Serious Skin Infection in Children: Midland Region James Scarfe Public Health Analyst Toi Te Ora Public Health Service

Transcription

1 Serious Skin Infection in Children: Midland Region James Scarfe Public Health Analyst Toi Te Ora Public Health Service Brent Harvey Analyst Clinical & Systems HealthShare Limited Dr. Justin Wilde General Paediatrician Bay of Plenty District Health Board On behalf of the Midland Child Health Action Group Skin Sub-Group December 2014

2 Acknowledgements The Midland Child Health Action Group and HealthShare Limited would like to express our appreciation to the Toi Te Ora Public Health Service who have made the completion of this report possible. In particular, we would like to acknowledge James Scarfe, who has kindly offered his time, expertise and skill in completing the raw data extraction and calculations on behalf of the project team. 1

3 Summary The purpose of this report is to describe the epidemiology of serious skin infection in the Midland region (covering five District Health Boards (DHBs) Bay of Plenty, Lakes, Tairawhiti, Taranaki and Waikato) and to provide a baseline to measure progress. Discharge data was analysed for all children admitted to hospital with a serious skin infection between 2001 and The incidence of skin infection in children in the Midland area is significantly higher than the incidence rate for all children living in New Zealand. The average incidence per year for the period was 57.5/10,000. Incidence for the period has seen an increase to 63.2/10,000, with a regional peak of 66.1/10,000 in There is significant variance between the five Midland District Health Boards (DHB), with Tairawhiti DHB having the highest incidence rates for all years between 2001 and 2012 ( : 110.6/10,000), followed by Lakes DHB ( : 79.1/10,000) and Bay of Plenty DHB ( : 77.9/10,000), with lower incidence observed in Waikato DHB ( : 48.6/10,000) and Taranaki DHB ( : 45.1/10,000). Higher incident rates of serious skin infection were observed in younger children (0-4 years), boys, and Maori. Maori rates are approximately three times higher than non-maori rates, suggesting significant inequalities. Admissions to hospital for serious skin infection are preventable through effective treatment in primary care or prevention of skin infection. From , Midland District Health Boards spent approximately $9.4 million on treating serious skin infection in hospital. The expenditure and resourcing invested in inpatient services is presently the only item we can easily quantify, with further expenditure in primary care, pharmaceuticals, and other secondary services (non-admitted attendances, outpatient events and readmissions) likely to be significant across the region. 2

4 Table of Contents Acknowledgements... 1 Summary... 2 List of Tables... 4 List of Figures Background Methods Definition of Serious Skin Infection Data Analysis Results Summary of Key Results Changes in the Incidence of Serious Skin Infection Over Time Age Ethnicity Category of Serious Skin Infection Geographic Distribution Cost of Serious Skin Infection Serious Skin Infection in Primary Care Conclusions Recommendations References

5 List of Tables Table 1: International Classification of Diseases Clinical Modification 10 (ICD-10) codes included in the O Sullivan & Baker (2010) Serious Skin Infection case definition... 8 Table 2: Colour Palette adopted in charts in this report Table 3: Counts of serious skin infection in children (0-14) Table 4: Yearly incidence (per 10,000) of serious skin infection in children (0-14) Table 5: Changes in incidence of serious skin infection for Midland area Table 6: Counts of serious skin infection by age at admission Table 7: Average yearly Incidence of skin infection by age and ethnicity Table 8: Counts of serious skin infections by age, category and ethnicity Table 9: Skin infections by type (principal diagnosis only) Table 10: The 25 CAUs with the greatest number of serious skin infection Table 11: The 25 CAUs with the greatest proportion of DHB serious skin infection Table 12: Secondary Services Inpatient cost of treating incident cases of serious skin infection in children (0-14) Table 13: Estimated number of primary care cases of skin infection

6 List of Figures Figure 1: Incidence of serious skin infection in children aged 0-14 years in New Zealand and the Midland region Figure 2: Incidence of serious skin infection in children aged 0-14 years in New Zealand and the Midland region Maori versus non-maori Figure 3: Incidence of serious skin infection in children aged 0-14 years in New Zealand and the Midland region by District Health Board Figure 4: Incidence of serious skin infection in children aged 0-14 years in New Zealand and the Midland region by District Health Board Maori Figure 5: Incidence of serious skin infection in children aged 0-14 years in New Zealand and the Midland region by District Health Board Maori versus non-maori Figure 6: Incidence of serious skin infection by age group Figure 7: Incidence of serious skin infection by age group - Maori Figure 8: Incidence of serious skin infection in children in New Zealand and Midland Maori versus non-maori Figure 9: Midland Incidence of serious skin infection by the category of infection

