Article Text

Download PDFPDF

An outbreak of scarlet fever in a primary school
  1. K H Lamden
  1. Cumbria and Lancashire Health Protection Unit, Chorley, Lancashire, UK
  1. Correspondence to Dr Kenneth Lamden, Cumbria and Lancashire Health Protection Unit, York House, Ackhurst Business Park, Chorley PR7 1NY, Lancashire, UK; kenneth.lamden{at}


Scarlet fever, due to infection with an erythrogenic toxin-producing Group A streptococcus, is an uncommon and generally mild illness, although serious sequelae do occur. In March 2009, 57 of the 126 (45%) pupils in a primary school in Lancashire, UK developed scarlet fever over a 4-week period. Infection was transmitted via direct contact between pupils, particularly among the youngest pupils. A significant degree of transmission also occurred between siblings. The median number of days absent from school was 3 (range 1–10 days). No children were hospitalised. Control measures, including hygiene advice to the school and exclusion of pupils for 24h while initiating penicillin treatment, were ineffective. The outbreak occurred against a background of an unusually high incidence of invasive Group A streptococcal infection. While there are national guidelines for the control of invasive disease, none exist for the control of scarlet fever outbreaks. This prolonged outbreak of scarlet fever highlights the need for an evidence based approach to outbreak management.

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.


Scarlet fever is a clinical syndrome presenting as sore throat and rash due to infection with an erythrogenic toxin-producing strain of Group A β-haemolytic streptococcus (GABHS), Streptococcus pyogenes. Infection can be associated with a range of streptococcal emm types, based on sequencing of the emm gene that encodes the major virulence factor, the M protein.

Scarlet fever is generally a mild illness with the main morbidity due to the long term sequelae of rheumatic fever (0.6%)1 and acute renal failure due to glomerulonephritis (1.8%).1 However, the risks of sequelae are based on data from the 1960s, which may not be clinically relevant today. Complications are more common in developing countries. The incubation period is usually between 1 and 3 days and the infectious period if left untreated is 10–21 days.2 Treatment with a 10-day course of penicillin aims to reduce suppurative complications, communicability and carriage. Antibiotic treatment virtually ends transmission after 24 h,2 and may reduce rheumatic fever3 but does not prevent post-streptococcal glomerulonephritis.3

National Institute for Health and Clinical Excellence (NICE) guidance on the treatment of sore throat4 is based on the minimal use of antibiotics, unless clinically indicated by the Centor criteria,5 and does not mention the use of throat swabs at all. However, similar guidance in France, Finland and North America places more emphasis on the need to diagnose GABHS using a rapid antigen test, and penicillin therapy to prevent rheumatic fever.6

What is already known on this topic

  • Scarlet fever notifications have declined steadily over the past two decades.

  • An unusual increase in invasive Group A streptococcal disease was observed in 2008/2009 and this was mirrored by an increase in notifications of scarlet fever.

  • Large outbreaks of scarlet fever are unusual.

What this study adds

  • Outbreaks of scarlet fever in enclosed settings may be prolonged despite implementation of standard control measures.

  • Treatment of contacts is a potential control measure but its value is uncertain.

  • In an outbreak, two thirds of eligible children received antibiotics and one third received a full 10-day course.

Scarlet fever is considered to be a disease of the pre-antibiotic era and to have fallen to low levels due to the combined effect of antibiotics and improved living standards. Most cases are sporadic, although the organism can transmit easily under certain conditions and outbreaks do still occur.7,,10 It is likely that some outbreaks remain undetected. Clinical notifications of scarlet fever average approximately 2000 per year in England and Wales. An unusual increase in laboratory reports of invasive Group A streptococcal disease was observed in 2008/2009 and this was mirrored by an increase in notifications of scarlet fever.11

This report describes an outbreak of scarlet fever in a primary school and highlights the need to develop evidence based guidance for outbreak control.


The outbreak occurred in March 2009 in a primary school in a socially deprived area of Chorley, Lancashire. The outbreak was reported to the local health protection unit by the health visitor attached to the school, who was concerned about the increasing number of children absent from school. The clinical presentation included pharyngitis, flushed facial appearance with inflammation and soreness around the mouth, and an extensive florid rash. Not all children had a rash. GABHS was isolated from throat swabs from two of the initial cases.

