The epidemiology of serious skin infections in New Zealand children: comparing the Tairawhiti region with national trends

Skin and subcutaneous tissue infections are a heterogeneous group of superficial bacterial infections, most commonly caused by opportunistic skin pathogens: Staphylococcus aureus and Streptococcus pyogenes.1While these infections are usually effectively treated within the primary care setting, several international studies have recognised an increase in the number of cases serious enough to require hospitalisation.2-4 This subset of more significant cases has been termed serious skin infections.In New Zealand (NZ) the increase has been particularly marked, with the rate of cellulitis double that of Australia and the United States of America.5 Between the years 1990 and 2007 the national incidence rate almost doubled,6,7 making these infections one of the most common reasons for childhood hospitalisation.8Within NZ significant inter-regional variation in the incidence of serious skin infections has been noted; these differences are hypothesised to be multi-factorial and in part reflect the distribution of population groups who are known to experience higher disease rates, notably M ori and Pacific children, children from lower socioeconomic backgrounds, and children less than 5 years old.5-10The Tairawhiti (Gisborne) region and District Health Board (DHB) is a geographically isolated area of 45 000 people on the East Coast of NZs North Island. The region is unique for its warm climate, large M ori population (47.3% of the total population and 58.0% of the 0-14 year old population), youthfulness (26.2% of people are aged less than 15 years old),11 and high level of deprivation (the region has the largest proportion of highly-deprived residents in the country).12In Tairawhiti, skin infections present a major challenge in both primary and secondary level care; recent research by the authors found that between 1990 and 2007 Tairawhiti District Health had the highest incidence of childhood serious skin infections out of all NZ DHBs.6This study aimed to describe the incidence and epidemiology of serious skin infections in children in the Tairawhiti region over the period 1990-2007, to compare these local patterns to equivalent national data, and to determine whether the infection incidence observed in the Tairawhiti region is greater than that which is expected given the high-risk population composition.Methods Case selection and data extractionHospital discharge data were obtained from the NZ Ministry of Health for all children aged 0-14 years, admitted at least overnight to a NZ public hospital between 1 January 1990 and 31 December 2007, with a principal or additional discharge diagnosis from a defined list of serious skin infection International Classification of Disease (ICD) codes (see Appendix 1 at the end of this article). Cases after July 1999 were identified using ICD-10 diagnostic codes, and cases prior to this date by ICD-9 codes which were forward and backward mapped from ICD-10. This case definition was developed in recent work which found the validity of the former definition was markedly improved by including categories of skin infections previously overlooked in epidemiological analyses. With the addition of skin infections of atypical anatomical sites, those secondary to either primary skin disease or trauma, and those recorded as additional diagnoses (see Appendix), the sensitivity of the case definition increased from 61.0% to 98.9% with little loss in specificity.13 Each discharge record included a unique patient identifier (encrypted National Health Index number) enabling transfers and readmissions within 30 days with the same principal diagnosis code to be removed. To ensure a better match with the census population, overseas visitors were excluded. Day cases were excluded from the case definition due to inconsistencies in the recording of these events between regions and over time. Patient variables including age, prioritised ethnicity, gender and home domicile code and admission variables such as the season, year, DHB, duration and outcome of admission were recorded and collated. Due to the small numbers of Pacific and other non-M ori ethnic groups in the Tairawhiti region, prioritised ethnicity used only two categories, M ori and non-M ori, with non-M ori including NZ European, Pacific, Asian and all other non-M ori ethnic groups. Assigning levels of socioeconomic deprivation used the New Zealand Deprivation Index (NZDep) and was based on the home domicile census area units (CAUs) of cases. The NZDep is based on nine variables extracted from census data;14 NZDep 1 indicates least deprivation and 10 indicates highest deprivation. In 2.21% of cases domicile codes could not be linked to CAUs due to retired codes and addresses outside of classification. To reduce the impact of these missing CAUs, retired domicile codes were linked to new codes using files from the Ministry of Health and Statistics NZ (R. Bishop, Statistics New Zealand, personal communication; CAU changes 1991-2006, Wellington, 2009; C. Lewis, New Zealand Health Information Service, personal communication; Domicile code mapping, Wellington, 2009). Data analysisThe data were analysed using Microsoft Excel 00ae and SAS 00ae. Denominators in rate calculations were derived from usually resident population counts from the 1991, 1996, 2001, and 2006 censuses. Counts from each census were used to approximate the population in the preceding and subsequent two years. Age adjustment used the World Health Organisation (WHO) standard population. Trends between populations were explored by the calculation of rate ratios (RRs) with 95% confidence intervals (95% CIs) calculated using the log-transformation method.15 Significant differences in RRs were indicated by a two-tailed p-value <0.05. Indirect standardisationThe final part of this analysis used indirect standardisation to adjust for variables in the Tairawhiti population that could affect disease rates, and hence establish whether the observed incidence (or crude incidence) of serious skin infections in the region was in line with the incidence expected after taking into account the high-risk age, ethnicity and deprivation composition of the population. Typically, direct standardisation is used to validly compare two or more groups that differ in health determinants, however this method requires a large population to ensure age, deprivation and ethnicity-specific rates remain stable. Due to the small numbers in some subgroups in the Tairawhiti population, direct standardisation could not be used. Age/ethnicity/deprivation-specific rates were calculated using interpolated usually resident population counts by CAU from the 1991, 1996, 2001, and 2006 censuses. Indirect standardisation was used to standardise each variable, both individually and in combination, across two time periods (1990-1999 and 2000-2007) with NZ in total (including Tairawhiti) used as the standard population. Expected discharge numbers for each age/ethnicity/deprivation group were calculated by multiplying the national rates for that stratum by the usually resident population for that stratum in the Tairawhiti region. Five cases with unknown deprivation scores were excluded from this analysis. The ratio of observed to expected (O:E) cases was then calculated. An O:E of 1 denoted the observed number of discharges was the same as the expected number, an O:E less than 1 indicated the observed number was less than the expected number and conversely an O:E greater than 1 indicated the observed number was greater than the expected number. Statistical significance was determined by calculating 95% confidence intervals for these ratios. Results Selection of cases, incidence and impactIn the Tairawhiti region a total of 1976 hospitalisations met the case definition. From this total, 10 (0.5%) overseas visitors, 50 (2.5%) transfers, 166 (8.4%) day cases, and 39 (2.0%) readmissions were excluded. This left 1711 (86.6%) cases of childhood serious skin infection for further analysis. Of these cases, 1 patient was reported to have been discharged dead from hospital (case fatality of 0.06%). Hospitalisation data recorded a total of 6459 hospital days over the study period. The median and mean lengths of stay were 2 and 3.8 days respectively. In New Zealand during the same period there were a total of 82 408 hospitalisations which met the case definition. From this, 213 (0.3%) private hospital admissions, 955 (1.2%) overseas visitors, 3109 (3.8%) transfers, 12 353 (15.0%) day cases, and 1210 (1.5%) readmissions were excluded. Of the remaining 64 568 cases, 29 were reported to have been discharged dead from hospital (case fatality 0.04%). Hospitalisation data recorded a total of 213 141 hospital days over the study period. The mean and median lengths of stay were 2 and 3.3 respectively. Table 1 shows the incidence of childhood serious skin infections in both the Tairawhiti region and NZ during 1990-1999 (ICD-9) and 2000-2007 (ICD-10). As recommended by the previous work developing the case definition, these data are disaggregated by category and level of diagnosis.13 During the earlier time period, 1990-1999, the age-adjusted total incidence of infections in Tairawhiti was 641.1/100 000 while the total NZ incidence was slightly over half this rate at 354.3/100 000. By 2000-2007 the incidence in Tairawhiti had increased by over 50% to 988.4/100 000, while that in NZ had increased by a similar proportion to 531.7/100,000. A more detailed version of this table is provided in the Appendix. Table 1. The incidence of serious skin infections in children aged 0-14 years in Tairawhiti and NZ, disaggregated by category and level of diagnosis, between 1990-1999 (ICD-9) and 2000-2007 (ICD-10) Category Level of diagnosis Tairawhiti region New Zealand 1990-1999 2000-2007 1990-1999 2000-2007 No. 2020 Rate 2021 No. 2020 Rate 2021 No. 2020 Rate 2021 No. 2020 Rate 2021 Serious skin infections of typical sites (previously used case definition) Principal All level 352 456 284.5 368.5 431 580 453.8 610.7 13541 17074 166.3 209.7 18177 24086 264.9 351.0 Serious skin infections of atypical anatomical sites Principal All level 72 100 58.2 80.8 37 45 39.0 47.4 3170 5233 38.9 64.3 1866 2270 27.2 33.1 Serious skin infections secondary to primary skin disease Principal All level 81 194 65.5 156.8 82 212 86.3 223.2 1406 5364 17.3 65.9 1909 6170 27.8 89.9 Serious skin infections secondary to external trauma Principal All level 25 48 20.2 38.8 15 76 15.8 80.0 635 1270 7.8 15.6 420 3101 6.1 45.2 Crude total serious skin infections Principal All level 530 798 428.3 644.9 565 913 594.9 961.4 18752 28941 230.3 355.4 22372 35627 326.0 519.2 Age-adjusted total serious skin infections All level 641.1 988.4 354.3 531.7 ICD: international classification of disease. 2020Total number of cases during time period. 2021Average annual incidence per 100 000 (based on usually resident population counts from NZ Census). Incidence by year and season, 1990-2007Figures 1 and 2 illustrate the incidence of serious skin infections in the Tairawhiti region and in the whole of NZ during each of the 18 years studied. Results are presented for incidence by category and in total. Between the years 1990 and 2007 the incidence of infections more than doubled in the Tairawhiti region (from 423.6/100 000 in 1990 to 952.6/100 000 in 2007), while the NZ incidence increased by just under double (from 298.0/100 000 to 547.3/100 000). In both settings the increasing infection incidence was a direct reflection of increases in the incidence of serious skin infections of typical sites, along with a small contribution from infections secondary to primary skin trauma. Infections secondary to primary skin disease increased less, and those of atypical sites declined over this period. Figure 1. The incidence of serious skin infections in 0-14-year-old children in the Tairawhiti region by category and year, 1990-2007 Figure 2. The incidence of serious skin infections in 0-14-year-old children in NZ by category and year, 1990-2007 Table 2 shows the seasonal variation in the incidence of serious skin infections. In NZ, the crude incidence of infections was significantly higher during summer and autumn compared to winter (RR 1.12 for both). This trend was less distinct in Tairawhiti, with no significant difference in the seasonal incidence of infections. There was, however, no statistically significant difference in this trend between Tairawhiti and NZ. Table 2. The crude incidence of serious skin infections in 0-14 year old children by season, gender, age group, ethnicity and deprivation level for the Tairawhiti region and NZ, 1990-2007 Variable Category