7 1 Background Skin infection is an important cause of morbidity in children in New Zealand. Skin infection contributes to ethnicity-related health inequalities and is associated with socioeconomic deprivation. The rates of admission to hospital for skin infection doubled between 1990 and 2007 (O Sullivan, Baker, & Zhang, 2011). During this time most Midland DHBs had incident rates of skin infection significantly higher than the national incident rate (O Sullivan, et al., 2011). Hospital treatment of skin infection often involves invasive and traumatic procedures resulting in hospital stays that can be of a significant length of time. Skin infection, if found in its early stages, can be effectively treated in primary care, therefore most skin infections are seen as an ambulatory sensitive admission. However, even though treatment of skin infection in primary care is cost saving and less traumatic for children, primary care treatment still does not address the fundamental causes of skin infection. In 2014 the Child Health Action Group (CHAG) identified Skin Infections in children as a priority issue to address in the 2013/14 Midland Regional Services Plan. CHAG is a clinical network representing a range of health stakeholder groups within the Midlands region, with membership covering the five DHBs within the region. Within the region, Toi Te Ora, the Public Health Service representing Bay of Plenty and Lakes DHBs have undertaken an excellent project to support their ambitious strategic goal of reducing childhood admissions for skin infections by two thirds by The Toi Te Ora work completed in September 2014 has enabled this report to be created promptly, and the Child Health Action Group would like to acknowledge their gratitude to Toi Te Ora for their willingness to collaborate regionally on this project. Hospital admissions provide the most convenient and simplest way to measure both baseline information, and as a method of recording progress and improvement. In order to achieve the goal baseline, rates of childhood skin infections need to be measured and follow up monitoring needs to take place in order measure progress in achieving the goal. It should be acknowledged that hospital admissions will likely account for a small proportion of the broad health sector activity (primary, pharmaceuticals, and non admitted events and attendances) treating skin conditions and infections. The current report is split into three main sections methods, results, and conclusions. The methods section will outline the definition of skin infections used and describe how the analysis was carried out. The results section displays the descriptive epidemiology of skin infection in the Midland region. The conclusion section describes the main findings that can be drawn from the descriptive epidemiology. 6

8 2 Methods 2.1 Definition of Serious Skin Infection The following analysis utilises the O Sullivan & Baker (2010) case definition of serious skin infections in children. The epidemiologic case definition they proposed is: A child aged 0-14 years, admitted to hospital with a principal or additional diagnosis of serious skin infection, with a diagnosis code either within the ICD skin infection sub-chapter, or within the categories of skin infection of an atypical site or skin infection following primary skin disease or external trauma. (O Sullivan & Baker, 2010, p181) The case definition was compared to a clinical description of a serious skin infection in children along with a number of alternative definitions. The above definition achieved a sensitivity of 98.9% and a specificity of 98.8% compared to the traditional definition which includes cases with a primary diagnosis from the International Classification of Diseases (ICD) skin infections subchapter which achieved a sensitivity of 61.0% and specificity of 100.0% (O Sullivan & Baker, 2010). In practice the O Sullivan & Baker (2010) case definition is split into 4 sub-categories. Category A corresponds to the ICD 10 skin infections sub chapters. Categories B-D include skin infections of an atypical site, skin infections secondary to primary skin disease, and skin infections secondary to external trauma. Table 1 displays the ICD10 codes included in each of the four categories. O Sullivan, Baker & Zhang (2011) utilised the case definition to study skin infections in New Zealand children from and applied four criteria in attempt to remove inconsistencies in data recording over time. Four exclusions were applied: (1) readmissions within 30 days with the same principle diagnosis, (2) transfers between hospitals and departments, (3) overseas visitors, and (4) day cases. This study applies the same four criteria to the data. It is important to note that many of the ICD diagnosis codes contribute to the Ambulatory-Sensitive Hospitalisation rates (ASH rate) which records hospitalisations that are potentially preventable through interventions and management in a primary health care setting. ASH rate improvement remains an important target for DHBs to address. 7