A team, comprising the consultant in communicable disease control (CCDC), the head teacher and the school health visitor was formed to manage the outbreak. All parents received a letter excluding symptomatic children from school until they had received at least 24 h of penicillin treatment.12 General practitioners (GPs) were advised of the need for 10 days' penicillin treatment for cases. Control measures were instituted within the school. This included closure of the water fountains, disinfection of the children's water bottles and extra washing of toys. Hand washing was extensively promoted.

Throat swabs were obtained from as many of the initial cases as possible via clinics held in two GP practices. An epidemiological investigation was undertaken to describe the nature of the outbreak, to determine the likely method(s) of disease transmission and to evaluate control measures.


For the epidemiological study potential cases were identified from the school absence register. Cases were classified on the basis of microbiology and symptoms reported by parents to the school. Cases were defined as confirmed (clinical symptoms plus throat isolate of GABHS), probable (rash and none or negative throat swab) or possible (sore throat alone). Information on demographic details, school class, date of first absence and length of absence was also obtained from the absence register. All analyses were conducted in EpiData Analysis v 2.1 (EpiData Association, Odense, Denmark).

In order to evaluate antibiotic prescribing advice, GP practices were contacted after the outbreak and asked to provide information on the number of antibiotic prescriptions written and the length of treatment.


The outbreak affected 57 out of the 126 children in the school, giving an attack rate of 45%. Group A streptococcus type emm3 was isolated from 9/13 throat swabs obtained. Of the four children with negative throat swabs, two had a sore throat and two had a rash plus sore throat. One of these children had a sibling with GABHS. In addition to the nine confirmed cases, there were 12 probable and 36 possible cases.

Twenty-nine (51%) of the cases were female and 28 (49%) were male. The median age was 8 years with a range of 4–11 years. The median number of days absent from school was 3 with a range of 1–10 days. No children were hospitalised. There was significant secondary spread within households with 25/57 (44%) cases occurring in 12 households.

The index case was absent from school on 3 March 2009 and the outbreak was reported to the CCDC on 10 March 2009. The outbreak persisted over a period of 4 weeks (figure 1). No cases were reported over weekends when the school was closed for 2 days.

Figure 1

Scarlet fever cases by date of first absence and school year.

The highest attack rates occurred in infant classes 0, 1 and 2 which contained children aged 4, 5 and 6, respectively (table 1). The outbreak curve (figure 1) suggests that the outbreak initially affected the youngest children and then progressed throughout the school. This suggests that person-to-person contact was the main route of transmission.

Table 1

Scarlet fever attack rates by school year

Evaluation of antibiotic prescribing

A total of 81% (46/57) children attended their GP, and 80% (37/46) of these were prescribed an antibiotic. Therefore, 65% (37/57) of the cases received an antibiotic. The length of treatment was 5 days in 54% (20/37) and 10 days in 46% (17/37).


This report describes a large outbreak of scarlet fever due to GABHS infection in a primary school. The outbreak occurred against a background of an unusually high incidence of invasive Group A streptococcal infection in England and Wales.11 Scarlet fever notifications were also high and numbered 4147 in 2008/2009, their highest level for 13 years.13

The outbreak strain was identified as S pyogenes Group B type emm3, which was the prevalent Group A streptococcal strain isolated from invasive disease cases at that time. Three months prior to the outbreak, a child from the same locality who attended a different school had died of invasive GABHS type emm3 infection. This raised the level of concern of the CCDC during the outbreak. The relationship between pharyngitis and progression to invasive disease is still uncertain and remains to be elucidated.

The clinical case definition identified 57 outbreak cases of whom 13 had throat swabs taken. It is not essential to obtain samples from every case in an outbreak once the causative organism has been confirmed and continued sampling can be an inefficient use of resources. It is possible that some of the cases had infection with co-circulating pathogens, however the typical pattern of the outbreak curve together with its progression through the school years, suggests that most cases were part of the outbreak.

In spite of the outbreak being recognised at an early stage, with just six children unwell, control measures had little impact on disease transmission. The control measures were implemented as well as they could have been. Exclusion of cases from school was rigorously enforced and although the minimum exclusion was 24 h, in practice it was usually 48 h. Parents and GPs responded well to the treatment advice with two thirds of children receiving antibiotics. The fact that one third of cases did not receive antibiotics at all and that only 54% of those receiving antibiotics were prescribed a 10-day course, probably contributed to ongoing transmission. In addition, lack of compliance with therapy is likely to have reduced the proportion of children receiving effective eradication therapy even further. Achieving greater compliance with therapy would, however, be a considerable challenge.

