Profile of Helicobacter pylori infections among children in the South Island of New Zealand (2010–2021)

Helicobacter pylori is a gram‐negative gut bacterium most often acquired during childhood. International guidelines state that children with suspected H. pylori infection should be referred to a gastroenterologist for investigation via gastroscopy and biopsy. Eradication therapy should be prescribed for children with peptic ulcer disease or following a treatment risk/benefit discussion for those with an incidental gastroscopy finding. Guidelines state that for children a “test‐and‐treat” approach is not warranted, contrasting recommendations for adults. The aim of this study was to profile pediatric H. pylori infections in the South Island of New Zealand (NZ) to determine diagnostic and management strategies, and adherence to international guidelines.


| INTRODUC TI ON
3][4] H. pylori is generally acquired during childhood, typically through intrafamilial spread via transmission routes such as food and resource sharing, hygiene practices, and water or food contamination. 2,5,6While most children with H. pylori are asymptomatic, 7,8 those that experience symptoms may present with gastrointestinal (GI) manifestations such as abdominal pain, vomiting, or extraintestinal features subsequent to these such as iron deficiency anemia and growth failure as defined by linear and ponderal growth, and height for age z scores. 5,9,10 to 90% of adults with stomach cancer are infected with H. pylori, 11 however, children rarely suffer long-term complications such as peptic ulcer disease (PUD), gastric lymphoma, and gastric carcinoma. 12[15][16] Testing for H. pylori may be carried out in several ways.
8][19] Invasive testing involves histologic assessment of mucosal biopsies obtained endoscopically with or without the use of a rapid biopsy urease test (CLO). 18ternational guidelines for the treatment and management of H. pylori in children were first provided in 2000 but have subsequently been updated and enhanced to reflect changing patterns of this infection. 7,12,20,21Several core recommendations have been maintained over time.These state that a "test and treat" approach is not warranted, that is, using noninvasive testing and prescribing eradication therapy to all that are positive. 7,12,20,21All symptomatic children should be referred for diagnosis via gastroscopy and biopsies only, and eradication therapy should only be given following discussion of the risks and benefits of treatment for children in whom the gastroscopy finding is incidental, or for those with PUD. 7,12,20,210][21][22][23] In New Zealand (NZ), national online pediatric guidelines available to all clinicians reinforce the same message as the international publications. 24veral small, regional studies in NZ have demonstrated that H. pylori was prevalent in 20%-39% of children screened prior to 2010. 25,26Children in NZ with H. pylori infection have a similar sociodemographic profile to adults, with higher rates in those identifying as Māori (21%-27%), Pasifika (26%-71%) and Asian (25%) ethnicity than among NZ European (7%-14%). 25,26An association with low socioeconomic status has also been reported. 2,25,26The current epidemiology of H. pylori infection during childhood, and longitudinal outcomes, have not been well established in NZ, and the management strategies have not yet evaluated against international standards.The aim of this project was to retrospectively analyze pediatric H. pylori infections across the South Island of NZ to describe which children present with infection, socioeconomic variables associated with infection, and to determine whether diagnostic and management strategies adhere to national and international guidelines.

| Study design
This study was a 12-year retrospective review of children diagnosed with H. pylori in the South Island of NZ.Positive cases were identified from regional pathology and hospital records, and a longitudinal review of each case was undertaken.

| Inclusion criteria
Children aged ≤18 years, resident in the South Island of NZ, a positive test for H. pylori infection between January 2010 and December 2021.

| Exclusion criteria
Children with an equivocal H. pylori serology result in isolation, those living in the North Island of NZ, and those whose positive H. pylori result from the testing laboratory was not registered in the hospital health records system.

| H. pylori testing data
Data on positive H. pylori tests between 2010 and 2021 were requested from all hospital and regional laboratory testing facilities throughout the South Island of NZ.The results from all relevant laboratory testing methods were requested (SAT, serum, and biopsy).Serum testing results were included in order to provide a comprehensive picture of testing practices within primary and secondary care settings children, eradication therapy, guidelines, Helicobacter pylori, retesting and adherence to guidelines, as has previously been done. 27The implications of this leading to possible overreporting are discussed as a study limitation.Confirmation was sought, and provided, that UBT was not used in children during this time period.All laboratory and hospital tests are recorded in the tertiary care hospital records system whether ordered from within the primary or tertiary care setting.