9 Table 1: International Classification of Diseases Clinical Modification 10 (ICD-10) codes included in the O Sullivan & Baker (2010) Serious Skin Infection case definition Category A Skin infections of typical Sites L01.0, L01.1 Impetigo L02.0 to L02.9 inclusive Cutaneous abscess, furuncle and carbuncle L03.01 to L03.9 inclusive Cellulitis L04.0 to L04.9 inclusive Acute lymphadenitis L05.0 Pilonidal cyst with abscess L08.0 Pyoderma L08.1, L08.8, L08.9 Other infections of skin and subcutaneous tissue Category B Skin infections of atypical sites A46 Erysipelas H00.0 Hordeolum/cellulitis/abscess eyelid H60.0 to H60.3 inclusive, H62.0, H62.4 Abscess/cellulitis external ear and infective otitis externa J34.0 Abscess/cellulitis nose K61.0 Anal abscess/cellulitis (excludes rectal, ischiorectal or intersphincteric regions) H05.0 Acute inflammation/cellulitis/abscess of orbit N48.2, N49.2, N49.9 Other inflammatory disorders of penis, scrotum and unspecified male genital organ(excludes deeper tissues) N76.4 Abscess/cellulitis of vulva Category C Skin infections secondary to primary disease of the skin B01.8 Varicella with other complications B86 Scabies L30.8,L30.9,L30.30 Dermatitis unspecified and other specified (eczema) and infective eczema Category D Skin infections secondary to trauma S10.13, S10.83, S10.93, S20.13, S20.33, S20.43, Insect/spider bites S20.83, S30.83, S30.93, S40.83, S50.83, S60.83, S70.83, S80.83, S90.83, T00.9, T09.03, T11.08, T13.03, T14.03, T63.3, T63.4 T79.3 Post-traumatic wound infection not elsewhere classified T89.01,T89.02 Open wound infection with foreign body (+ infection) and open wound with infection 8

10 2.2 Data Analysis Hospital discharge data was obtained from the Ministry of Health. The CHAG project team are grateful to Toi Te Ora for sharing this data and replicating the calculations for Midland DHBs in order to accelerate progress in preparing this report. The data included all children in New Zealand aged 0-14 with a principal or additional diagnosis displayed in table 1. The analysis was undertaken in Microsoft Excel To maintain consistency with the analysis of O Sullivan, Baker & Zhang (2011) the following filters were applied to the data: Day cases were removed Data for any discharge that had a length of stay of less than 1 day was removed from the data set. Transfers To eliminate multiple discharges for the same case any discharge data that was classified as a transfer was removed. Readmissions Any data for discharges that occurred within 30 days of previous discharge with the same principle diagnosis were removed. Overseas visitors Data for non-residents were removed from the dataset. The pivot table function in Microsoft Excel 2010 was used to generate counts of hospital discharges per calendar year of skin infections for 0-14 age group and 0-4, 5-9, and age groups for all five Midland DHBs, as well as a consolidated Midland region and New Zealand view. The estimated usually resident population was used as the denominator to calculate rates of serious skin infections. The estimated usually resident population counts were obtained from the Statistics New Zealand population projection release to the Ministry of Health (November 2013). Population projections within this dataset are based upon projections from the 2006 census. Ethnicity is prioritised based upon the standard rules for prioritised ethnicity. In the absence of earlier consistent information, populations prior to 2006 were set at the 2006 figures in order to accelerate the production of this report. Rates per 10,000 people were calculated. Ninety five percent confidence intervals were calculated where appropriate. The following section presents the results of the analysis. Key assumptions on the data and our calculations are made. These include: That ICD diagnosis coding is consistent across both DHBs, and the years analysed. The focus of our analysis is on Maori and non-maori populations. Due to this the Pacific Island population (denominator) and skin infections (numerator) are both included as non-maori. This may result in some data inconsistencies when comparing or benchmarking across DHBs where Pacific populations may be proportionately larger than other DHBs (Waikato for example). 9

11 As skin infection data is based on both primary and all diagnosis codes, where a case records two relevant skin infection diagnosis codes (for example a primary diagnosis within category A, and another diagnosis within category D) a basic prioritisation is undertaken that records only one case categorised by selecting category A before B and so on. The readers of this report are anticipated to be broad across the Midland region. To assist the readers of this report to identify and dissect for the information that is most important to them, a colour palette has been adopted uniformly across all graphs as indicated below. Table 2: Colour Palette adopted in charts in this report Midland New Zealand Bay of Plenty DHB Lakes DHB Tairawhiti DHB Taranaki DHB Waikato DHB Where Maori and non-maori are charted on the same graph, Maori will be indicated by a bold line, with non-maori dotted. The three age groups (0-4, 5-9 and 10-14) will be indicated by different dashes. 10

12 3 Results 3.1 Summary of Key Results Figure 1 shows that the Midland region consistently has a higher rate of serious skin infections in children than the overall rate for New Zealand (NZ). Figure 1: Incidence of serious skin infection in children aged 0-14 years in New Zealand and the Midland region Figure 2 shows that the Midland region consistently has a higher rate of serious skin infections in Maori children than the overall rate for New Zealand Maori. The Midland Maori rate is significantly higher than the Midland non-maori rate, with the incidence rate being approximately times higher for Maori. Figure 2: Incidence of serious skin infection in children aged 0-14 years in New Zealand and the Midland region Maori versus non-maori 11

