Epidemiology of skin infections in Auckland, New Zealand

Aotearoa New Zealand has one of the highest rates of bacterial skin infections among high-income regions, with over 15,500 cases per 100,000 population documented by the Global Burden of Diseases in 2019.¹ Predominantly, these cases manifest symptomatically as pyoderma (82%), with cellulitis comprising the remainder.¹ These skin infections represent a significant health concern for New Zealand, particularly affecting children under the age of 14 years, a group with a 2019 hospitalisation rate of 3.2 per 1,000.² The risk of being hospitalised with a skin infection is notably higher among children, Māori, Pacific populations and individuals in socio-economically deprived areas.^2–4

Staphylococcus aureus (S. aureus) is the primary cause of skin infections in New Zealand,⁵ aligning with patterns observed in other countries reporting high rates of skin infections, such as India and the United States.^6,7 Streptococcal pyogenes (S. pyogenes), commonly known as Group A Streptococcus or GAS, is also an important cause of bacterial skin infections in New Zealand and worldwide.^5,8,9 In Auckland, New Zealand between 2010 and 2016, positivity rates for S. pyogenes and S. aureus in primary healthcare skin and wound swabs were reported to be 4.8 and 14.1 per 1,000 person-years, respectively.⁵ Cellulitis caused by S. pyogenes is a substantial contributor to total economic and health burdens in New Zealand, with an estimated annual economic burden of NZ$11.3 million in 2015.¹⁰

Beyond acute infection, S. pyogenes infection can instigate autoimmune disorders, including acute post-streptococcal glomerulonephritis (PSGN), a rare kidney disease,¹⁰ and potentially acute rheumatic fever (ARF).¹¹ ARF is a serious condition that can result in permanent heart damage, known as rheumatic heart disease (RHD).¹² Historically, ARF was thought to follow GAS pharyngitis,¹² but recent evidence indicates that New Zealand Māori and Pacific peoples face a fivefold increased risk of ARF within 8–90 days after a skin swab positive for S. pyogenes compared with those with negative swabs.¹¹ ARF/RHD are preventable and rare in high-income settings but prevalent among New Zealand Māori and Pacific peoples, who also face a disproportionate burden of S. pyogenes skin infections.^5,12,13

The study is an extension on a previous study looking at S. aureus and S. pyogenes in skin swabs from 2010 to 2016.⁵ This study includes data up to 2020 and aims to investigate how these bacteria impact different demographic groups both in isolation and as co-infections, as well as to investigate the potential influence of S. aureus on the risk of S. pyogenes infection. Previous research has indicated a high prevalence of S. pyogenes in skin swabs from children under 10 years old, along with evidence suggesting that such early exposure contributes to the onset of ARF.⁵ Therefore, we posited that a particular demographic, specifically children under 10 years, may face an increased risk of skin infections attributed to S. pyogenes.

Method

Study data were obtained from Awanui Labs (formally Labtests), which has been Auckland’s only accredited community pathology laboratory service provider since mid-2009.¹⁴ The Auckland Region was defined as Waitematā, Auckland and Counties Manukau catchment areas (at the time of the data collection, district health boards [DHBs]). Data included all Auckland residents who visited primary care for a wound or suspected skin infection that resulted in a wound/skin swab being sent to Awanui Labs for microbiological culture between 1 January 2010 and 31 December 2020. Samples for these swabs were collected during primary care consultations from patients presenting with symptoms characteristic of wound or skin infections that required microbiological culture for further assessment. The dataset was restricted to symptomatic cases, thereby omitting any asymptomatic individuals. Serving approximately 1.5 million people, or one-third of New Zealand’s population,¹⁵ Awanui Labs primarily receives specimens from primary care, with over 95% of these coming from such facilities. Cultures were grown on tryptic soy agar with 5% sheep blood and incubated for 48 hours at 37 degrees Celsius in 5% CO₂. Until 2012, identification relied on streptococcal grouping latex and coagulase tests, respectively; thereafter, identification utilised MALDI-TOF MS Biotyper (Bruker, Germany).

