The impact of low socioeconomic status and primary health care access on emergency department presentations in young children in regional Queensland

The purpose of this study was to empirically evaluate if children from low socio‐economic status (SES) families in regional southeast Queensland utilise acute care services for low acuity health care rather than utilising primary health services.

1 International literature identifies that families with greater social disadvantage tend to access acute care services more frequently.

What this paper adds
1 The Australian concession/health care card (AC/HCC) is a sensitive measure for identifying low SES individuals and families in Australian research.2 In one regional setting in Queensland, families who held a AC/HCC accessed hospital acute care services more frequently than families who were not eligible for AC/HCCs.3 Accessing primary health services such as child health does not ameliorate the frequency of acute care presentations by families with young children.
5][6] The availability of affordable and appropriate health services influences how family's access and utilise health care.
Poorer health outcomes are often experienced by those living in rural and regional Australia, 7 where access to health services may be limited.The Australian health-care system allows Australian's access to publicly funded hospitals with no associated costs, and access to primary health services such as General Practitioners (GPs) at a reduced fee. 8GPs are medical doctors who work in private practice in the community and who provide generalist medical care across the lifespan.There is variance in out-of-pocket costs experienced by the patient due to individual GP fees.Child health services are community-based publicly funded primary health-care services that promote the health and well-being of young children and families.
They are staffed by Registered Nurses and are available in most Australian communities.0][11][12][13] In contrast, in Queensland Australia, Williams et al. 14 found that although presentations at a major metropolitan paediatric hospital ED were multifactorial, socio-economic factors did not contribute to the presentations.Although Alele et al. 15 reported on a large data set collected over a 4-year period in Cairns, Queensland, families of very high socioeconomic status (SES) were 'twice as likely' to have a low acuity presentation to the hospital ED, compared with low SES families.Both Williams et al. 14 and Alele et al. 15 used the Australian Bureau of Statistics Socio-Economic Indexes for Areas (SEIFA) data and postcode as a means of indicating SES.This is consistent with the international literature where postcode and census data have been used to develop indices that measure a community's level of deprivation. 12,13This block method of calculating SES does not reflect the individual or family status, but more accurately reflects the overall status of the community. 16Block-level measures such as SEIFA data are insensitive to individual circumstances because the measures are unable to account for substantial variations in socioeconomic levels within a community. 17,18Therefore, postcodes in combination with SEIFA data may not be a sufficiently sensitive measure of SES at an individual or family level in regional Australia, where postcodes cover large geographical areas.The Australian Government provides a variety of low-income concession and health care cards for those residents whose income falls below the threshold of the means test for their family situation.Approximately 30% of people in living in regional communities hold an Australian concession/health care card (AC/HCC). 19The AC/HCCs account for the age of card holder, income, the number of children within a family, disability and illness.Therefore, a more specific assessment of low SES for individuals and families within a regional setting could be whether the family reaches the threshold for holding an AC/HCC. 8,16he purpose of this study is to empirically evaluate if children from low SES families in regional Southeast Queensland (SE QLD) utilise acute care services for low acuity health care rather than utilising primary health services, and to further explore the relationships between type and frequency of presentation to acute care services and access to primary health services.

Methods
This study was a non-experimental quantitative research design, conducted at a regional hospital in SE QLD.The medical records of all children who had not reached their fifth birthday and who presented to the regional hospital ED between 1 June 2019 and 31 May 2020 were audited.Ethics approval was gained from Human Research Ethics Committees at Darling Downs Health and the University prior to commencing this project.A medical record audit was identified as the best method available to collect the required information necessary to address the research questions.Using an audit process is a commonly used methodology for collecting retrospective data and provides a systematic format for collecting and collating data. 20The audit tool was developed as a spreadsheet.The Emergency Department Information System (EDIS) was used to identify the medical records of all children who met the criteria.A 12-month review period was chosen to account for potential seasonal variations in health conditions.The initial lockdowns and social restrictions associated with the SARS-CoV-2 pandemic commenced in Australia in March 2020.A small decrease in the number of ED presentations by young children was noted compared to the previous year.However, the impact of the pandemic was limited at this time in regional Queensland, unlike what was being experienced internationally.
The unit record number was used to match the medical record with the presentation as it was recorded on EDIS.A four-digit code was allocated to each medical record in order that the data could be de-identified prior to transfer to the secure data storage site.Each medical record was reviewed against the audit tool.Demographic details, such as age, date of presentation, gender and ethnicity, were collected.The eligibility to hold an AC/HCC was importantly noted as it was used to identify low SES families.The Australasian Triage Scale was recorded as it provided an indicator of the acuity of the presentation. 21Five categories of triage were applied, ranging from Category 1 requiring immediate resuscitation to Category 5 being nonurgent. 21The presenting problem/s as identified by the parent, medical diagnosis and outcomes such as discharge, admission or transfer to other health/hospital services were noted as the information gathered was expected to identify the type and acuity of the problem that contributed to the hospital presentation.Accessing child health services and a GP was viewed as the family utilising primary care health services within the community.Following the audit, a further random selection of 5% of the charts was reviewed against the audit tool to verify the accuracy of the collected data.The data were checked for completeness and accuracy.Any potential errors in the data were rechecked against the medical record.Frequency analysis was applied to the demographic data and Pearson chi-square was used to test the relationship between child health attendance and holding an AC/HCC.Poisson regression was used to explore the relationships between acute care presentations and primary health presentations.

