pyogenes is suspected, on account of fever, tonsillar exudate and cervical lymphadenopathy, then antibiotics may be given.^5,6 Confirmation by culture of throat swabs is an ideal that is seldom possible in general practice. Furthermore, identification of a pathogen from a culture specimen does not necessarily have a causative implication, as this may be part of the "normal" flora for that particular individual. On the other hand, the absence of an identifiable pathogen does not rule out the possibility of bacterial infection.⁹

The recent concern about antibiotic abuse and the absence of local prospective study on this topic prompted this project to be conducted.

Method

Participants

A prospective study was conducted on all patients with sore throat as the chief complaint seen in a primary care setting clinic in Wong Tai Sin by the same physician on Mondays to Thursdays from 2 July 1999 to 26 July 2000. Sore throat secondary to cough or those with definite symptoms of influenza and common cold were excluded.

Clinical and bacteriological measurements

Ear temperatures were taken on the participants. Those with temperature of 101 degrees F. or higher were included in one group and those with less in another. The throat was looked at in all patients and the presence or absence of exudative tonsillitis noted. Enlarged and/or tender neck glands in the anterior triangle were felt for in all patients and their presence or absence noted. The presence or absence of rhinorrhoea was particularly asked for and examined for in all patients.

Patients were informed of the nature of the study and that the tests were free. The throat swabs were all taken by the same doctor from the posterior pharyngeal wall and the tonsillar crypts bilaterally using an applicator, and transported in the Venturi Transystem to the Department of Health, Institute of Pathology (Microbiology) for culture and sensitivity test within 24 hours. They were inoculated in Columbia blood agar and incubated both under 5% carbon dioxide and anaerobic conditions overnight. Beta-haemolytic colonies identified with streptococcal morphological features were grouped by latex agglutination test. The NCCLS guideline was used to determine the sensitivity of beta-haemolytic streptococci to penicillin, erythromycin, and tetracycline.

Results

171 throat swabs were taken from 171 patients between July 2, 1999 and July 26, 2000. Twenty-six cultures were positive for bacteria (15%) and 145 were negative (85%). Of the positive cultures, one reported a pure growth of staphylococcus aureus and another grew a mixture of staphylococcus aureus and streptococcus pyogenes. (The laboratory reported staphylococcus aureus and not streptococci only because it was also involved in studying carrier state of staphylococcus aureus at that time. Subsequent to the period of this study, staphylococcus aureus was no longer reported in throat swab culture results).

The patients from the two groups were looked at separately. Of the 26 positive cultures, 12 yielded Group G haemolytic streptococcus (46%) and 13 grew Group A haemolytic streptococcus (50%) (including the mixed growth mentioned above) and one reported as staphylococcus aureus (see above) that was probably not the pathogen of the sore throat.

All the streptococci were sensitive to penicillin, 75% of group G Streptococci and 46% of Group A Streptococci were sensitive to erythromycin. All Group G haemolytic streptococci were resistant to tetracycline while only 23 % of Group A streptococci were resistant.

The clinical features of both the culture-positive group and the culture-negative group were examined. The results are tabulated below:

Positive cultures (26)   Negative cultures (145) Fever >101 degrees F 58%   29% Exudative tonsillitis 42% 10% Cervical lymphadenopathy 23% 10% Rhinorrhoea 19% 47%

Of the 26 patients with positive cultures, 14 were female and 12 male and their age varied from 5 to 66. More than half had a temperature of more than 101 degrees F and slightly less than half had exudative tonsillitis. About one quarter had cervical glands and only 19% had rhinorrhoea. Of the 145 with a negative culture, their age varied from 1 to 81 and the female to male ratio was 84 to 61. Only slightly more than one quarter had a significant temperature, and only about 10% had exudative tonsillitis and neck glands respectively. Rhinorrhoea was a complaint in about half of the patients.

Analysis of data

Using the SPSS, the age, sex, temperature, cervical lymphadenopathy, rhinitis and exudative tonsillitis were matched with a positive culture. Only exudative tonsillitis showed a positive co-relation with positive culture. (p=0.001) The other variables were not significantly related to a positive bacterial culture.

Discussion

Incidence

Bacterial sore throats make up about 15% of pharyngitis in this study. Almost all the bacteria cultured were streptococcus. This is in line with most of the authority articles where figures of 5-10% of all cases are quoted. The reports from the particular laboratory did not mention bacteria other than streptococcus and staphylococcus because it did not consider the other organisms pathogenic in the throat even if they were grown.⁶ (The reason for including staphylococcus was given).

Bacteriology

Although Group A streptococci is considered the most common pathogen for acute pharyngitis or tonsillitis, 46% of the cultured streptococci in this study were Group G streptococci.

