Rheumatic fever trends in the context of skin infection and Group A Streptococcal sore throat programmes in the Bay of Plenty: an observational study, 2000–2022

Although acute rheumatic fever (ARF) triggered by Group A Streptococcal (GAS) infections affected New Zealanders of all ethnicities in 1920–1950, Indigenous Māori had the highest rates.¹ A century later, NZ Europeans/Pākehā rarely get rheumatic fever, which now affects mainly Māori and Pacific peoples. Rheumatic fever rates increased from the 1980s to 2009^2,3 and increased for young adults until 2018.⁴ Nationally, Aotearoa New Zealand’s Rheumatic Fever Prevention Programme (RFPP) 2012–2017 significantly lowered ARF admission rates, more so at school-age than for young adults, particularly in Counties Manukau^5,6 and within the Bay of Plenty (BOP).⁷ The programme reviewers hypothesised five possible mechanisms for this reduction: treatment of GAS pharyngitis within school-based programmes and primary care; skin infection treatment; declining GAS prevalence/transmission;^6,7 and these mechanisms in combination.⁵

While sore throat management should remain a key ARF prevention strategy, the national RFPP review recommended a “new focus” to address skin infections.⁵ The organisms most commonly cultured from skin infections are GAS and/or Staphylococcus aureus. Aotearoa New Zealand case-control, cohort and community ARF studies identified both GAS pharyngitis and skin infections,⁸ including scabies^8–10 via secondary GAS infection, as specific modifiable risk factors. Self-reported scabies preceded 6% of ARF cases.⁸ Auckland population-level evidence supported the hypothesis that GAS skin infection triggers ARF. A review of 267 ARF cases found GAS-positive skin swabs preceded ARF in 9% of cases, pharyngeal GAS in 88% of cases and both in 3% of cases.¹¹

Other risk factors for GAS skin infections include doctor-diagnosed eczema.¹² In Auckland school-age children, 52% with a GAS-positive skin infection had eczema, as did 32% with GAS pharyngitis.¹² Recent immunology evidence found that multiple more GAS infections precede ARF, on average 11 episodes, priming the immune system (compared to healthy children) probably from as early as 2 years of age, challenging current intervention strategies.^13,14

GAS pharyngitis, skin infection and ARF case-control studies also recommend adequate housing, minimising crowding and reducing barriers to primary care access.^8,12 Nationally, primary care became free for most children under 6 years of age from 2007 but was not extended to those under 15 years of age until 2015.

The BOP District Health Board has funded twice-weekly school-based sore throat swabbing ARF prevention programmes delivered by Māori health providers, Hauora, since 2011; there are four programmes in the rural East and one in the West in Tauranga, with general practice support,⁷ following Heart Foundation NZ guidelines (Appendix 1). The East, which has high socio-economic deprivation, includes Whakatāne, Kawerau and Ōpōtiki districts, and the West, with heterogenous and more moderate deprivation, includes Tauranga and Western Bay of Plenty districts.⁷ From 2018, once-weekly symptomatic swabbing began in five of 18 Whakatāne schools, where ARF had doubled without RFPP. Of note, half of the East’s school-age Māori attend Whakatāne schools.⁷

Community pharmacies in Eastern BOP and Te Puke, Western Bay of Plenty, have dispensed antibiotics for GAS pharyngitis since 2015 on standing orders for those aged 3–19 years at risk of ARF, and for their skin infections since 2020.¹⁵

Concurrently, decades of increasing childhood skin infection hospitalisations prompted sentinel Aotearoa New Zealand publications in 2010.^16–18 Regional skin infection programmes followed. Health promotion, findings and protocols were shared nationally.¹⁹

Five BOP skin infection initiatives commenced in 2010–2012 and continue. Three involved public health, primary and secondary healthcare. Firstly, Toi Te Ora Public Health (TTO) undertook a health needs assessment of hospitalised skin infections for children aged 0–14 years in 2010.²⁰ Its strategies addressed health promotion, advocacy for affordable primary health access, training—including paediatric nurse seminars for practice and preschool nurses—and research.²⁰ Secondly, public health nurses, Hauora and paediatricians developed local skin infection resources from 2011 and intensified their skin infection service delivery. High schools appointed nurses trained to use standing orders to treat students’ skin infections and sore throats. In 2011, the Infectious Diseases service led a third initiative, a primary–secondary care collaboration that developed a new guideline website, “Bay Navigator”.²¹ Its pathways guiding the treatment of childhood and adult skin infections were analogous to its successor, the Midland Region Community HealthPathways.²¹

Two other initiatives were targeted: Whakatāne Hospital’s paediatric home visiting team of a nurse and social worker²² who, from 2012, supported whānau to manage eczema and prevent recurrent skin infections and started paediatric skin clinics. In Kawerau, one school-based sore throat team initiated the Kiri Ora (Healthy Skin) programme in 2013. Pharyngeal GAS point prevalence, which declined 23–11% with their sore throat programme in 2010–2013, declined further to 8% in 2014.⁷

TTO’s strategic goals included reducing ARF and respiratory and skin infections, and they reported declining childhood skin infection admission rates in 2000–2016.²³ Nationally, the socio-economic disparities of those admissions were closing.²⁴ However, while BOP’s childhood skin infection rates in 2013–2017 declined below the national average,²⁵ Auckland’s three districts’ rates (where ARF rates were increasing) remained above the national average.²⁶

This study in BOP aimed to observe if both skin infection admissions rates and ARF rates (admissions and Ministry of Health [MOH] notifications) declined contemporaneously for preschool, school-age and young adults since 2011, when health initiatives and programmes addressing GAS sore throats and skin infections commenced; this is illustrated and detailed in Appendix 1. The study sought to monitor equity, evaluate possible programme effectiveness and identify evidence of areas for quality improvement, where available.

Methods

Skin infections

Admissions of children and young adults aged under 30 years with serious skin infections, fulfilling O’Sullivan and Baker’s case definition 2010,¹⁷ and used in both previous TTO reports^20,23 (Appendix Figure 2 with 53 ICD-10 discharge codes), were identified by NHI from BOP’s two hospitals, Whakatāne, with its Eastern referral catchment, and Tauranga for the West. De-identified aggregated patient demographic data for these NHI and admissions were extracted, which informed rate sub-analyses by age, gender, hospital, year and MOH single prioritised self-identified ethnicities aggregated into Māori, Pacific peoples, NZ Europeans and Others (MELAAA: peoples of Middle East, Latin American, African and Asian ethnicities).²⁷ A total of 17,615 admissions were identified, and 19 records (0.1%) were discarded with “ethnicity not stated”. Their family names, on digital note scrutiny, suggested similar ethnic distribution to BOP’s population. Although of another 31 people, 0.2% for whom ethnicity response was “unidentifiable”, four had Māori and one had Tongan names, hospital medical information recoded them as Other European. Another coding error was re-audited where Eskimo-USA was assigned to 1.1%, which affected 198 admissions from 2000–2007, of whom 103 were identified and recoded as Māori, 90 as NZ European, 3 as Asian and 2 as Pacific peoples. Previously filed self-identified ethnicity data completed by whānau is now destroyed after clerical coding on admission. The direction of error to underestimate Māori morbidity is consistent with prior studies.²⁸

Case numbers informed rates over the 23 years of this study, 11 years from 2000 to 2010 prior to RFPP and skin infection programmes and 12 years 2011 to 2022, inclusive, since their inception. However, the years 2011 to 2019 informed our BOP post-intervention findings prior to the COVID-19 pandemic from 2020 to 2022 when, nationally, admissions for many infectious diseases, including ARF, declined.^29,30 Rate ratios estimated time trends when rates, with 95% confidence intervals (CI), were compared between time periods. Risk ratios within time periods informed equity of rates between ethnicities, genders and deprivation and, when compared between time periods, informed equity trends. Further annual rates/10,000 person-years were estimated for five 3-year and two 4-year periods: 2000–2002, 2003–2006, 2007–2010 pre-intervention and 2011–2013, 2014–2016, 2017–2019, 2020–2022, following. To evaluate some programme outcomes, prior annual rates from 2007 to 2010 vs post-intervention comparisons from 2017 to 2019 were made. Sub-group age bands were studied, because preschoolers 0–4 years and school-age children aged 5–14 years both receive childhood-specific services, including public health nursing plus primary care, while young adults aged 15–29 years receive mainly primary care alone; hence, age band comparisons, to some extent, compare service model outcomes.⁵ Preschool rates may be relevant to address GAS immune priming.

