Group A β-haemolytic streptococci (GABHS) are gram-positive, facultative anaerobic streptococci that are a frequent cause of disease in children. They are present in a large proportion of individuals in the oropharynx and skin1 and may colonise these regions asymptomatically or causing diseases, usually not invasive, such as tonsillitis, scarlet fever or impetigo.2 However, at times GABHS may cause severe invasive disease associated with a high morbidity and mortality.3

In the past 30 years, there has been an increase in the incidence of invasive infection by GABHS in different geographical regions,4–10 including cases of very severe disease,2 the cause of which is not well understood.1,5,10–14 In 2002, 11 European countries started a project for the investigation and surveillance of invasive disease by GABHS (iGASD), that was named Strep-EURO whose goal was to elucidate the epidemiology of this disease in the European continent. Thus, it is known that in Europe the incidence of iGASD ranges between 0.4 and 4.8 cases per 100000 individuals per year, while in the United States it is estimated at 3.54 cases per 100000 individuals per year, with a mortality rate in children of 0–14%. Spain did not participate in the Strep-EURO project, and since iGASD is not a notifiable disease, its incidence in our country is not known.6 In addition, there are other aspects that have not been established in regard to iGASD in children in Spain, such as its epidemiological and clinical characteristics or the differences in comparison with the same disease in adults.

The objectives of our study were: (1) to describe the epidemiological, clinical and microbiological characteristics of iGASD in children treated in a tertiary care hospital over a period of 8 years, as well as to determine risk factors for severe disease and (2) to analyse potential changes in the incidence and severity of this disease in our geographical area during the period under study.

The main findings of the study were the following: (1) we found an increase in the incidence of iGASD in children in our area in the past few years; (2) this observed increase in incidence was not associated with any other changes in epidemiological, microbiological or clinical parameters related to iGASD; (3) the severity of iGASD was considerable, with a high percentage of patients requiring admission to the PICU and surgical intervention, and the risk factors associated with increased severity were age less than 48 months, diagnosis of pneumonia with pleural effusion and elevation of CPR; and (4) the rate of resistance of GABHS to macrolides and clindamycin remained low throughout the period under study.

The incidence of iGASD seems to have been increasing in the past few years worldwide, as evinced by recent studies conducted in children10,12,18 and adults.3,7,18,19 However, the epidemiology of iGASD is not well known in most countries, including Spain, as there are no adequate surveillance systems in place,14 so that the data on its incidence are limited, especially in the paediatric population. In our case series, the incidence tripled from the beginning to the end of the study period, increasing from 5.6 to 18.9 cases per 105 paediatric emergency visits per year, which diverged from the findings of a recent study in a children's hospital in Barcelona, where there was no increase in the incidence of iGASD in the 6 years included in the analysis.20. The incidence in P2 was of 3.6 cases per 105 inhabitants per year, which was very similar to the incidence found by Montes et al.6 in Spain (3.11 cases/105 inhabitants), but lower compared to the incidence reported for the United States (5.3 cases/105 inhabitants/year), although it was greater compared to the mean found in other European countries, such as Sweden (2.1 cases/105 inhabitants in 2012)1 or Finland (2.5 cases/105 inhabitants in the 2006–2010 period).10

The causes of this increase in the incidence of iGASD are not well understood, but increases in risk factors or improvement in microbiological methods may be among them. Varicella has classically been one of the most important risk factors for development of iGASD in children,2,10,21 especially in those that experience a severe invasive cutaneous infection.22 However, our study and other recent works20,23 have not found evidence of an increase in the incidence of varicella, consistent with improvements in the coverage of vaccination against this disease.24 There is evidence that the diagnostic yield of molecular techniques is higher for certain microorganisms compared to conventional diagnostic methods, especially in patients previously treated with antibiotics.25,26 There has also been a proven increase in the accuracy of diagnosis of pneumonia with empyema caused by GABHS in children, both with the use of genetic techniques27 as well as antigen detection methods.28,29 In our hospital, 16S PCR was introduced in 2009, and 19% of the diagnosis of iGASD in P2 were made with this technique. Since many of the infections were suppurative and at least 43% of the children in whom diagnosis was achieved by means of 16S PCR had previously taken antibiotics, the use of molecular diagnostic methods may have played an important role in the increase in diagnosis of iGASD. However, we did not find any clear change in specific clinical forms nor in the isolation of other bacteria, nor epidemiological differences between the 2 periods, so the increase in incidence may be related to other factors that we did not analyse. Thus, numerous studies have found an increase in the severity of iGASD caused by certain genotypes in children,6,10,14,23,30–32 while other authors have reported that the incidence of iGASD may be associated with the susceptibility of specific subpopulations to those strains.33