Results
Approximately 1040 children under 5 years of age resided in the main postcode area during the study period. 22A total of 888 children who had not reached their fifth birthday presented to the hospital in the designated 12-month period.About 34% (n = 302) of children lived in families who held an AC/HCC.There was no documentation regarding a GP in 82.1% (n = 729) medical records, whereas 71.3% (n = 633) had accessed the local child health services with the average number of visits to child health being 5.56 (SD = 6.331).No association between attending child health services and holding an AC/HCC was identified (χ 2 (1) = 0.04, P > 0.001).The demographic characteristics of the children are displayed in Table 1.
For the designated period, there were 1691 presentations to the regional hospital ED of children younger than 5 years, with a mean number of 1.91 presentations (SD = 1.483).The low number of ED presentations was evident across all ages.In 83.55% (n = 1412) presentations, the family resided in the local postcode area.Most presentations (50.86%, n = 860) were triaged as Category 4semiurgent presentations.Most children, 95% (n = 1610), were bought to ED by their parents/guardians using private transport, and 88.23% of presentations (n = 1492) were discharged home.The characteristics of the ED presentations are displayed in Table 2.
Poisson regression was applied to identify if holding an AC/HCC or accessing a primary health-care provider such as a GP or child health services predicted ED presentations.A total of three cases, which had presented greater than 11 times to the ED, were deleted from the analysis to ensure equidispersion. 23The likelihood ratio chi-square test indicated that the full model was a significant improvement in fit over a null model (P < 0.001).
Holding an AC/HCC was a significant predictor for the number of ED presentations by a child under 5 years of age (b = 0.171, SE = 0.0481, P < 0.001).It resulted in a 15.75% (7.4-24%) increase, on average, in hospital presentations.
Accessing child health services was a significant predictor for ED presentations (b = 0.140, SE = 0.0638, P < 0.05).There was a 13% (1.4-23%) increase, on average, in ED presentations for children who had accessed child health services.
Having a GP was a significant predictor for ED presentations (b = 0.280, SE = 0.0560, P < 0.001).On a closer analysis of the data pertaining to GP through chi-square tests, crosstabulation and boxplots against ethnicity, concession card and child health, the accuracy of medical record documentation in relation to the child has ever seen a GP was highly questionable.Therefore, no further analysis was conducted into the association between ED presentation and seeing a GP.

Discussion
The purpose of this study was to evaluate if children from low SES families in regional SE QLD utilised acute care services for low acuity health care.Most presentations of young children at ED were for semi-urgent paediatric health concerns and they were discharged home following medical review.The results indicated that having an AC/HCC was associated with a significant increase in ED presentations.0][11][12][13] However, the findings are inconsistent with local results that either higher SES families utilised acute care services more frequently, 15 or that SES had no effect on the utilisation of emergency services for non-acute concerns. 14 major difference between this study and other Queensland research is the measure of SES.Alele et al. 15 and Williams et al. 14 used the block measures of postcode and SEIFA data for measuring SES.The use of block-level data is in keeping with the literature. 16owever, regional and rural Australian postcodes cover large geographical areas, where there is heterogeneity of income, education levels and resources.Due to the population density in major international cities, more homogeneity within an area or postcode can be expected.In this study, the family measure of low SES was holding an AC/HCC.This provided a measure that did not rely on self-report, residential address, or neighbourhood.Therefore, the AC/HCC may be a suitable measure of low SES that is useful with smaller, more diverse populations, and where the individual and family level of deprivation is the variable of concern.Further study is needed to replicate these results in other Australian localities that are rural, regional, and metropolitan.
It was anticipated that accessing primary care services would reduce the number of acute care presentations for a child.Investigation into the possible relationship between GP access and ED presentations was unable to be pursued due to a consistent lack  of documentation within the medical records.The results indicated that accessing child health services was associated with a small but significant increase in ED visits.One possible explanation for these results lies within the family's ability to access and engage with health services.Those families who engage with primary health services such as child health appear in this study, more likely to engage with other health services such as hospital ED services, in order to provide health care for their children.Although this can be seen as positive for the health outcomes of the child whose parents are able to search for and access appropriate health care, it raises questions about the health outcomes for those families who encounter difficulties accessing or engaging with health services, and this warrants further investigation.
Due to the medical records containing both hospital acute care records and primary health records such as child health, the relationship between ED access and primary health-care access was able to be explored with confidence.This study was conducted in a regional area in SE QLD; further studies in other regional and rural areas will be necessary to validate the transferability of these results.The medical record was the only source of data, and hence the results are as accurate as the documentation within the records.These results could have been strengthened by verifying medical record information directly with parents/guardians.

Conclusions
Most families accessed acute care services for their young child for semi-urgent presentations, arrived via private transport and were subsequently discharged home.This study identified that the AC/HCC can be used to identify low SES families and individuals, and that holders of AC/HCC tend to use acute services more frequently than non-concession card holders.Furthermore, families that engage with primary care services such as child health also access acute care services more frequently.Further research is needed to investigate the effects of health service access and engagement on the health outcomes of young children in regional areas.

Table 1
Demographic characteristics of children under 5 years presenting to ED. [Correction added on June 22, 2023, after first online publication: Table 1 has been updated] NESB = non-English-speaking background.‡ Australian concession/health care card holder.ED, emergency department.

Table 2
Characteristics of ED presentations for children under 5 years † 13 Health = A 24-h confidential phone service that provides a health advice to people living in Queensland.‡ Poisons information = A 24-h telephone service that provides information to the public and health professionals about poisonings from plants and animals.Medicines and chemicals.ED, emergency department.