Acute rheumatic fever

First presentation ARF case numbers utilised discharges ICD codes I00.0–I02.9 and MOH notifications with cases confirmed after case note scrutiny, as previously reported.^7,31 They informed estimated ARF rates/100,000/person-years for 2000–2010, 2011–2022 and 2011–2019 inclusive and trends. The two sources were utilised together for all time periods (and alone, estimating ascertainment bias risk), utilising the same age bands as skin infections. For ease of comparison with previous publications, this study’s ARF findings are reported in several formats: all ethnicities at all ages (MOH, Te Whatu Ora – Health New Zealand); under-30-years-age, wherein 93% of cases in Aotearoa New Zealand occurred between 2000 and 2018;⁴ and Māori specific in two age bands, school-age and young adults, to permit comparison of interventions and outcomes.^5,7,32 This comparison is because many at school age, mainly Māori, are served by BOP school-based prevention programmes with primary care support,⁷ while most young adults are served by primary care alone, as also noted within the national review of RFPP.⁵ Equity analyses were as for skin infection admissions. A possible spike in ARF cases in early-COVID January–June 2020 was evaluated, as were contextual factors, comparing the odds ratio of ARF with the surrounding 6 years, and ARF rates 2011–2019 vs 2020–2022 inclusive.

Statistics New Zealand and BOP population data informed denominators, using whole district, which included Whakatāne, Kawerau and Ōpōtiki territorial land authorities (TLAs) for Eastern BOP (East) and Tauranga and Western TLA for Western BOP (West) age, ethnic-specific and gender numbers. Census data from 2006 informed 2000–2010, 2013 informed 2011–2016 and 2018 informed 2017–2022.³³ While NZ Census 2018 limitations meant there was only a 70% completion rate by Māori, it informed population growth/shifts since 2013. Population-weighted socio-economic deprivation decile scores were calculated, deriving a New Zealand Index of Deprivation (NZDep) 2006 average for the East of 9 and West 6,³⁴ which changed minimally when re-estimated using 2013 data. Rate and risk ratios with 95% CI were estimated using StataCorp Texas Statistical Software: Release 14.2, 2015. Results were considered statistically significant where CIs of rates did not overlap, rate and risk ratios did not include 1, which were tabled without reliance on P-values, which in those circumstances were <0.05 but not always tabled.

The attributable proportion of ARF cases from 2015 to 2022 (since both pharyngeal and skin GAS triggers have been actively considered) with a recent preceding GAS pharyngeal or skin infection was estimated by comparing those with a positive culture taken from either or both sites. The study also assessed whether admitted skin infections, while fewer than community infections, reflected total infections and if both changed in parallel (rather than shifting one management setting to another). As an example with accessible data, scabies rates were compared over time in three settings, in the under-30-years-age group, for admissions (ICD Code B86 primary or secondary diagnoses), emergency department (ED) presentations (Read code AD30 or scabies mentioned) and community permethrin 5% total prescriptions (unique in that it is used exclusively for scabies) in 2014–2019 available on MOH’s Datapharm website. Community scabies prescriptions for the under-30-years-age group were estimated by applying their proportion (74% of all ED scabies presentations 2000–2019) to total prescriptions.

Data availability

Where derived rates appear alone in the article, case numbers and exposed populations’ person-years are published for replicability as full tables in the Appendices.

Ethics

This study adheres to World Health Organization Ethics Guidelines, NZ Health and Disability Ethics Committees’ National Ethical Standards for Health and Disability Research and Quality Improvement 2019 and Standard Operating Procedure 2019 and its exemptions, as an observational study with minimal risk. Locality approvals were given by the BOP District Health Board for the initial ARF study of 2000–2018 and its successor Hauora a Toi for the extension of the ARF study to 2019–2022, inclusive, and skin infection data from 2000–2022, alongside Māori Research Panel review. The locality applications detailed the retrospective nature of the study of the impact of prevalent healthcare interventions, from their initiation within the district using routine de-identified data for skin infections and ARF, and NHI-linked data where use was limited to ARF case-certainty and prior skin and pharyngeal GAS infections and as having low privacy impact.

Māori Health consultation/support was sought alongside both locality agreements. This study adheres to the Health Research Council of New Zealand Guidelines for Researchers on Health Research Involving Māori 2010 and is guided by Tangata Whenua Determinants of Health 2011, Health Equity Tool 2008 and Te Toi Ahorangi 2019. Kawerau GAS prevalence studies obtained parental swabbing-consent and ethics approval NZ/1/77C3019.32. Reporting follows STROBE guidelines.