13 Figure 3 illustrates the large regional variance by charting each DHB against Midland and New Zealand. Tairawhiti DHB consistently records the highest incidence rates, approximately twice the national rate. Two DHBs, Lakes DHB and Bay of Plenty DHB are above both the Midland and national incidence rates, with Waikato DHB and Taranaki DHB below both. Figure 3: Incidence of serious skin infection in children aged 0-14 years in New Zealand and the Midland region by District Health Board 12

14 Figures 4 and 5 illustrate the regional variance for Maori and non-maori by charting each DHB against Midland and National. All DHBs record inequity between Maori and non-maori incidence rates. Greater variation is evidenced in Maori incidence rates when comparing between DHBs than non-maori rates which appear more tightly clustered and consistent. Figure 4: Incidence of serious skin infection in children aged 0-14 years in New Zealand and the Midland region by District Health Board Maori Figure 5: Incidence of serious skin infection in children aged 0-14 years in New Zealand and the Midland region by District Health Board Maori versus non-maori 13

15 3.2 Changes in the Incidence of Serious Skin Infection Over Time Table 3 displays the counts per year for the Midland region and by DHB. The absolute number of infections per year has increased over the 12 year time period, for all DHBs and the Midland region. The incidence per 10,000 people has increased for all DHBs and the Midland region (Table 4). Over the entire time period ( ) boys are at higher risk of skin infection compared to girls. In the last time period ( ) the difference between girls and boys decreased. The decrease in rate ratio was not due to a decreasing incidence in boys but was due to an acceleration in the incidence in girls. Over the entire time period, the incidence in total counts increased by 20% in girls, but only 6% in boys. This risk profile exists for all DHBs (and for both Maori and non-maori), and for most years, with the exception of Lakes DHB which has a greater incidence rate in girls than boys in 2010 and The incidence of serious skin infection in Maori children has increased over the last twelve years. The incidence of skin infection in Maori is approximately three times the incidence of skin infection in Non-Maori. The disparity between Maori and Non-Maori has increased over the last twelve years. Additional data and analysis has been conducted in a range of segments, including ethnicity (Maori versus non-maori), age band (0-4, 5-9, and 0-14), region (all five DHBs, Midland region and New Zealand), diagnosis (all or primary), gender (male, female and total) and year ( individually and three, four and six yearly groupings). If required, these can be sourced from the authors of this report. 14

16 Table 3: Counts of serious skin infection in children (0-14) DHB / Region Level Total ,727 All levels Bay of Plenty DHB ,426 Principal only Lakes DHB Tairawhiti DHB Taranaki DHB Waikato DHB Midland New Zealand Midland Percentage of New Zealand All levels Principal only All levels Principal only All levels Principal only All levels Principal only All levels Principal only All levels Principal only All levels Principal only , , , , , ,684 1,000 1,026 1, ,058 1,111 1,203 1,161 1,116 12, ,026 4,460 4,550 4,432 4,444 4,267 4,402 4,657 4,941 4,884 5,219 5,186 4,920 56,362 2,830 2,888 2,801 2,897 2,738 2,872 3,086 3,212 3,190 3,275 3,296 3,114 36,199 22% 23% 24% 22% 22% 22% 21% 21% 23% 23% 22% 23% 22% 23% 23% 23% 23% 22% 21% 22% 20% 22% 22% 22% 22% 22% 15

17 Table 4: Yearly incidence (per 10,000) of serious skin infection in children (0-14) DHB / Region Level All levels Bay of Plenty DHB Principal only Lakes DHB Tairawhiti DHB Taranaki DHB Waikato DHB Midland New Zealand All levels Principal only All levels Principal only All levels Principal only All levels Principal only All levels Principal only All levels Principal only

18 Table 5: Changes in incidence of serious skin infection for Midland area Variable f Incid.* RR (95%CI) f Incid.* RR (95%CI) f Incid.* RR (95%CI) f Incid.* RR (95%CI) Total 3, , , , Gender Male 1, Female 1, Age 0-4 1, Ethnicity Maori 1, Non-Maori 1, ( ) 1.00 ⱡ 2.57 ( ) 1.17 ( ) 1.00 ⱡ 2.58 ( ) 1.00 ⱡ 1, , , , , ( ) 1.00 ⱡ 2.65 ( ) 1.16 ( ) 1.00 ⱡ 2.93 ( ) 1.00 ⱡ 1, , , , , ( ) 1.00 ⱡ 3.06 ( ) 1.25 ( ) 1.00 ⱡ 2.87 ( ) 1.00 ⱡ 1, , , , , ( ) 1.00 ⱡ 2.67 ( ) 1.10 ( ) 1.00 ⱡ 2.77 ( ) RR, Rate Ratio; CI, confidence interval f, Frequency is the number of cases in the three year period * Average annual incidence per 10,000 based on estimated usually resident population (from Statistics NZ population projections to Ministry of Health based on 2006 Census) ⱡ Arbitrary reference category 1.00 ⱡ 17

