Children's scabies survey indicates high prevalence and misdiagnosis in Auckland educational institutions

Here, we present results of a survey of scabies prevalence in childcare centres and primary schools in Auckland.

1 Scabies is likely to lead to a range of childhood conditions, including cellulitis, abscess and impetigo.Strong statistical associations exist between scabies and both post-streptococcal glomerulonephritis and acute rheumatic fever (ARF). 2 Statistical associations exist between diagnosis and treatment for scabies and the incidence of ARF in Auckland, New Zealand.

What this paper adds
1 The prevalence of undiagnosed scabies in Auckland is high particularly in early childcare centres, particularly those situated in lower socioeconomic areas. 2 Undiagnosed scabies cases who were treated with non-specific anti-pruritic skin lotions and topical steroids suggest that misdiagnosis of scabies is common in primary care.
Scabies is associated with a range of diseases of childhood including bacterial skin infection, post-streptococcal glomerulonephritis 1 and acute rheumatic fever (ARF) a strong possibility. 2We have recently shown early evidence of a higher-than-expected prevalence of scabies in Auckland children, with 21% having typical lesions and 33% of skin samples tested for scabies PCR being positive. 3Here, we present the final results, with analysis adjusted for the survey design and more subjects recruited.
In addition to reporting an estimate of prevalence, we consider possible reasons for the high prevalence of scabies that we have found.Possible explanations for the high prevalence include poor diagnosis, with the condition being frequently mistaken for insect bites or eczema and ineffective treatment being given.Case-series indicate that topical and systemic corticosteroids and antihistamines are frequently prescribed in the case of misdiagnosis, with diagnosis delayed for several months. 4A further possibility is that normalisation occurs in high prevalence communities, such as has been described among indigenous Australians. 5Socioeconomic deprivation, which particularly includes overcrowding, poor access to healthcare and shared sleeping conditions may also contribute to a high prevalence. 6o investigate the prevalence of scabies in New Zealand further, we conducted a designed survey of early childhood education centres and schools in the Auckland Region.

Study design and education setting selection
A designed survey was initially proposed, which aimed to assess the prevalence of scabies among children attending educational settings, including early childcare centres and schools.An area-based measure of socioeconomic status (NZDep) was used to stratify the survey, since scabies was likely to be most prevalent in poorer areas.The design consisted of screening 23 childcare centres in a stratified manner, with half taken from socioeconomically deprived strata (most deprived quintile or other, by area).We also proposed to sample a total of eight schools, including students of four randomly selected classes from within each school.Assuming a worst case 8% prevalence of scabies (by the International Alliance for Control of Scabies [IACS] criteriaexplained later), this would enable an estimate to be within three percentage points.A design effect of three was included to account for clustering of observations within educational centres.Schools and childcare centres were sampled at random proportional to their estimated size, with half taken from each sociodemographic stratum.Lists of schools and childcare centres in the Auckland region were accessed from Ministry of Education websites.
Our recruitment was hampered by the COVID-19 pandemic and associated public health measures which resulted in closure of educational facilities and reticence on the part of these institutions to engage in activities that involved contact with visitors.Parents of children attending each sampled centre were invited to participate and gave informed consent for their child to participate.Parents were invited to fill in a written survey which consisted of questions relating to skin symptoms, signs and recent diagnoses and treatments of their child or household, based on the 2020 IACS criteria for scabies diagnosis.Sociodemographic information including age, gender, and ethnicity were also collected.
If parents selected more than one ethnicity, we determined a 'prioritised' ethnicity for analysis, consistent with standard practice in the New Zealand health sector. 7hysical examination of the child's skin consisted of assessment of the child's arms, legs and abdomen for papules or crusted lesions which suggest scabies or impetigo, respectively.Other skin conditions were also considered.A general practitioner conducted the screening procedures after training and testing from an experienced Paediatrician (DE), with expertise in the diagnosis of scabies.8][9] Skin lesions were classified by appearance as either typical scabies, atypical for scabies or inconsistent with scabies, according to global consensus definitions (2020 IACS criteria). 10ypical lesions occur in groups, are erythematous or may be darkly pigmented papules and are solid and about 2 to 3 mm in diameter.With infection, groups of red ulcerated and crusting lesions were considered typical.Atypical lesions are less than three typical lesions in any body area.Based on the combination of examination features, and history reports (from the parental questionnaire), participants were then assigned to a diagnosis of 'Clinical scabies', 'Suspected scabies', or 'Not scabies'.Confirmation of diagnosis using skin scrapings or dermoscopy was not used.Since there was some reticence to fill in the symptom questionnaire meaning missing data was high, we also reported outcomes from the clinical findings only (typical or atypical scabies skin lesions), since complete information was available for this outcome.
If any typical or atypical or other skin lesion was found, and the child provided verbal assent, samples were taken, by rubbing the lesions with FLOQ swabs saturated with saline, and frozen (À18 C) in a lysis buffer ATL (Qiagen) before PCR analysis, which was carried out in Queensland (CP).If children had several lesions, sites were rubbed with the same swab.The primers for the PCR were designed to detect specific abundant non-coding regions and mitochondrial regions of the Sarcoptes scabiei var.hominis genome. 11Three different targets were tested, including the cox1 (mitochondrial region), and microsatellite highly abundant SSR5 and SSR6 regions.The cox1 gene was thought to be more likely to be conserved. 11The amplification of any target was considered positive for scabies.Testing of samples in our study was conducted in triplicate.
Parents of children who were assessed as having either suspected or clinical scabies from the appearances of their child's skin or from a later positive PCR test were offered permethrin lotion treatment for the participating child and all household contacts and were recommended to follow-up with their family doctor.