Results

Skin infections

Of all the skin infection admissions in 2000–2022, 29% occurred in preschoolers (on average 1,037/year of age), 27% over 10 school-age years, (475/school-age year) and 44% in young adults over 15 years (512/young adult year).

The proportion of admissions for Māori was 58%, NZ Europeans/Pākehā 34%, Pacific peoples 4% and Others 3% (of which 52% were Indian). Māori make up 30% of BOP residents under 30 years of age, NZ Europeans/Pākehā 60%, Pacific peoples 3.7% and Others (MELAAA peoples) 6.3%.

Skin infection admission rates for under-30-year-olds increased by 8% comparing the mean rate pre-intervention 2000–2010 with 2011–2019 (Table 1), which excludes the potentially confounding 2020–2022 COVID-19 pandemic period. (The skin infection rates in 2011–2022 were lower than 2011–2019; available from authors). For sub-groups, the changes were significant: NZ European rates were the lowest and declined significantly at all ages except young adults; Māori preschoolers’ rates were the highest, six times that of NZ European preschoolers’, and both declined, Māori by 6% and NZ Europeans by 14%. However, the inequity gaps and Māori vs NZ European risk ratios were unchanged. School-age Māori rates appeared unchanged, while NZ European rates decreased by 10%. Young adult rates were lower than preschoolers, but higher than those at school-age. Young adult Māori experienced 7% increased rates.

Pacific peoples’ skin infections were 4% of the total. Pacific peoples’ admission rates increased 34%. Their highest rates, for preschoolers, approached Māori rates and increased substantially for young adults, with significantly increased inequity risk ratios compared to NZ Europeans (Table 1).

Skin infection admissions rates for “Others” ethnicities increased significantly by 27% to rates similar to NZ Europeans. (The case numbers and denominator person-years exposed are in Appendix Table 3).

Annual skin infection admissions rose from 2000, peaked in 2010–2011, then declined to baseline numbers in 2022, a pattern most apparent for preschoolers (Figure 1).

View Figure 1–2, Table 1–3.

Therefore, analysis was undertaken of annual rates in seven time periods. Pre-intervention rates in 2007–2010 were then compared with 2017–2019 post-intervention, the most recent pre-COVID 3-year period (Table 2), rather than 2020–2022, as the decline in admissions observed for Aotearoa New Zealand during the COVID-19 pandemic also occurred in BOP and confounded observations of programme outcomes since implementation. Skin infection admission rates declined significantly from 2007 to 2010 to 2017 to 2019 by 39% for NZ European preschoolers, 31% at school-age and 32% for young adults. Rates declined 40% for Māori preschoolers, 14% at school-age and 24% for young adults. (Appendix Table 4 presents cases/exposed population person-years, and peak rate/2017–2019 comparisons).

Comparing early COVID-19 pandemic rates in 2020–2022 with 2017–2019 for BOP Māori, skin infection admissions did decline significantly at preschool and school-age but increased slightly for young adults. Rates for NZ Europeans were similar in both time periods in all age bands (Table 2).

Māori:NZ European skin infection admission risk ratios compared pre-intervention in 2007–2010 and 2017–2019 remained almost six-fold higher for Māori preschoolers, which increased to three-fold at school-age and remained two-fold for young adults.

Māori young adults’ 2007–2010 skin infection admission rates were significantly higher than school-age, risk ratio 1.48 (95% CI 1.32–1.66) and 2017–2019, risk ratio 1.30 (95% CI 1.14–1.49). School-age Māori rates in 2017–2019 declined significantly from baseline 2000–2002 rate ratio 0.82 (95% CI 0.71–0.95), whereas young adult Māori rates in 2017–2019 were not dissimilar to baseline rates rate ratio 1.04 (0.91–1.19).

For BOP Māori under-30-years-of-age, skin infection rates were unchanged in the East and declined 43% in the West from similar peak rates. An East:West high deprivation:moderate deprivation inequity ratio of 1.4 emerged for all age bands. Comparing Māori skin infection admission rates in 2007–2010 to pre-COVID 2017–2019 rates, significantly larger declines were observed in the West at all ages, with a smaller 31% decrease for Eastern preschoolers. The only statistically significant male:female skin infection rate disparity for Māori was the emergence of a risk-ratio of 1.4 in the Eastern BOP for young adults (Appendix Table 5).