The continuance of the outbreak was probably due to carriers of GABHS in the school and household spread, as suggested by the high proportion of linked family cases. It is also possible that cases in the wider community helped to sustain the outbreak. Work from the 1950s showed that between 25% and 50% of sibling contacts will develop GABHS and that delay in treatment of the primary case beyond 2 days is associated with a rapidly increasing attack rate in sibling contacts.14 Searching for and treating carriers may seem to be a rational control measure, however it is time consuming and is not backed up by convincing evidence of effectiveness.15 Carriers appear less likely to transmit the organism than acutely infected patients and are at significantly reduced risk of developing rheumatic fever.15 It may also be more difficult to eradicate the organism from a carrier than from an acute case.16

There are surprisingly few reports of scarlet fever outbreaks in the literature. In 1992, an outbreak affected 12 children in a Swedish day-care centre over a 3-week period.7 Screening of 31 children 2 weeks into the outbreak revealed 61% were colonised with GABHS. Control was based on repeated throat cultures with antibiotic treatment of cases and asymptomatic carriers. Contamination of the environment with GABHS was demonstrated, however this was thought to be an unlikely source of infection. Although streptococci are a relatively hardy organism and can survive in the environment, for example, in bedding17 and dust,18 they dry within hours and appear to be non-infectious for pharyngitis.19 However, environmental persistence is a factor in transmission of streptococcal pyoderma.20 In 2000, 21 cases of GABHS, including scarlet fever, pharyngitis and impetigo, occurred in a Dutch primary school over a 4-week period.8 Cases received treatment with a 3-day course of azithromycin. In 2003, 13 children in a Perth primary school acquired scarlet fever over a 5-week period.9 The outbreak ceased after the treatment of pharyngeal carriers. An outbreak of 16 cases of scarlet fever in a nursery in Southwest England was reported in 2006.10 Control was based on exclusion of cases for 5 days after commencement of penicillin.

Scarlet fever remains something of a conundrum. It is a notifiable disease but there is large under-reporting and some outbreaks are probably undetected. Is it important to detect outbreaks, and if so how is surveillance best undertaken? Nurseries and schools would need to be encouraged to report outbreaks, followed up by throat swab or rapid diagnosis streptococcal antigen test. This test has not proved popular in the UK due to a low level of sensitivity.21 Either test would entail additional time and costs. Increasing the detection of sporadic cases would require increased testing of children with a rash and compatible illness by GPs. This would be more likely to happen if a significant rheumatic fever risk could be demonstrated.

The question remains, how aggressive should CCDCs be in attempting to curtail an outbreak of scarlet fever in an enclosed setting, bearing in mind the potential for invasive disease and long term sequelae? Evidence based guidance to support decision making and intervention would be helpful. Current guidance relates to the management of close community contacts of invasive disease22 and not to outbreaks of pharyngitis or scarlet fever. Guidance would need to consider whether swabbing and treatment of contacts, or prophylaxis of contacts, has a role to play, and whether 24 h exclusion of young children is sufficient. It would also need to consider antibiotic policy taking compliance and completion rates into consideration. Penicillin is dispensed as a 5-day course of suspension and parents need to attend a pharmacy twice as a 10-day course of antibiotics is recommended. Low compliance or completion may lead to prolonged transmission in the short term and an increased incidence of rheumatic fever in the longer term.

One of the problems attempting to synthesise available evidence is that much comes from the 1950s and 1960s when sequelae of GABHS in developed countries were more common than they are today. Nevertheless, collation of current knowledge of the biology of GABHS, complication rates and the behaviour of carriage could form an initial evidence base for outbreak control interventions.

Outbreaks of scarlet fever will continue to occur and may be prolonged. Although illness is generally mild, there is concern about sequelae and outbreaks generate anxiety among parents. CCDCs, GPs, teachers and parents would all benefit from a standardised and evidence based approach to control.


The author gratefully acknowledges the assistance of Claire Smith, head teacher at Clayton Brook Primary School, Hilary Sangster and Sheila Jones who obtained the throat swabs, and the Streptococcus and Diphtheria Reference Unit of the Health Protection Agency Centre for Infections for advice and strain typing.



  • Funding The study was conducted as part of an outbreak investigation and as a result the author was funded by the Health Protection Agency.

  • Competing interests None.

  • Provenance and peer review Not commissioned; externally peer reviewed.