19 3.3 Age For all age groups children in the Midland area have higher incidence rates of serious skin infection than New Zealand children (Figure 5). The 0-4 age group has both the highest incidence of serious skin infection in both the Midland area and New Zealand, and the largest negative variation from the New Zealand incidence rate. Figure 6: Incidence of serious skin infection by age group For all age groups Maori children in the Midland area have higher incidence rates of serious skin infection than New Zealand Maori children (Figure 5). The 0-4 age group has the highest incidence of serious skin infection for Maori in both the Midland area and New Zealand. The age group reports a significant variation between Midland and New Zealand Maori. Figure 7: Incidence of serious skin infection by age group - Maori 18

20 To provide additional insight the proportion of counts attributable to each age group is displayed in Table 6. Children aged 0-2 years together account for approximately 42% of admissions for serious skin infection. It is clear that a subgroup of the 0-4 age group account for the majority of skin infection admissions (2012: 56% Total population and 62% Maori, and 55% of the total count from ). This is a key point because it indicates that if the Midland region DHBs are to reduce admissions for serious skin infection, it needs to carefully consider strategies to influence interventions with the age group most in need of support. Table 6: Counts of serious skin infection by age at admission Age at Admission (years) Count (%) Count (%) Count (%) (14.7%) 1,330 (20.0%) 2,203 (17.5%) (13.9%) 1,008 (15.2%) 1,831 (14.6%) (9.8%) 652 (9.8%) 1,235 (9.8%) (8.0%) 457 (6.9%) 932 (7.4%) (6.1%) 365 (5.5%) 725 (5.8%) (5.6%) 364 (5.5%) 694 (5.5%) (4.9%) 293 (4.4%) 583 (4.6%) (5.1%) 274 (4.1%) 574 (4.6%) (5.0%) 308 (4.6%) 604 (4.8%) (4.2%) 255 (3.8%) 505 (4.0%) (5.5%) 290 (4.4%) 614 (4.9%) (4.2%) 276 (4.2%) 522 (4.2%) (4.2%) 258 (3.9%) 509 (4.1%) (4.5%) 240 (3.6%) 506 (4.0%) (4.4%) 266 (4.0%) 524 (4.2%) Total 5,925 6,636 12,561 19

21 3.4 Ethnicity The incidence of serious skin infection is higher in Maori than Non-Maori throughout New Zealand and the Midland region (Figure 7). Furthermore Maori in the Midland region have a higher incidence of serious skin infection than Maori throughout New Zealand. Comparing the Non-Maori incidence may not be valid due to the demographic features of the Non-Maori group in the Midland region being statistically different to those of Non-Maori from New Zealand. Figure 8: Incidence of serious skin infection in children in New Zealand and Midland Maori versus non-maori In the Midland region, Maori children of all age groups display a higher incidence of skin infection than Non-Maori (Table 7). In fact the oldest year old age group (lowest risk age group) in Maori has a higher incidence than the 0-4 age group (highest risk age group) in Non-Maori. Table 7: Average yearly Incidence of skin infection by age and ethnicity Ave. Incidence per year, per 10, Maori Non-Maori Rate Ratio Maori Non-Maori Rate Ratio Maori Non-Maori Rate Ratio Maori Total 0-14 Non-Maori Rate Ratio

22 3.5 Category of Serious Skin Infection The majority of serious skin infections in the Midland region are from category A (Figure 8). This is in line with national trends. The category A rates display a significant amount of year to year variation, with the variation under pinning the trend for all serious skin infections. There is not sufficient evidence to suggest that rates for category A are increasing. Categories B and D have remained stable over the 12 year period while the incident rate of category C has increased at a largely steady rate. Infections included in category C appear to be driving the overall increases seen in the Midland region. Awareness of these conditions, in particular dermatitis (Table 9), may need to be raised. Also infections in category C are not part of the Ministry of Health ASH monitoring so as a result less emphasis may be put of reducing these types of admissions. Figure 9: Midland Incidence of serious skin infection by the category of infection The proportion of counts attributable to each category changes with age and ethnicity (Table 8). The 0-4 age group has the lowest proportion of category A counts. The 5-9 and age groups have a higher proportion of category A counts. The 0-4 age group has a larger proportion of skin infections from category C than the 5-9 and age groups. While the proportion of category D counts increase with age from 4.0% for 0-4 to 17.5% for There are significant differences between skin infections in Maori and Non-Maori. Maori children have a higher number of infections arising from infections from category C while Non-Maori have a higher proportion from category D. 21