| Study outcomes
Study outcome data were collected from available tertiary healthcare records.In NZ baseline, demographic data are available in tertiary health-care records for all patients, but information on clinical encounters taking place within the primary care setting is not accessible to researchers in the tertiary setting.Therefore, data available on all patients (primary or tertiary setting) included age, sex at birth, ethnicity, address at time of diagnosis, and H. pylori testing method and results.For those seen in the tertiary care setting only, key casedefining variables were collected regarding family history of H. pylori infection, presenting GI symptoms, initial and longitudinal H. pylori testing and retesting methods, endoscopic and histologic findings, and the eradication regimen prescribed.Longitudinal outcomes for the 12 months following baseline positive H. pylori test were collected and included retesting, reinfection, re-eradication, gastrointestinal or extraintestinal symptoms, and medications.

| Ethnicity
Ethnicity data of the study cohort were compared to the 2018 NZ census data for the whole country, and against summarized South Island data to account for known differences in ethnicity distribution between the North and South Islands. 28

| Socio-geographic deciles
Socio-geographic domains were assessed using the Index of Multiple Deprivation (IMD) 29 that reports deprivation domains for individual addresses of overall deciles of IMD, employment, income, crime, housing, health, education, and geographic access (Table S1).Each decile is ranked on a scale of 1-10, with 10 being the highest level of deprivation.

| Cost
Baseline and longitudinal testing within 12 months for all patients were assigned unit costs to assess the economic burden to the health-care system of H. pylori testing in children.

| Ethics
This study was approved by the University of Otago Human Ethics Committee (Health) (HD21/048), with locality approval provided by the Canterbury District Health Board (RO21125).

| Statistical analysis
Descriptive data presented as number (%), mean (standard deviation [SD]), and median (interquartile range [IQR]).Individual IMD data were summarized for all children residing in the same District Health Board (DHB) region, their data compared to the known population estimates for that DHB, and then summarized as the percentage difference over or under (observed minus expected) the general population.The mean percentage from all regions were presented. 30tween-group differences were examined using Fisher's exact test, with significance level set at p < 0.05 for all.Data were analyzed using SPSS version 28.0 (2021.Armonk, NY, US: IBM Corp).The mean age of the cohort overall was 13.2 years (SD 4.3), and 105 (44%) were male (Table 1).Baseline sociodemographic data were available on most children from tertiary health records, with additional variables available on only a proportion of the cohort seen in the tertiary care setting.

| Socio-geographic factors
When the IMD deciles 29 for study participants were summarized by DHB region and compared to known population data, it was possible to review the percentage difference between the study cohort, and that expected in the regional general population (Figure 1).When the proportion of children in each decile of each IMD domain were examined (Figure 1), it could be seen that children living in areas with low income, and children living in areas with high crime rates, were more likely to have H. pylori infection, as well as those, conversely, in areas with good geographic access to amenities, and those in areas with better health indicators (Table S1).

| Ethnicity
Data were available for the whole cohort on the ethnicity of each participant (Table 1  overrepresented in the study cohort (Figure 2). 28When compared to the summarized South Island data Pasifika children were also overrepresented and participants identifying as Māori on a par with the census.Limited data on refugee status or country of birth were available, showing that 11 (5%) of the cohort were known refugees, with 36 (15%) overall having been born outside of NZ.

| Primary care setting
One hundred and fifty-five (65%) of the cohort had their baseline testing carried out in the primary care setting and had no data available on related symptoms or medications at the time of testing.

| Tertiary care setting
Eighty-five (35%) children had their baseline test for H. pylori within the tertiary care setting and had data available on symptoms and medications at the time of testing.In this subgroup, symptoms were experienced by all children and for a mean duration of 10.7 months (SD 12.4), median 6 months (IQR 1.5-18), with a range of 0.1-60 months.The main reported symptoms were abdominal or epigastric pain (65% and 45%, respectively), nausea (39%), vomiting or hematemesis (34%), and associated weight loss (24%) (Figure 3).
For the 85 children seen in the tertiary care setting for symptom assessment or H. pylori testing, their referral had been made to the pediatric gastroenterologist for 36 (42%), general pediatricians for 40 (47%), and pediatric surgeons for nine (11%).This represents adherence to the international guidelines to refer to a pediatric gastroenterologist for assessment by only 15% (36 from 240) of the overall cohort.

| Baseline testing
Of the 155 children seen in the primary care setting, H. pylori testing was by SAT for 96 (62%) and serum tests for 59 (38%) (Figure 4).
Of the 24 children that had gastroscopies, 12 (50%) had a positive CLO result confirming H. pylori infection and 16 (67%) had histologic confirmation.Four of the children had multiple positive test results.
Endoscopic and histologic findings for each child with available reports were also examined (Table S2).
Of the whole cohort, 138 (58%) children had positive SAT tests, and 78 (32%) positive serum tests.Only 24 (10%) of H. pylori cases adhered to international guidelines in being confirmed via gastroscopy.