Weight estimation
Each participant was assigned a weight commensurate with the sampling design of the study.A 'centre' here was considered either a childcare centre or primary school.First, the probability of the institution being sampled was estimated as the total number attending the selected centres, divided by the total in the school or childcare centre strata, multiplied by the number of centres sampled.Next, the probability of the individual being sampled in the school was made by dividing the total number within the centre who participated divided by the total recorded in the list of attendees.Next, a non-response probability was estimated for each school using a logistic regression model for the centre taking part, given the type of institution and the deprivation decile.Predicted probabilities of participating were then divided into two at the median.The within-group response weight was estimated by calculating the total respondents out of the total sampled, within each 'propensity to respond' group.The three probabilities were then multiplied and the reciprocal taken to convert the probabilities into individual weights.The survey weights were then raked 12 according to the marginal totals by age group (0 to 4 years, and 5 years of age or older) derived from the sampling frame of childcare centres and schools.

Missing data
A substantial proportion (43/181; 23%) of the cohort did not complete a questionnaire, which prevented individuals to be diagnosed with scabies according to the 2020 IACS criteria.

Summary statistics
Descriptive analysis of the proportions of children testing positive from various clinical and laboratory tests were given, by demographic factor and centre, along with the degree of overlap between each factor.Chi-square, Fisher tests and t-tests were used to check for significant associations between sociodemographic characteristics and scabies diagnoses.R software (version 4.2.1) was used for analysis. 13The survey package was used to adjust the point estimate and standard errors of calculations for the survey design by using survey weights. 12Clusters were considered educational centres, and strata were by socioeconomic characteristics of the schools.The srd package was used to illustrate categorical data relationships. 14

Ethics
Ethical approval was granted by the New Zealand Ministry of Health (Manat u Hauora), Health and Disability Ethics Committee (20/STH/41).