Awanui Labs’ microbiological testing data were provided with encrypted National Health Index (NHI) numbers supplied by the New Zealand Ministry of Health, enabling the pairing with de-identified demographic data. Auckland’s demographics and population size were estimated using the 2013 New Zealand Census. The 2013 Census data provided the ethnic composition of Auckland: 11% Māori, 13% Pacific peoples, 22% Asian and 53% NZ European/Other.¹⁵ Ethnicity data used a prioritised approach, with prioritisation given to Māori, followed by Pacific peoples, then Asian and finally NZ European/Other.¹⁶ Those with multiple ethnic identifications were prioritised to the first ethnic grouping in that schema, e.g., individuals identifying with both Māori and Pacific ethnicities were classified as Māori. The socio-economic status of residential areas was indicated by the New Zealand Index of Deprivation (NZDep) 2013.¹⁷ NZDep is a granular, area-specific metric of socio-economic deprivation that categorises New Zealand’s geographical areas into 10 deciles or five quintiles, ranging from regions of least socio-economic deprivation (deciles 1–2 or quintile 1) to regions marked by highest deprivation (deciles 9–10 or quintile 5).

Data analysis

Data analysis was conducted with Stata/SE 17.0. Data missing essential information like encrypted NHI numbers were excluded. Analysis included stratified analysis by age, ethnicity, socio-economic status (using NZDep2013), gender, region and season. Age and season were determined by birth and test result dates, respectively. The methodology for defining cases of skin infections from skin swab data involved grouping swabs taken for the same person within a 3-month period as a single event, with any skin swabs taken after this period considered a new infection event. The 3-month cut-off period was determined based on the distribution of swabs collected over the 11-year period to differentiate between distinct episodes of skin infections (Appendix Figure 1). Cases were defined as infections caused by either S. aureus, S. pyogenes, co-infections of both or negative results for both pathogens.

The frequency of disease measurements was based on the fraction of Auckland’s population (N=1,431,189) who sought primary care and underwent swab tests (N=360,861) for skin infections over an 11-year period. We calculated presentation and infection rates per 1,000 person-years, including rate ratios (RR) with 95% confidence intervals (95% CI) to compare risks across demographic groups. Poisson regression was applied, adjusting for age, ethnicity, socio-economic status, gender and geographic location (the three constituent DHBs within the greater Auckland Region).

To estimate the risk of S. pyogenes following exposure to S. aureus, infection events initially swabbed and re-swabbed within a 3-month period were stratified into two groups: one group tested positive for S. aureus only during the initial swab collection, and the other group tested negative for S. aureus (either positive for another skin pathogen or no significant pathogen detected) at the time of the initial swab. The S. pyogenes positivity rate was then examined for both groups during the subsequent re-swabbing within the 3-month follow-up period.

Results

Overview

Between 2010 and 2020, 360,861 people with unique NHI numbers underwent clinical evaluations for suspected skin infections at primary care facilities, resulting in skin swabs being sent for bacterial culture. Annually, an average of 44,903 individuals had a swab taken, a rate of 31.4 individuals swabbed per 1,000 person-years. Approximately 22.6% of the swabs were collected within 3 months of an initial swab. The swabbing rate decreased steadily after the initial 3 months. Hence, all swabs collected within 3 months of the initial swab were categorised together to identify a single infection incident (Appendix Figure 1).

A total of 514,280 skin infection cases were identified, with 87.9% requiring a single visit and 12.1% necessitating multiple visits within 3 months, which resulted in additional wound swab/s being collected for microbiological assessment. Among these patients, 26.0% (n=93,957) experienced re-infection (excluding positive swabs in the 3 months after their previous swab) a median of 508 days later (interquartile range [IQR] 231–1,097), resulting in 153,419 recorded re-infections. Laboratory analyses revealed that 37.8% of cases were culture negative and 5.1% were positive for other pathogens, while 43.5% were positive for S. aureus, 4.9% were positive for S. pyogenes and 8.8% were positive for both (Table 1).

View Table 1–2, Figure 1–4.

Figure 1 shows an increase over the 11-year period in the rate of presentations/visits at primary care for suspected skin infections, with a decline noted in 2020, corresponding to the COVID-19 pandemic and response.

Age

The population who presented for evaluation of skin infections had a median age of 36.2 years (IQR 13.6–61.7). At a population level across the study period, 38.6 primary care visits necessitated microbiology culture testing for skin infections per 1,000 person-years. The incidence rate was higher among children under the age of 10 years (45.5 per 1,000 years), in contrast to 29.1 per 1,000 years for those aged 10 years and above. Over the 11-year period, 37.8% of individuals under the age of 10 years sought primary care for microbiology testing for skin infections, in contrast to 23.1% of individuals outside this age bracket. Individuals under the age of 10 years comprised the majority of cases (42.4%) of all infections related to S. pyogenes, and 18.4% of all infections solely attributed to S. aureus. Infections caused by S. aureus alone were predominantly (55.9%) found in adults aged 30 years and above (Table 1).