Annual scabies admission rates/10,000 increased significantly for under-30-year-olds from the baseline years 2000–2010 (rate 1.80 [CI 1.51–2.09]) to the peak year 2011 with the years 2011–2015 (rate 3.57 [CI 2.97–4.17]), then declined in 2016–2019 to 1.37 (CI 0.99–1.75). Similarly, scabies ED presentations/10,000 for 2000–2010 was 3.21 (CI 2.82–3.59), rose in 2011–2015 to 8.80 (CI 7.87–9.74), then declined in 2016–2019 to 2.60 (CI 2.07–3.13). The estimated community permethrin 5% prescription rate in 2011–2015 of 220.06 (CI 215.89–225.29) also declined in 2016–2019 to 127.19 (123.49–130.89). Māori and Pacific 0–29-year-olds experienced 93% of under-30-years-age scabies admissions and made up 81% of ED under-30-years-age scabies presentations.

Diagnosed scabies preceded ARF within the 3 months before ARF diagnosis¹¹ for 0.7% (1/142) Māori and Pacific 5–14-year-olds presenting with ARF in 2000–2019, whereas 2.8% (4/142) had a documented scabies admission, ED presentation or community anti-scabies prescription in the 10 years prior to ARF diagnosis.

Acute rheumatic fever

Of 49 ARF cases in BOP in 2015–2022, 35 had recent preceding positive community or hospital GAS swabs. Of positive swabs, pharyngeal GAS preceded 69% of cases, skin GAS 17% and both pharyngeal and skin GAS 14%. That is, pharyngeal GAS may have preceded ARF in 69–83% of cases and skin GAS may have preceded ARF in 17–31% of cases.

Although the plot of 194 first presentation ARF cases in BOP by year in 2000–2022 under-30-years-age shows similar numbers per year in 2020–2022 to the immediately preceding 6 years (Figure 2), ARF rates were estimated in 2011–2019 vs 2000–2010 for consistency with skin infection admission analysis, lest the COVID-19 2020–2022 findings confound estimated programme outcomes.

From 2000 to 2019 ARF cases in BOP at all ages, 13 (7%) were recurrences,⁷ leaving 181 ARF cases at all ages, of whom 162 (90%) identified as Māori, 11 (6%) Pacific peoples and 8 (4%) NZ European. Nine cases (5%) were aged 30–45 years. First ARF presentations at all ages for all ethnicities (the Te Whatu Ora – Health New Zealand, MOH website reporting measure) declined significantly by 26% using discharges plus notifications, then case note scrutiny, and appeared unchanged utilising notifications alone (Appendix Table 6).

For those under-30-years-age, the focus of this article and a recent study of New Zealand’s ARF-RHD trends until 2018,⁴ BOP ARF rates declined significantly for all ethnicities by 29% and declined for Māori by 31% when comparing 2011–2019 with 2000–2010 (Table 3). The ARF rate for Māori under-30-years-age, compared with NZ Europeans, for 2000–2010, informed a risk ratio of 24.39 (11.29–54.40), which increased as low NZ European rates declined by 85%, faster than Māori rates declined. Most NZ European ARF cases occurred in 2000–2003, except the last one in 2011.

Pacific peoples reside mainly in the West. Pacific ARF case numbers rose and NZ European cases declined markedly for the same period. With small numbers and wide CIs, neither change reached statistical significance. Pacific peoples’ rates were between NZ European and Māori rates, with significant Pacific peoples:NZ European risk ratios (Appendix Table 7; ARF case numbers, exposed person-years).

One preschool Māori tamariki experienced ARF. Significant declines in ARF rates occurred for Māori school-age children (36%) but were unchanged for young adults. Male:female two-fold higher risk ratios continue (Appendix Table 8). The Māori school-age ARF rate compared with Māori young adults for 2000–2010 was risk ratio 10.07 (4.88–20.76) and in 2011–2019 was risk ratio 4.74 (2.40–9.37).