| Baseline eradication therapy
The 155 children who had their testing ordered from within the primary care setting had no records available relating to eradication therapy so were determined to have eradication status unknown at baseline and follow-up.negative (Figure 4).This finding may imply that these children had been prescribed eradication therapy and were retested to determine successful eradication.
The method of baseline test was not associated with the method of initial retest (χ 2 4.7, p = 0.32), nor the method of baseline test with the result of the initial retest (χ 2 1.78, p = 0.41).The method of initial retest was not associated with having a positive initial retest result (χ 2 3.0, p = 0.22).

| Re-eradication
Three children had documentation of being prescribed eradication therapy for a second time within the 12-month follow-up period, but details of the drug combination were not recorded.

| Further retesting
Within the 12-month period following the baseline test, and in addition to the initial retest, 29 (12%) children from the cohort overall were tested for a third time yielding seven positive tests (24%), with these being carried out by SAT for 19 (66%), serum for two (7%), and gastroscopy for eight (27%).Three children were tested for a fourth time (all SAT) yielding one positive test, and two children were tested for a fifth time (both SAT) within the 12-month period with one testing positive.

| Patterns of testing
When the 105 children who were retested throughout the 12-month follow-up period had all their results reviewed, it was seen that 78 had either one or two negative retests and no other positives, 13 had either one or two positive retests and no negatives.Twelve children had either one or two positive retests followed by subsequent negative(s).Two children initially retested negative and then tested positive once again, and for one of these children, this H. pyloripositive status continued for three subsequent tests during the 12month follow-up.

| Cost
For the overall cohort, there was a total of 379 tests carried out during the study period, including baseline testing and all retests within the 12-month period.These comprised 243 (61%) SAT, 83 (22%) serum, and 53 (13%) gastroscopies.When calculated, the sum costs for each test, at each time point, totaled $102,873 NZD, with a mean sum of $206 (SD $499) per patient for baseline testing, and $429 NZD (SD $778) across the entire study period (Table 2).

| DISCUSS ION
This study presented data on a cohort of children in NZ that tested positive to H. pylori infection during an 11-year period.Only a small proportion of children were diagnosed via the national and internationally recognized standard of care.Eradication therapy was successful for the majority with confirmed treatment, but many additional retests were negative for those with no records to confirm eradication, strongly indicating empiric treatment.
Given the number of children in this study retesting negative to H. pylori infection but with no records available to confirm eradication therapy, it must be considered whether an assumption of eradication is correct.The alternative assumption is that spontaneous eradication occurred, although this is unlikely to account for the numbers seen in this study.Previous work has reported a range of spontaneous eradication rates, from 2.2% among children 1-7 years in Brazil, to 2.9% in children aged 7-10 years in China, 31,32 and up to 4.7% per year in a Mexican cohort of children aged 5-13 years. 33It is suggested that spontaneous eradication rates are higher in children within the first 2 years of life, 34 but 92% of the cohort in this study were aged over 6 years.Spontaneous eradication may be linked to incidental antibiotic treatment prescribed for infections other than H. pylori, but this is thought more to influence the duration of H.
pylori infection and only explains a small portion of spontaneous clearance. 357][38] This disparity may be due to a number of factors.Previous articles reporting NZ data have mostly presented results of screening studies, with a report based on the percentage of each ethnicity group testing positive.The screening method is likely to identify children testing positive across a more diverse group than the data in this study.There are known inequities in health access and engagement for Māori and Pasifika in NZ, 39,40 as a result of recognized structural, cultural, and communication barriers. 41Previous studies highlighting that preventable hospitalization rates among these two ethnicity groups are high, in particular for children, indicating that primary care access and utilization is low. 42 context of the findings of this study, this may have reduced the number of children in these ethnicity groups being tested.0][21][22][23][24] The guidelines state that diagnosis should be carried out via gastroscopy only, and eradication therapy prescribed for those who have PUD, or following a risk/benefit discussion for children diagnosed following incidental gastroscopy findings. 7,12,20,21With levels of antibiotic resistance of H. pylori shown to be rising throughout Australasia, 44 and in NZ rising general antibiotic resistance 45 and resistance to H. pylori eradication drugs specifically, 46 strategies to reduce unnecessary exposure should be of high priority.The cost of carrying out unnecessary testing in NZ has been shown to be a substantial burden to the health-care system in the current pediatric study, and among adults in NZ. 38 In addition, serum antibody testing does not distinguish between current infection and past exposure, thereby indicating in this study that 88 children had undergone venipuncture unnecessarily. 47number of factors in this study indicate poor adherence to pediatric H. pylori guidelines.[21][22][23][24] This is not an uncommon finding, as previous research at the tertiary care level reported only 49%-52% were diagnosed via gastroscopy, 27,48 30% underwent testing by a primary care physician prior to referral, and 68% received eradication treatment before tertiary referral, thus not following international guidelines. 27For the adult population, NZ and international H. pylori guidelines indicate that a "test and treat" strategy is appropriate, [49][50][51] thereby highlighting the TA B L E 2 Costs (NZD) associated with Helicobacter pylori testing at baseline and within a 12-month follow-up period for study cohort (N = 240).
Costs were calculated in New Zealand Dollars (NZD).Testing costs were derived from NZ testing laboratories (Canterbury Health Laboratories, MedLab South), and the Canterbury District Health Board where all gastroscopies were carried out.Cost of SAT tests in 2023 were $35.66 NZD (excluding goods and services tax [GST]), H. pylori serum antibody testing in 2020 $49.50 NZD (excluding GST), and gastroscopy costs in 2020 $1700 NZD (excluding GST).