Age

Some authors suggested that in children less than 3 years of age, pharyngitis is almost always viral.¹ In this study, the youngest child to have streptococcal sore throat is 5 years old. Acute pharyngitis and tonsillitis are reported to be most frequently seen in children aged 4 to 7. In this series, streptococcal sore throat was not shown to be co-related with any particular age group.

Clinical features

Although it is reported that bacterial sore throats present more often with exudative tonsillitis, higher fever, cervical lymphadenopathy and prostration, and less often with rhinorrhoea or other systemic upsets compared with non-bacterial sore throats, it is difficult for clinicians to judge whether a sore throat is bacterial based on clinical features alone. In this study, only exudative tonsillitis was found have statistical significance when matched against a positive culture.

Antibiotic of choice for sore throat

The sensitivity tests show that penicillin is still the antibiotic of choice for streptococcal sore throats. There is already some degree of resistance to erythromycin and marked resistance to tetracycline. This is in line with findings elsewhere including Australia where the rate of macrolide resistance is 8% (National Antibiotic Resistance Surveillance Programme).⁷ The over 50% resistance to erythromycin may correspond to the amount of macrolide used in Hong Kong. This phenomenon was also observed in Japan where more than 60% of streptococcus isolates were resistant to erythromycin during a period where macrolide use was increasing.¹ In Finland, the level of erythromycin resistance was also very high. After restrictions on the use of macrolides this resistance was significantly reduced.⁷ In Hong Kong, physicians should be advised to restrict their use of macrolides and tetracyclines in pharyngitis, unless the patient is penicillin-sensitive. They should also be advised to choose the narrowest- spectrum antibiotic, penicillin V, for streptococcal sore throats, which is least likely to affect the patient's normal flora which is not the target of therapy. Consequently, superinfections such as candidiasis and selection for resistance in the normal flora which is not the target of therapy are less likely. Thus penicillin V is more appropriate than amoxycillin.⁴

Throat swabs for culture

Some experts are of the opinion that in acute sore throat, swabs are seldom helpful because S. pyogenes is frequently isolated from children, whether symptomatic or not.⁸ Other experts, however, charged that failure to perform investigations before treatment of an infection is akin to diagnosing and treating a cardiac arrhythmia without performing an ECG. They also suggested that the collection of specimens only when the patient has failed treatment with the initial choice is unsatisfactory. Colonisation rather than infection, of the infected site or of nearby normal flora by organisms resistant to antibiotic is common, and will often generate misleading results in the laboratory.⁴ In Hong Kong, because of the price differential between the private and public sector, the test may cost more than the treatment in private practice and patients would have difficulties in paying for the test. As Professor Hull wrote, "confirmation by culture of throat swabs is an ideal which is seldom possible in general practice".

Latex agglutination test for group A lancefield antigen

Although this claims a sensitivity of 80-90% and a specificity of >95%, and a result is available within 10 minutes,¹⁰ most general practitioners in Hong Kong are unaware of its availability. Wong and Chung recently reported a disappointing 52.6% sensitivity and 98.2% specificity with Accustrip in their local study when 514 tests were done.¹⁰ Moreover, since about half of the streptococcus in the current study belongs to Group G, they may be missed if only group A antigen is tested.

Conclusion

The analysis of a group of patients with acute sore throat as the primary complaint shows that the predominant bacterial organism is group A beta-haemolytic streptococcus. However, group G streptococcus seems to play a significant role as well. Both group A and G streptococci are penicillin sensitive; however, there is a high incidence of erythromycin and tetracycline resistance to suggest over-use of these antimicrobials. It is difficult to be sure whether a sore throat is bacterial based on the clinical features alone and rapid tests may assist the physician in his decision to use antibiotics. The use of latex agglutination tests may need to be advertised but confusion may be created if the organism is group G streptococcus. The only clinical feature found to be significant is exudative tonsillitis. Since streptococcal sore throats comprise only 15% of acute sore throats, general practitioners should be cautious in their use of antibiotics in this condition.

Acknowledgement

The author thanks Professor J A Dickinson for his encouragement and is indebted to Professor Y T Wun for his assistance in the statistical analysis of the data. Special thanks should be addressed to Professor K Y Yuen and his Postgraduate Diploma course in Infectious Disease that initiated the project. Thanks is also due to Miss S L Yim for her patience and secretarial assistance and the Department of Health Institute of Pathology (Microbiology) for performing the bacterial cultures.

Key messages

1. Sore throat is a common condition in general practice.
2. Bacterial sore throat comprises only 15% of all sore throats.
3. The commonest bacterial organisms in sore throat in Hong Kong are Group A and Group G streptococcus.
4. The most significant clinical feature that may be related to bacterial sore throat is exudative tonsillitis.
5. Penicillin V is still the best narrow-spectrum antibiotic for streptococcal sore throats6. Erythromycin and Tetracycline should not be used in sore throat because of the high percentage of resistance.

---

---

References