More than two-thirds (69% [111/162]) of all-age Māori who experienced ARF in BOP in 2000–2019 resided in the more socio-economically challenged East, which has a mean NZDep decile of 9 and is where 44% of Māori live. Conversely, one-third of Māori acquiring ARF reside in the more heterogenous West, which has a mean NZDep decile of 6. Of those Māori who experience ARF while living in the West, a third live in Census areas with NZDep deciles of 2–6, belying the stereotype that ARF is always a disease of poverty.³⁵ Higher school-age ARF rates in the East declined by 23%, but not significantly. Larger significant 49% declines from lower rates occurred in the West at school-age, but not for young adults. East:West inequities with two three-fold higher rates continued (Appendix Table 9).

During COVID-19’s first wave, eight cases of ARF affected BOP school-age Māori and Pacific peoples in the first 6 months of 2020, with an odds ratio of 3.8 (CI 1.53–7.48), compared with the remainder of July 2015 to July 2021. (Presented by lead author to The Paediatric Society of NZ, Virtual ASM, November 2021). From all under-30-years-age in BOP, 66 ARF cases presented in 728,721 person-years in 2011–2019 and 22/268,043 person-years in 2020–2022, deriving estimated rates/100,000/year of 9.06 and 8.21 respectively, with rate ratio 0.91 (CI 0.56–1.47) comparing the intervention period of 2011–2019 and the COVID-19 period. As ARF numbers reduced then stabilised from 2014 (Figure 2), the rate for the period 2014–2019, inclusive 38/498,393, was also compared with the 2020–2022 COVID-19 period, deriving an annual rate of 7.62 vs 8.21/100,000 with a rate ratio of 1.08 (0.64–1.81).

Discussion

There is mounting evidence that skin infections contribute to ARF in Aotearoa New Zealand,^8,11 supported by our finding where cultures were taken of pharyngeal GAS associated with possibly two-thirds of ARF presentations, skin GAS with one-sixth and both one-third. Immune studies are now able to confirm or refute whether such cultured GAS triggered ARF,^13,14 while concluding that interventions to reduce cumulative childhood GAS exposures could prevent ARF.¹³

This study found that skin infection admission rates rose, then both skin and ARF rates declined contemporaneously and significantly within BOP from 2011 after specific programmes for both commenced. The decline in skin infection rates was seen for both Māori and NZ Europeans at preschool, school-age and as young adults, while the decline in ARF rates for Māori was seen at school-age but not for young adults. NZ Europeans’ ARF rates declined at school-age, with the last NZ European ARF first presentation in 2011.

We found, for one sentinel skin infection of scabies, that admission and community rates declined in parallel. However, scabies seldom (0.7%) preceded ARF in BOP within the critical preceding 3 months wherein GAS triggers ARF.¹¹ Skin infection programmes began²⁰ at the time of their peak rates in 2010–2011, and RFPP GAS pharyngitis school-based programmes with primary care support began simultaneously.⁷ As well as general and targeted direct programmes effects, both skin infection and GAS pharyngitis programmes lower GAS prevalence and thereby transmission, as well as immune priming, which may lower ARF rates.^5–7,36,37 The wider context included free children’s primary healthcare and healthy homes initiatives.

Māori make up 30% of the under-30-years-age population in BOP but experience 58% of the skin infection admissions and 90% of the ARF. Eight of every 10 people presenting with ARF in the BOP are Māori at school-age. By 2017–2019, school-age Māori skin infection admission rates declined 14% from pre-intervention in 2007–2010, down 24% from their peak and 18% from baseline years, and ARF declined by 36%. Young adult Māori are 1 in 10 of those who experience ARF, a group without targeted programmes, served by primary care alone. Their skin infection admission rates were higher than school-age Māori rates, declined 26% from 2007–2010, 33% from peak rates and increased 4% from baseline, and their ARF rates were unchanged. (Their COVID-19 2020–2022 skin infection admission rate rose 12% from 2017–2019).

NZ Europeans, who make up 60% of the under-30-years-age population in BOP, experience 34% of the skin infection admissions and 4% of the ARF. School-age NZ Europeans experienced the largest pre- post–skin infection rate (32%) and ARF rate declines (85%), such that inequities increased. While both Māori and NZ European preschoolers had equal, almost 40% skin infection admissions rate reductions, Māori:NZ Europeans risk ratio, a measure of inequity, was unchanged, with almost six-fold higher rates.

Pacific peoples’ highest skin infection rate was for preschoolers, rising to a rate similar to that experienced by Māori. Their skin infection rates increased at all ages except school-age. Pacific peoples experience 6% of the ARF in BOP, with unchanged ARF rates at all ages.