Results
The study took place between 11 March 2021 and 28 September 2022, with 181 children taking part from five early childcare centres and two schools (Fig. 1).Non-response at the educational setting level was a particular issue in the survey, with the highest number of centres coming from socioeconomically deprived childcare centres (n = 4), although these comprised only 26.7% of those invited.In the other strata (schools or childcare centres in the two socioeconomic strata), only one centre participated.
The cohort is tabulated by IACS scabies category and by socioeconomic status (Tables 1 and 2).The mean age of children was 5.5 years (standard deviation: 2.9; range: 8 months to 14.1 years).
Most parents of the study population identified as either Pacific (29.7%) or M aori (17.2%).Age in early childcare centres ranged between 8 months and 4 years 8 months, whereas in schools, the age of subjects ranged from 5 to 14 years.
The prevalence of scabies lesions was 51.2% (95% CI: 40.6 to 61.8) among preschool children and 17.8% (95% CI: 0.00 to 35.7) in school children.No PCR positive tests were identified in school children; however, the overall prevalence of PCR positive tests was 14.4% (95% CI: 10.8 to 18.0), very similar to that estimated by the IACS criteria.No substantial difference in PCR positive tests were observed by socioeconomic status of childcare centres.
A total of 64 PCR tests were taken.Of the total, 15 were positive, with five positives diagnosed with 'clinical scabies', five assessed as 'suspected' and four not meeting the IACS criteria despite having suspicious lesions.One child with a positive sample had typical scabies lesions, but no history was available.Of the four who did not meet the criteria, two presented with atypical and two typical scabies lesions.Of the three targets assayed, only the cox1 and SSR5 targets tested positive.Only one specimen tested positive for both targets.Ethnic differences were evident in the lower socioeconomic strata, with M aori and Pacific children more likely to have diagnosed scabies than children of other ethnic groups (P 0.033).These differences were also evident when PCR positive scabies was considered.Weighted estimates showed the highest proportion in M aori, 12.2% (95% CI: 0.0 to 27.1), followed by Pacific 9.5% (95% CI: 3.3 to 15.7), and then European 9.0% (95% CI: 1.3 to 16.8).South Asian and Southeast Asian ethnic groups had no PCR positive cases.
A scaled rectangle diagram depicts the degree of overlap between different classifiers with survey weights applied (Fig. 2a) and with raw counts (Fig. 2b).The area of the rectangles, and degree of overlap, are proportional to the count with each attribute.The figures depict the total population (outer rectangle), with the grey rectangle proportional to children who had skin lesions which were classified as 'typical or atypical' for scabies.The green rectangle is proportional to those who met the IACS clinical criteria.The light-yellow rectangle represents the PCR positive cases.The blue rectangle in both diagrams represents those undergoing topical steroid treatment, presumably for a diagnosis of eczema, and the beige were undergoing non-specific anti-pruritic skin treatment.
The scaled diagrams (Fig. 2) illustrate that use of topical steroids and other anti-pruritics were common in children with evidence of scabies, from clinical examination or laboratory test results.Of the 10 PCR positive cases with history information, 5 (50%) reported the use of a topical anti-pruritic treatment, with two subjects using calamine lotion and three topical steroids, presumably for medically diagnosed eczema.In the study overall, 25.9% of subjects reported use of a topical skin treatment.Those with missing history information are included in the 'missing' column.† Anti-pruritic: includes emollient, anti-fungal, calamine lotion and topical or systemic anti-histamine.ECEC, early childcare education centre.
Weighted estimates for topical skin treatment use were higher than for crude estimates and were 35.8% (95% CI: 8.0 to 63.6) of the surveyed population.The disagreement between IACS clinical criteria for scabies and PCR results is also illustrated.