23 Table 8: Counts of serious skin infections by age, category and ethnicity Age Group Category A Category B Category C Category D Total Counts 0-4 4,178 (60.3%) 393 (5.7%) 2,079 (30.0%) 275 (4.0%) 6, ,108 (71.2%) 161 (5.4%) 304 (10.3%) 387 (13.1%) 2, ,852 (69.2%) 161 (6.0%) 193 (7.2%) 469 (17.5%) 2,675 Maori 5,087 (64.5%) 390 (4.9%) 1,878 (23.8%) 531 (6.7%) 7,886 Non-Maori 3,051 (65.3%) 325 (7.0%) 698 (14.9%) 600 (12.8%) 4,674 Total 8,138 (64.8%) 715 (5.7%) 2,576 (20.5%) 1,131 (9.0%) 12,560 The proportion of infections from each subcategory was investigated (Table 9). In this case only the principal diagnosis was considered. The three most common subcategories are cutaneous abscess, furuncle and carbuncle (category A), cellulitis (category A), and dermatitis (category C). These three subcategories accounted for 77% of all principal diagnosis skin infections from 2001 to

24 Table 9: Skin infections by type (principal diagnosis only) % % % Category A Skin infections of typical Sites Impetigo Cutaneous abscess, furuncle and carbuncle Cellulitis Acute lymphadenitis Pilonidal cyst with abscess Pyoderma Other infections of skin and subcutaneous tissue Category B Skin infections of atypical sites Erysipelas Hordeolum/cellulitis/abscess eyelid Abscess/cellulitis external ear and infective otitis externa Abscess/cellulitis nose Anal abscess/cellulitis (excludes rectal, ischiorectal or intersphincteric regions) Acute inflammation/cellulitis/abscess of orbit Other inflammatory disorders of penis, scrotum and unspecified male genital organ(excludes deeper tissues) Abscess/cellulitis of vulva Category C Skin infections secondary to primary disease of the skin Varicella with other complications Scabies Dermatitis unspecified and other specified Category D Skin infections secondary to trauma Insect/spider bites Post-traumatic wound infection not elsewhere classified Open wound infection with foreign body (+ infection) and open wound with infection Total

25 3.6 Geographic Distribution Tables 10 and 11 display the top 25 by census areas units (CAU) by highest proportion of serious skin infections and by highest incidence across the Midland region for the , and total time periods. For clarity, and to allow for targeting of remedial approaches, the DHB and Territorial Authority are shown alongside the CAU. Table 10 ranks CAUs by the number of skin infections. It displays the 25 CAUs with the greatest number of skin infections. In total, these 25 CAUs represent approximately 28 percent of the Midland regions cases. This indicates that if effective measures are taken in a small number of neigbourhoods the burden of serious skin infection can be significantly reduced. The absolute number of skin infections in the 25 highest ranking CAUs has increased marginally from 1,692 in the period from to 1,797 from In the time period nine of the 25 CAUs are within Bay of Plenty DHB, seven in Waikato, six in Tairawhiti, and three in Lakes DHB. No CAUs are represented in the top 25 by total incidence from Taranaki DHB, although Taranaki DHB does have CAUs that represent a higher proportion of DHB incidence (Table 11). Table 10: The 25 CAUs with the greatest number of serious skin infection Census Area Unit Territorial Authority DHB % of skin admission by DHB % of skin admission by DHB Total % of skin admission by DHB Kawerau Kawerau Bay of Plenty Western Heights Rotorua Lakes Kaiti South Gisborne Tairawhiti Opotiki Opotiki Bay of Plenty Te Hapara Gisborne Tairawhiti Ngaruawahia Waikato Waikato Gisborne Central Gisborne Tairawhiti Fordlands Rotorua Lakes Whakatane North Whakatane Bay of Plenty Te Puke Community East Western BOP Bay of Plenty Hamilton Central Hamilton Waikato Mangapapa Gisborne Tairawhiti Gisborne Airport Gisborne Tairawhiti Yatton Park Tauranga Bay of Plenty Outer Kaiti Gisborne Tairawhiti Bader Hamilton Waikato Whakatane West Whakatane Bay of Plenty Gate Pa Tauranga Bay of Plenty Otakiri Whakatane Bay of Plenty Melville Hamilton Waikato Brymer Hamilton Waikato Huntly East Waikato Waikato Riverlea Hamilton Waikato Pukehangi North Rotorua Lakes Murupara Whakatane Bay of Plenty