Others (MELAAA peoples) had increased skin infection rates, to rates similar to NZ Europeans’, and experienced no diagnosed ARF.

This study’s strengths include bringing skin infection and ARF rates alongside in identical age bands by ethnicity, gender and two districts with medium and high deprivation over 23 years from one region, with sufficient power to permit temporal observations for age-groups with and without targeted interventions.^7,38,39 The risk of confounding has been reduced by excluding 2020–2022 from all prior and post-intervention rate comparisons in tables when considering programme outcomes. However, the COVID-19 period has been retained for ARF time trend analysis only in Appendix 10 in light of the finding that ARF did not decline in BOP during that period, although overall ARF declined in Aotearoa New Zealand. Future studies may test our hypothesis, which we evaluated only for scabies, that trends in admitted skin infections may reflect community infections too, by accessing primary care and secondary care diagnostic data. Future ARF studies may estimate more refined proportionate attributable risks from skin and throat GAS infections and guide programme weighting.

This study’s multi-sourced scrutinised ARF data methodology is robust and validated,^6,7,31,35,40 in contrast to single-sourced notifications (utilised in Aotearoa New Zealand’s RFPP effectiveness evaluation),⁵ which in BOP, with ascertainment bias, suggested unchanged ARF rates.

Skin infection admission rates declined 40% for both Māori and NZ European preschoolers exposed to public health nursing, primary care and some targeted paediatric interventions, possibly lowering recurrent GAS infections’ induction of immune priming and risk for ARF.^13,14 That skin infection admission rate reductions for Māori were more marked in the West than East, and that significant Māori:NZ European inequities continue, has strategic implications. All Hauora-delivering school-based ARF programmes might also be contracted for focussed Kiri Ora (Healthy Skin) programmes, informed by current evidence, and workplaces might be considered too. Hauora complement and collaborate with targeted pharmacies’ services, predominantly in the East, but also in Te Puke in the West, for at-risk children and young adults.¹⁵ Hauora programmes with primary care support have reported positive outcomes, including for skin infections,^41,42 and have addressed GAS pharyngitis and ARF equity gaps for Māori by encouraging presentations.⁷

The prior BOP retrospective cohort study reported substantial (60%) school-age ARF rate reductions within the rural East,⁷ where twice-weekly sore throat swabbing occurs and where the inequity gap of two-fold higher rates for Māori males closed a little, but not significantly, compared with females.⁷ The current finding that the BOP-wide ARF rate reductions were greater in the lower deprivation West than East suggests that the East’s contracted service delivery model should be reviewed. A change from once to twice weekly targeted GAS sore throat swabbing in at least five Whakatāne schools is indicated, a district with moderate–high deprivation, where half the East’s school-age Māori live, while twice-weekly swabbing continues in the adjacent highest deprivation areas. In the West, the substantial 49% ARF decline for school-age Māori from lower baseline rates suggests that there the mixed model of mainly primary care and school-based programmes in 3/58 schools remains responsive to Māori health needs.

Behind the higher ARF rates for Māori, especially in the Eastern Bay, lie their documented housing, education and employment inequities and challenges.⁴² The findings provide further evidence for focussed needs-based health and housing interventions to build on community strengths and cultural capital and enable Hauora and providers engaged with their communities, especially those with greatest socio-economic deprivation.

BOP’s COVID-19 spike in 2020 highlighted the challenges for public health to enable other health delivery to continue during epidemics, including both GAS and COVID-19 swabbing in vulnerable communities. Initially, drive-by stations swabbed solely for COVID-19. Meanwhile, school-based GAS programmes closed, primary care became virtual, paediatric hospital presentations declined by 30% and functional household crowding increased. In mid-2020, swabbing for both COVID-19 and GAS at BOP COVID-19 stations was enabled. Pandemic aftermaths include post-COVID absenteeism and caution.

Future co-designed skin infection and ARF strategies are indicated with Pacific peoples, and for all young adults, as is timely re-evaluation of all altered BOP programmes.

ARF and skin infection rates declined contemporaneously across the BOP with programmes for both, especially at school-age. Ethnic, gender, age and socio-economic inequities persist, and appear amenable to targeted interventions, increasing the coverage and frequency of focussed programmes.

View Appendices.