Discussion
This study shows a high prevalence of scabies in children in the Auckland region.This is the first formal study of scabies prevalence in New Zealand for several decades.The use of PCR as an adjunct to the clinical findings suggests its utility as an objective method of diagnosis.These findings suggest an important contributing factor for the high burden of scabies in New Zealand is poor diagnosis, leading to ineffective treatment and person-to-person spread of the mite.Such findings have been found in other settings. 15 strength of the study was an objective measure of scabies diagnosis, PCR.This latter test is likely to reduce measurement error in prevalence, and given the uncertainty of diagnosis, strengthens our conclusion that scabies is commonly found in Auckland early childcare centres.Although PCR has limited sensitivity, in educational settings, where at least one child is found to be positive, this increases the degree of confidence in the clinical diagnosis that children with similar skin lesions in the centre have scabies.
The study had low success with recruitment, which was partially explained by COVID-19 which disrupted recruitment from schools and childcare centres.Only two schools were recruited, and only one childcare centre associated with a wealthier demographic.This is reflected in the non-representative sample, which includes a high proportion of South Asian children.A high proportion of children had missing demographic details.Several strata only had one centre recruited, so this limits the generalisability of these findings.Our recruitment was relatively higher from socioeconomically deprived early childcare centres, so there is relatively greater confidence in these findings.A  detractor from the quality of the study was the degree of missing information.Studies in the Pacific report much higher participation rates than reported here. 16This makes the generalisability of our results less convincing than for other studies, and there is still some uncertainty about the true prevalence, particularly in wealthier areas.We had a higher participation rate from socioeconomically challenged childcare centres which may be related to the perception of a higher prevalence of skin-related issues in children.This would tend to over-estimate the prevalence of scabies in our results.Also, we note that the PCR test is yet to be formally validated, although preliminary studies have been carried out. 11he PCR test that we used has had early testing against small numbers of clinical scabies cases, diagnosed with other methods.The test consists of three genomic targets and underwent in vitro with the human scabies mite and tested positive, whilst also testing negative to other common human and animal skin parasites.The assay was positive in 5/7 dermatologist confirmed scabies cases, and uniformly negative when samples from people with other dermatological conditions such as dermatitis, psoriasis and tinea were tested (19 were negative for all three targets). 11 recent New Zealand study with which to compare our findings is a case-control study which addressed potential risk factors for rheumatic fever, including scabies. 17In that study in which individuals were asked to report a recent scabies diagnosis, 5.9% (7/117) were positive, whereas in demographically matched controls, the proportion was 1.4% (5/364).This contrasts with our findings of 51.2% of children with typical or atypical lesions in childcare centres and 17.8% in schools.In childcare centres, the prevalence of PCR positive lesions was 14.4% (95% CI: 10.8 to 18.0).In contrast, in our survey, the weighted estimate of responses to the question of whether a family has had scabies treatment in the last 6 months was 3.9% (95% CI: 2.1 to 5.6) in the most deprived strata and no such cases were detected in the least deprived.This comparison highlights the importance of clinical examination over self-report methods for more accurate estimation of scabies prevalence.
Many different opportunities exist to reduce the prevalence of scabies.Our study highlights several, one of which is to improve the accuracy of the diagnosis of skin conditions through raising clinician awareness of the clinical features of scabies, and use of objective laboratory methods, such as PCR.Our experience was that the skin swab used here was a convenient and acceptable method for gaining objective evidence for the diagnosis.
A recent publication suggests a formal framework for scabies control, based on prevalence of the condition in that region. 18he framework proposes three key strategies: 1 Mapping the disease burden. 2 Control interventions, using annual, ivermectin-based mass drug administration in regions with a community prevalence of 10% or more.3 Monitoring and evaluation of prevalence and effects of interventions.
Our findings suggest that improved clinical diagnosis is likely to help.They also suggest that healthcare professionals should have a high index of suspicion for scabies, when reviewing children with itchy rash.We are currently working on using image recognition to standardise and improve the diagnosis of scabies, and we also believe that screening of education centres are likely to be useful, since many cases of scabies in our study had either not been diagnosed or sought medical attention.We note that half of children identified with PCR positive scabies were undergoing treatment for other skin disorders.Further, of those who were PCR positive, none were treated for scabies.
In summary, we highlight the high prevalence of scabies in Auckland children attending pre-school educational settings.Improving the diagnosis and ensuring treatment success of this important disease is likely to improve ethnic inequality in health status, with the potential to reduce the incidence of bacterial skin infection, post-streptococcal glomerulonephritis and likely ARF, 2 thereby promoting ethnic health equity.On the basis of our findings, we believe public health authorities should carefully consider further screening for scabies in educational settings and initiating mass drug administration with ivermectin in centres with a prevalence of scabies greater than 10%.

Fig. 2
Fig. 2 Scaled rectangle diagram showing (a) weighted and (b) raw counts of survey sample with various scabies diagnoses, topical skin treatments and polymerase chain reaction (PCR) test results.

Table 1
Study cohort by IACS scabies diagnosis category in low socioeconomic area education centres

Table 2
Study cohort by IACS scabies diagnosis category in high socioeconomic status education centresThose with missing history information are included in the 'missing' column.† Anti-pruritic: includes emollient, anti-fungal, calamine lotion and topical or systemic anti-histamine.ECEC, early childcare education centre.