26 Incidence of skin admissions in raw numbers alone may miss identifying CAUs with high need. An example of this is Waitara West in Taranaki DHB which although not listed in Table 10, has a significant proportion of it s DHBs skin admissions. Table 11 ranks CAUs by the proportion of skin infections within each DHB. This will help to identify communities with both incidence and proportions, as well as enable DHBs with lower incidence to identify any communities with high proportions. Table 11: The 25 CAUs with the greatest proportion of DHB serious skin infection Census Area Unit Territorial Authority DHB % of skin admission by DHB % of skin admission by DHB Total % of skin admission by DHB Kaiti South Gisborne Tairawhiti Te Hapara Gisborne Tairawhiti Western Heights Rotorua Lakes Gisborne Central Gisborne Tairawhiti Mangapapa Gisborne Tairawhiti Gisborne Airport Gisborne Tairawhiti Outer Kaiti Gisborne Tairawhiti Fordlands Rotorua Lakes Waitara West New Plymouth Taranaki Tamarau Gisborne Tairawhiti Kawerau Kawerau Bay of Plenty Frankleigh New Plymouth Taranaki Pukehangi North Rotorua Lakes Opotiki Opotiki Bay of Plenty Hawera South South Taranaki Taranaki Owhata West Rotorua Lakes Koutu Rotorua Lakes East Cape Gisborne Tairawhiti Waitara East New Plymouth Taranaki Kuirau Rotorua Lakes Marfell New Plymouth Taranaki Mangakakahi Rotorua Lakes Bell Block New Plymouth Taranaki Westown New Plymouth Taranaki Ngaruawahia Waikato Waikato

27 3.7 Cost of Serious Skin Infection The Toi Te Ora report on serious skin infection estimated that the cost of treating each infection during the financial year ending 2010 was $2,659 (Lowe, Ingram-Seal, & de Wet, 2011). This cost had been stable since the financial year ending A brief analysis of Midland cases in 2013/14 suggests that the average caseweight attributed to inpatient events is approximately 0.6 caseweights per event. Converted to dollars, this is approximately $2,700 per discharge, materially confirming that the Toi Te Ora finding remains a fair judgement. Assuming the cost of treating serious skin infection is similar for each DHB the cost of treating serious skin infection can be estimated (Table 10). From 2010 to 2012 serious skin infection is estimated to have cost Midland District Health Boards $9.396 million, or $3.132 million per annum. Table 12: Secondary Services Inpatient cost of treating incident cases of serious skin infection in children (0-14) DHB Number of cases Cost over 3 years (cost per year) BOPDHB 1,035 $2,750,000 ($920,000) Lakes DHB 526 $1,400,000 ($470,000) Tairawhiti DHB 404 $1,091,000 ($364,000) Taranaki DHB 305 $ 824,000 ($275,000) Waikato DHB 1,210 $3,267,000 ($1,089,000) Midland 3,480 $9,396,000 ($3,132,000) The estimate given will be an underestimate of the true cost of treating serious skin infection because the estimate is based on the number of incident cases. Re-admissions of the same case have not been taken into account. The wider cost, inclusive of readmissions, primary care, nonadmitted events, outpatient events and pharmaceuticals as examples cannot presently be quantified with any degree of confidence or accuracy. Indicatively, Midland DHBs spend on average $800,000 per annum (calendar years 2011 and 2012) on the drug cost (excluding dispensing fees) for the dermatological therapeutic group of pharmaceuticals in 0-14 year olds. Included within this is emollient expenditure of approximately $140,000 per annum. 26

28 3.8 Serious Skin Infection in Primary Care O Sullivan (2011) conducted an analysis comparing the number of skin infections treated in hospital to the number of skin infections treated in primary care in Tairawhiti District Health Board and found that, for every serious skin infection treated in hospital, a further 14 were treated in primary care. No attempt to verify this ratio has been undertaken as part of this report. Using this ratio the number of primary care cases in the Midland area is estimated in Table 11. These estimates should only be taken as a rough estimate as the ratio used is from one DHB from within our region, which has not been verified by other DHBs. The estimates suggest that in addition to the 6,635 cases of serious skin infection treated in hospital from 2007 to 2012, a further 92,890 may have been treated in primary care. To assist in understanding this 1:14 ratio further, work will be undertaken to analysis pharmaceutical dispensing patterns and the ratio between primary and secondary prescriptions, as well as whether primary data is available to support the ratio. This will be investigated in the next phase of data collection. Table 13: Estimated number of primary care cases of skin infection Number of hospital cases of serious skin infection Estimated number of primary care cases ,058 1,111 1,203 1,161 1,116 13,804 14,812 15,554 16,842 16,254 15,624 27