The overall rate of skin infection declined with increasing age. Although the rate of S. aureus infections remained relatively consistent across age groups, infections involving S. pyogenes, either as solitary infections or co-infections with S. aureus, peaked in individuals under the age of 10 years and demonstrated a steady decrease thereafter (Figure 2). Children under the age of 10 years exhibited a 1.3-fold higher rate of S. aureus infections (95% CI 1.3–1.3) compared with those aged 10 years and older, and a 3.1-fold (95% CI 3.1–3.2) higher rate of S. pyogenes-related infections. Analysis further established that, on average, individuals infected with S. pyogenes were 23.3 years younger (95% CI 23.0–23.7), with a median age of 12.9 years (IQR 5.7–31.0), as opposed to those solely infected with S. aureus, who had a median age of 36.2 years (IQR 14.4–63.2).

Ethnicity

Individuals who identified as Asian exhibited the lowest incidence of microbiological testing in primary care. Pacific peoples reported the highest annual rate of primary care visits for skin infection microbiology testing, at 65.1 visits per 1,000 person-years, followed by Māori (46.8 visits per 1,000 person-years) and NZ European/Other (43.1 visits per 1,000 person-years) (Appendix Table 1).

The NZ European/Other group accounted for over half (56.0%) of all identified S. aureus-exclusive infections. In contrast, the Māori and Pacific groups accounted for 72.8% of all S. pyogenes-related infections (Table 1). While infections exclusively caused by S. aureus showed similar incidence rates across all three ethnic groups, S. pyogenes infections differed substantially by ethnicity. For S. pyogenes alone, higher rates compared with NZ European/Other were seen for both Māori (RR 2.2, 95% CI 2.1–2.3) and for Pacific peoples (RR 2.7, 95% CI 2.6–2.8). Higher rates were also seen for S. pyogenes and S. aureus co-infections, again compared with NZ European/Other, for both Māori (RR 3.9, 95% CI 3.8–4.0) and Pacific peoples (RR 5.6, 95% CI 5.4–5.7) (Table 2).

Socio-economic deprivation

Skin infections increased with socio-economic deprivation (Figure 4). People in the most deprived areas (Quintile 5) had 54.8 primary care microbiological examinations per 1,000 person-years (Appendix Table 1). S. aureus had a similar incidence across all socio-economic deprivation levels (Table 2, Figure 4). However, 54.5% of infections involving S. pyogenes occurred among individuals living in the most deprived areas (Table 1). For S. pyogenes alone, higher rates were seen for individuals in the most deprived areas (RR 2.4, 95% CI 2.2–2.5, compared with the least socio-economically deprived areas). Higher rates were also seen for S. pyogenes and S. aureus co-infections for the most socio-economically deprived areas (RR 3.0, 95% CI 2.8–3.1) (Table 2, Figure 4).

Gender, geographical area and season

No significant gender differences were seen in primary care testing for skin infections (Appendix Table 1). Infection rates across these regions were comparable (Table 2). Seasonal variation was observed, with a rise in skin infections during summer and autumn, resulting in 26,338 more cases than in winter and spring (Table 1). There was a slightly higher rate of S. aureus-exclusive infections in these warmer and dry months, with a 1.1-fold (95% CI 1.1–1.1) increase, and a 1.5-fold (95% CI 1.5–1.5) increase in the rate of S. pyogenes-related infections, compared with the cooler and wet seasons (Table 2).

Role of S. aureus on risk for S. pyogenes

Over half (51.4%) of co-infections with S. aureus and S. pyogenes were seen within Pacific populations (Table 1).

A follow-up analysis examined the risk for subsequently testing positive (within 3 months) for S. pyogenes based on initial test results. Individuals initially testing positive for S. aureus were more than twice (2.1, 95% CI 1.9–2.3) as likely to subsequently test positive for S. pyogenes during a skin infection, compared with cases where the individual initially tested negative for S. aureus or was positive for a different skin pathogen (Appendix Table 2).

Discussion

Utilising data from 2010 to 2020 for Auckland, our study shows an average annual rate of 38.6 visits per 1,000 population using primary healthcare services for skin swab culture tests for suspected skin infections. This exceeds international figures; for example, Finland reported an average of 10.8 primary care visits for skin infections per 1,000 population between 2015 and 2019.¹⁸ The true burden is probably greater than what our study shows, since not everyone seeks primary care or receives skin swabbing during their visits. For example, our estimate of 46.2 visits per 1,000 children under 15 years is significantly lower than the rate found in a 2008 small-area study in Tairāwhiti, which reported 106.7 cases per 1,000 children.¹⁹ This indicates that in some areas the burden of skin infections could be much higher.