3. 1 |
Description of cohort 3.1.1| Demographics Positive H. pylori test results were provided by testing laboratories on 286 children.Forty-six were subsequently excluded from analysis; 11 lived in the North Island of NZ (with testing carried out in South Island laboratories), 11 had equivocal serum tests in isolation, 15 were over the age of 18, and nine of the test results were not found in the tertiary care records.Thus, 240 children aged between 2 and 18 years with a positive H. pylori test were included in the final analysis.

TA B L E 1
Demographic and health variables for 240 children testing positive to Helicobacter pylori in the South Island, NZ, between 2010 and 2022.
Helicobacter pylori-positive study cohort IMD domain deciles and the known population data, expressed as a percentage difference.Positive percentage values indicate a greater percentage of the study cohort in a decile than the general regional population, negative percentage values indicate a smaller percentage of the study cohort in a decile than the general regional population.IMD, index of multiple deprivation, decile scale: 1 = least deprivation, 10 = highest deprivation.F I G U R E 2 Ethnicity of study patients compared to proportions from New Zealand Census (2018) both as a comparison to whole country data, and summarized South Island of NZ data.MELAA = Middle Eastern, Latin American, African; NZ, New Zealand.

F I G U R E 3
Among the group of 85 children with testing ordered from within the tertiary care setting, 59 (69%) had a record of eradication therapy being prescribed, and 26 (31%) were either recorded as not being prescribed treatment, or no record was available, so were determined as having eradication status unknown.Of the children prescribed eradication therapy, it was noted that 15/24 (63%) that had a gastroscopy-received eradication therapy, and 44/61 (72%) with positive SAT or serum tests were recorded as having treatment.There were 33/59 (56%) children with information available on the eradication therapy prescribed.Twenty-five children were prescribed the combination of omeprazole, clarithromycin, and amoxicillin, with the remaining seven having substitutes of pantoprazole, metronidazole, clindamycin, and erythromycin.Seven children had records of having received eradication prescribed for 1 week, and 13 for 2 weeks.In summary, no records were available to confirm eradication therapy for 181 (75%) children from the overall cohort, and 59 (25%) were prescribed eradication therapy.Proportion of children with Helicobacter pylori experiencing symptoms among those seen in the tertiary care setting (N = 85).F I G U R E 4 Flowchart of Helicobacter pylori testing, initial retesting and eradication for the whole cohort (N = 240).Data presented as N (%).Denominator for total numbers in bold are based on previous flowchart level, denominator for subsequent data based on total for that box.

3. 5 |
Twelve-month follow-up 3.5.1 | Initial retesting One hundred and five (44%) of the overall cohort of 240 children had an initial (first) retest for H. pylori within 12 months of their baseline test.These were carried out a mean of 13.3 weeks (SD 11.8), median 9.3 weeks (IQR 4.7-19.1)after their baseline positive test.The main retesting format was SAT for 81 (77%), serum for 3 (3%), and gastroscopy for 21 (20%).From the cohort of 59 children that had a record of eradication therapy, 39 (66%) were retested and 7 (18%) remained positive, and 32 (82%) had successful eradication.Of the 181 children with no records to confirm receipt of eradication therapy, 66 (28%) had a record of being retested, with 18 (27%) testing positive again, and 48 (73%) testing

Test Number of tests overall Number of SAT tests SAT cost Number of serum tests Serum cost Number of gastroscopy tests Gastroscopy cost Sum Mean (SD)
Abbreviations: SAT, stool antigen test; SD, standard deviation.