29 4 Conclusions The incidence of serious skin infection in children within the Midland region has increased over the years A spike occurred in 2009, 2010 and Rates in 2012 fell back to levels seen in Serious skin infection incidence in children is statistically significantly higher in compared to Maori children are at much higher risk of serious skin infection than non-maori. The disparity remained relatively constant over the period with Maori nearly three times as likely to have a serious skin infection as non-maori. Maori in the Midland region have a higher incidence that Maori throughout New Zealand. The incident rate varies dramatically by DHB across the region, with Tairawhiti DHB having the highest incident rate, nearly twice the national average, Lakes DHB and Bay of Plenty DHB both significantly above national average, and Waikato DHB and Taranaki DHB both below national average. This is not to suggest that health gain cannot be made in Waikato DHB and Taranaki DHB, with both of these DHBs recording some CAUs with high incidence, but does indicate that the area of greatest need and regional inequity is in the other three DHBs. There is a clear relationship between serious skin infection and age group. The risk ratio for serious skin infection decreases as the age group increases. Higher incidence of skin infection is observed in young children, especially in 0-2 year olds. Forty-two percent of children admitted to hospital in the Midland region for serious skin infection were aged two or younger. Increased risk of serious skin infection was observed in boys. For the period % of spin infection incidences were observed in boys. The disparity between boys and girls is decreasing, with results since 2010 showing increased gender equity (2012: Boys 54%). The most common category of infection across all age groups are category A infections. This is partly due to how the categories are defined. The proportion of counts in each category varies by age and ethnicity. Younger children (0-4 age group) have a higher proportion of category C infection. Maori children have a higher proportion of category A, B and C infections than non- Maori. Cutaneous abscess, furuncle and carbuncle (31%), cellulitis (34%), and dermatitis (11%) account for 77% of all principal diagnosis cases of serious skin infection in Midland region. The majority of incident cases of serious skin infection originate from a limited number of census area units. Twenty eight percent of cases come from just 25 census area units. In order of the number of admission to hospital for serious skin infection the top ten CAUs with greatest incidence tabled below. 28

30 Census Area Unit Territorial Authority DHB % of skin admission by DHB % of skin admission by DHB Total % of skin admission by DHB Kawerau Kawerau Bay of Plenty Western Heights Rotorua Lakes Kaiti South Gisborne Tairawhiti Opotiki Opotiki Bay of Plenty Te Hapara Gisborne Tairawhiti Ngaruawahia Waikato Waikato Gisborne Central Gisborne Tairawhiti Fordlands Rotorua Lakes Whakatane North Whakatane Bay of Plenty Te Puke Community East Western BOP Bay of Plenty Serious skin infection represents a significant cost to District Health Boards approximately $3.1 million per year for admitted hospital care within the Midland region alone. Overall serious skin infection is increasing in the Midland region. The increase in serious skin infection has largely been driven by increases in category C infections (skin infections secondary to primary disease of the skin). 29

31 5 Recommendations The following recommendations are made: 1. That the Child Health Action Group adopt this report. 2. That the CHAG disseminate this report to all relevant health organisations and groups, noting to them in particular: a. Skin conditions / infections prevention and management be prioritised in relevant organisational work plans. b. Intervention should focus on Maori children 0-4 years. c. Prevention and earlier detection, intervention, supported patient self management and treatment should be part of any care pathway. 3. That the CHAG direct the Skin sub group to continue to work on a set of resources to assist organisations to decrease the incidence of serious skin infections. 30

32 References Lowe, L., Ingram-Seal, R., & de Wet, N. Bay of Plenty District Health Board, Toi Te Ora - Public Health Service. (2011). Health needs assessment: Cellulitis and skin infections in children in the Bay of Plenty O'Sullivan, C. (2011). Serious skin infection in New Zealand children. (Master's thesis, University of Otago) O'Sullivan, C. E., & Baker, M. G. (2010). Proposed epidemiological case definition for serious skin infection in children. Journal of paediatrics and child health, 46(4), O'Sullivan, C. E., Baker, M. G., & Zhang, J. (2011). Increasing hospitalizations for serious skin infections in New Zealand children, Epidemiology and infection, 139(11), Toi Te Ora - Public Health Service. Bay of Plenty District Health Board, Toi Te Ora - Public Health Service. (2013). Strategic plan Toi Te Ora - Public Health Service, Scarfe, J. Bay of Plenty District Health Board, Toi Te Ora - Public Health Service. (2014). Serious Skin Infections September

33 32