The most frequent form of iGASD observed in our study was skin and soft tissue infection, which was consistent with many previous studies.6,10,20,34 However, overall, in our case series infections were most frequently located in the ear–nose–throat region. Contrary to other case series, we only found 1 case of primary bacteraemia,10,14,24,34,35 1 case of necrotising fasciitis and no deaths,10,20,32 which could be attributed to the low incidence of varicella and differences between age groups in clinical presentation (fewer cases of necrotising fasciitis and septic shock in children aged less than 10 years)33,36 or to technological advances in the PICU and the multidisciplinary management of these patients.23

Invasive infection by GABHS can cause significant morbidity and a high proportion of affected patients may require surgery and/or admission to the PICU. Our study, one of the largest conducted in Spain, contributes objective data on the severity of iGASD in children in our region. Thus, we found that 18% of patients with iGASD were admitted to the PICU and 63.5% required some form of surgical intervention. These data were consistent with those of previously published paediatric case series, in which the proportion admitted to the PICU ranged between 19% and 57%10,24,35 and the proportion that required surgery between 43% and 64%.10,24,32,35 However, in the study conducted in Barcelona, despite the geographical proximity, the proportion admitted to the PICU was significantly greater (32.7%), while the proportion that underwent surgery was significantly lower (34.6%) compared to our case series in Madrid,20 which could be explained by the higher proportion of children with chronic skin diseases. We ought to underscore that both younger age and the presence of pneumonia were predictors of severe disease in our study, which was consistent with the findings of other authors,5,18,23,24 which suggests that when iGASD is diagnosed in younger children or children with pneumonia, clinicians should contemplate a more aggressive approach to management. Elevation of CPR at admission was another independent risk factor for admission to the PICU, which was also consistent with previous studies.13,20,32 Despite the considerable severity of the disease, it does not seem as if the clinical presentation or the outcomes of iGASD have changed in our region in recent years.

Last of all, in our series we found a very low rate of macrolide and clindamycin resistance in GABHS compared to other studies.6,17 Although antibiotic use in Spain may be high,37 considerable efforts are being made to develop protocols, strategic plans and training programmes for the rational use of antimicrobials.38,39 In this regard, in our study we found that children that required surgery frequently received a narrow-spectrum oral antibiotic compared to children that did not require surgery. We did not find a clear reason for this difference, although it may suggest a different approach to the use of antibiotics.

Some of the limitations of our study are its retrospective design, the small sample size and that it was conducted in a single centre, which limits its statistical power and hinders the generalisation of its result. Furthermore, tests for serotyping or determining the virulence of GABHS strains were not performed, so we could not know whether these factors changed over the period under study or could have influenced the severity of iGASD. Nevertheless, our study is still the largest study on iGASD in children conducted in Spain to date along with the one published by Arias-Constantí et al.,20 and it provides valuable data on the current situation of this disease in Spain, which can help design future studies to improve our knowledge of iGASD. Since the importance of this disease is well known, in 2015 a network was created to investigate iGASD in the paediatric population in the Autonomous Community of Madrid.

In conclusion, our study on iGASD in children found an increase in incidence in the catchment area of our hospital in Madrid. This disease continues to have the potential to be very severe, with a high percentage of affected patients admitted to the PICU and requiring surgery. Given the heterogeneity of the studies published to date, it would be interesting to establish a prospective, multicentre nationwide register of iGASD, to accurately evaluate its epidemiology, risk factors and severity, and to establish the characteristics of the strains isolated in affected children.

The authors have no conflicts of interest to declare.