Infections caused by S. aureus were reported at a rate of 14.2 per 1,000 person-years and S. pyogenes at 4.5 per 1,000 person-years, the latter aligning closely with previous annual estimate of 4.0 cases per 1,000 population for S. pyogenes in New Zealand.¹⁰ The estimates remained consistent even when individuals with distinct NHI numbers were accounted for only once each year. Consistent with similar studies in the Auckland population, skin infections solely due to S. aureus are prevalent across all ages and ethnicities, regardless of socio-economic deprivation levels.⁵ Our study identifies S. aureus as a risk factor for S. pyogenes, with individuals testing culture-positive for S. aureus having a 2.1-fold increased risk of subsequently testing positive for S. pyogenes upon being re-swabbed at some point during a 3-month follow-up period.

S. pyogenes involvement in skin infections, either in isolation or as a co-infection with S. aureus, predominantly affects young children, Māori and Pacific communities and individuals in the most socio-economically deprived areas. Approximately 42.4% of all S. pyogenes-related infections were identified in children under 10 years of age, at a rate of 10.7 infections per 1,000 child-years—three times higher than the rate in individuals aged 10 years and older. Studies in Australian Aboriginal and Torres Strait Islander descent children also report a higher risk of skin infections from birth, gradually declining with age.^20–22 Māori and Pacific peoples in New Zealand are disproportionately affected by S. pyogenes, with a 4.7-fold increased risk compared with the NZ European/Other population. Additionally, Pacific peoples represent over half (51.4%) of all S. aureus and S. pyogenes co-infections, mirroring observations among Indigenous Australian children in remote areas of the Northern Territory.⁹

The prevalence of S. pyogenes in these demographics is particularly concerning in Australasian countries due to the associated risks for ARF.^11,23 Persistent exposure to S. pyogenes through common infections or colonisation in early childhood can potentially increase the risk of immune system dysregulation and potential development of ARF/RHD later in life.^24–26 Additional skin conditions, such as scabies and eczema, may also serve as risk factors for skin infections within at-risk communities by facilitating bacterial colonisation and onset of secondary infections. Research demonstrates that S. pyogenes skin infections are twice as prevalent in Aboriginal and Pacific children afflicted with scabies compared with their counterparts without scabies,^9,27,28 and New Zealand children affected by S. pyogenes skin infections are four times more likely to receive an eczema diagnosis compared to children who did not have S. pyogenes infection.²⁹

While the study benefits from its well-defined population base, standardised microbiological protocols and analysis by socio-demographic factors, there are limitations, such as the reliance on skin swab data reflecting primarily purulent, weeping or leaking infections, the absence of clinical data and the inability to include individuals who did not seek medical attention. The true disease burden likely surpasses the reported figures, particularly among the most vulnerable populations who are less likely to access healthcare. Additionally, the study is limited to testing, and did not ascertain whether infections were treated and, if so, the nature and impact of such treatments.

This paper highlights the steady rising trend in primary care visits for skin infections, which may be due to an increased underlying risk of infection or more people seeking healthcare services for these conditions due to awareness or changes in access (including funding for children’s visits to general practitioners). A decline in skin swab cases was noted in 2020, coinciding with the onset of the COVID-19 pandemic. We hypothesise that the pandemic may have played a role in this reduction as individuals were in isolation, potentially impacting the frequency of in-person visits to healthcare facilities. This paper ultimately stresses the persistent issue of skin infections in New Zealand, with important implications for both clinical practice and public health policies.

The findings accentuate that health disparities are prevalent among Māori and Pacific peoples, with the root causes of these disparities remaining insufficiently understood. The disproportionate prevalence of S. pyogenes skin infections in Māori and Pacific people suggests that primary care practitioners should prioritise thorough assessment and treatment of skin infections within these populations. Additionally, there is an urgent necessity for more comprehensive, effective and culturally informed public health initiatives designed to address these inequities and improve health outcomes for these communities. These initiatives should address social determinants of health, such as improving access to healthcare and healthy homes, which are strongly patterned by ethnicity.²⁹ Additionally, the establishment of clinical guidelines is essential for the effective management of these infections. Establishing a primary care surveillance network to monitor skin infections and transmission of S. pyogenes in high-risk communities could yield valuable insights into disease mechanisms and understanding risk disparities. Such a network would be instrumental in generating critical data to better understand susceptibility to S. pyogenes and related conditions, such as ARF, ultimately informing strategies to effectively lessen the observed disparities.

View Appendix.